4

qTudinal STRESS OS- re

4 Os
RA Donated In Memory of
y =o
mA Brian Schilling, BSME, MBA
V ix Asay Global Drug Delivery
 Digitized by the Internet Archive
in 2022 with funding from
Kahle/Austin Foundation

https://archive.org/details/machinedesigntheOOO0Odeut
Machine Design
Theory and Practice
Aaron D. Deutschman
Walter J. Michels
Charles E. Wilson
Newark College of Engineering
New Jersey Institute of Technology
 Machine
Design
Theory and
Practice

Macmillan Publishing Co., Inc.

NEW YORK

LONDON
Copyright © 1975, Macmillan Publishing Co., Inc.

Printed in the United States of America

All rights reserved. No part of this book may be reproduced

or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or any
information storage and retrieval system, without permission
in writing from the Publisher.

Macmillan Publishing Co., Inc.

866 Third Avenue
New York, New York 10022

Collier-Macmillan Canada, Ltd.

Library of Congress Cataloging in Publication Data

Deutschman, Aaron D
Machine design; theory and practice.
Bibliography: p.
1. Machinery—Design. I. Michels, Walter J.,
joint author. Ii. Wilson, Charles E., joint author.
III. Title.
TJ230.D45 621.8715 73-14430
ISBN 0-02-329000-5 (Hardbound)
ISBN 0-02-979720-9 (International Edition)

Printing: 13 Year: 45

ESEN = -0-Oe—3e3800-S
 This book was written for the student interested in beginning a study of
the art and science of machine design. Assuming that the reader has already
completed basic engineering and mathematic courses, the authors have
demonstrated how engineering fundamentals are applied to mechanical
engineering design. This course assimilates knowledge from many disciplines,
and this book is oriented toward practical engineering principles that
demonstrate how functional and economically feasible solutions can be
obtained through proper design. Modern analytical concepts, by which
essential mechanical theories can be visualized, have also been included and
derived.
The first four chapters are devoted not only to a review of important
material properties and fabricating techniques but, more significantly, to
establishing how these broad disciplines affect design. The authors recognize
that many engineering schools still offer formal courses in materials science,
engineering metallurgy, and production processes. Nevertheless, it is equally
true that many institutions are leaning in the direction of engineering science
and, consequently, have reduced emphasis on, or virtually eliminated, these
courses from their curricula. Thus, these early chapters represent an intro-
duction to mechanical design as well as a sound, reasonably comprehensive
review of previously studied topics, which now must be applied to the develop-
ment of mechanical systems. Depending upon the background required for
this course, this material may be studied rigorously or serve as an overview
and review. Where courses in material properties and fabricating techniques
are not required or not available, study of the first four chapters should
prepare the student to read with reasonable understanding and compre-
hension the detailed information in later chapters related to specific machine
elements.
The remaining chapters deal with the methods of stress and deflection
analysis and techniques involved in designing and selecting individual
machine parts. In addition, emphasis has been given to the engineer’s ability
to solve relevant problems concerning mechanical systems. These chapters
on specific machine elements, which constitute the bulk of the book, are
largely independent of one another. This permits the instructor a great deal
of latitude in structuring his course and makes this book sufficiently flexible
to be used with a variety of existing course syllabuses without major
modifications.
Although the book has been designed primarily for use by mechanical
engineering students, the thoroughness of the presentation has been main-

Vv
tained at such a level that the professional or the industrial designer will find
it a valuable reference. Much of design involves more than a single method
of solution to a given problem, and an attempt is made to show alternate
approaches. Some methods not covered in the text are illustrated by computer
problems at the end of the chapter.
Well-defined analytical procedures are used and illustrated via carefully
selected example problems. The book also contains illustrative problems
that demonstrate modern methods; for example, numerical, graphical, and
digital methods are employed. The methods have been selected for generality
of application so that the reader may build upon these techniques to solve
problems not treated in a textbook.
The authors wish to express their deep indebtedness to Professor Alfred R.
Holowenko of Purdue University, Professor Barton L. Jenks of Penn State
University, and Professor L. J. Powers of Texas Tech, who read, thoughtfully
critized, and made helpful suggestions for improving the original manu-
script. Their penetrating commentaries (both pro and con) resulted in a
substantially improved book. We greatly appreciate their invaluable
assistance. And for the task of organizing the amounts of typed and illustrative
material into a printable book, we wish to extend our sincere thanks to
Mr. John J. Beck, Technical Editor at Macmillan. Most especially, we are
grateful to him for his patient and kind attitude in working with us. The
authors would invite students, instructors, and practicing engineers to write
them concerning any questions that may arise. All comments and suggestions
would be appreciated.

ASD)
IVE:
Baw:

vi
Introduction 1

1 The design flow diagram 2 / 2 Design analysis 8 / 3 Factor of safety & /

4 Reliability 77 / 5 Cost 13 / 6 Safety 14 / 7 Closure /5 / References 16

Metallurgical Properties of Engineering Materials 17

1 Brittle and ductile fracture /7 / 2 Improving the strength of materials /8 /

3 The iron-carbon diagram /9 / 4 Wrought iron 22 / 5 Cast iron 24 /
6 Heat treatment of steels 3/ / 7 Isothermal transformation diagrams 34 /
8 Hardenability 39 / 9 Heat treatment definitions 45 / 10 Supplementary
metallurgical terms 53 / 11 Alloying elements in steel and their
functions 54 / 12 Classification of steels 55 / 13 AISI-SAE designations for
wrought steel 65 / 14 Wrought aluminum alloys 67 / 15 Aluminum casting
alloys 77 / 16 Copper and copper alloys 7/ / 17 Magnesium alloys 72 /
18 Nickel and nickel alloys 73 / 19 Plastics 74 / 20 Elastomers 76 /
Problems 78 / References 79

Mechanical Properties of Engineering Materials 80

1 Mechanical properties of materials 8/ / 2 Homogeneity 8/ /

3 Elasticity 8/ / 4 Isotropy 8/ / 5 Plasticity 82 / 6 Tensile strength 82 /
7 Stress 83 / 8 Strain 83 / 9 Fracture strength 84 / 10 Proportional limit 84 /
11 Elastic limit 84 / 12 Proof stress 84 / 13 Yield point 85 / 14 Modulus of
elasticity 87 / 15 Hooke’s law 88 / 16 Direct shear strength 88 / 17 Shear
yield point 89 / 18 Ductility 90 / 19 Malleability 9/7 / 20 Modulus of
resilience 92 / 21 Modulus of toughness 92 / 22 Hardness 94 / 23 Endurance
strength /00 / 24 Endurance limit /02 / Effects of material variation—
reliability factor /09 / 26 Influence of size—size factor //0 / 27 Effects of
methods of manufacture-surface finish factor /// / 28 Effect of stress
concentration //3 / 29 A working equation for the endurance limit //9 /
® Effect of some important nonquantitative factors 120 / 31 Creep 134 /
32 Other temperature sensitive properties of materials 135 / 33 Wear 138 /
34 Radiation effects 139 / Problems 140 / References 145

Manufacturing Processes and Design 147

1 Introduction 148 / 2 Primary processes 148 / 3 Forming of plastics /67 /

4 Welding /7/ / 5 Secondary production processes /8/ / 6 Engineering
drawings and dimensions /89 / 7 Definitions of dimensioning terms /90/
vil
 8 Classes of fits /93 / 9 Selecting tolerances /96 / 10 Vague and
superfluous dimensions 203 / 11 Accumulation and nonaccumulation of
tolerances 204 / 12 Tolerance stackup 205 / 13 Statistical determination of
tolerance 205 / 14 Geometric dimensioning and tolerancing 207 /
15 Surface quality 208 / 16 Design hints 2/2 / Problems 23/ /
References 234

Analysis of Stress and Displacement 235

1 The analytical model 236 / 2 The relationship between “‘exact theory”
and the “strength of materials” approach 236 / 3 The nature of stress and
strain 237 / 4 Bending stresses in machine members 242 / 5 Deflection of
machine members due to bending 250 / 6 Shear stress and shear deflection
due to lateral load—shear center 260 / 7 Torsion 265 / 8 Thermal stress 274 /
9 Energy methods—Castigliano’s theorem 278 / 10 Elastic stability 28/ /
11 Numerical methods 290 / Problems 294 / References 299

Theories of Failure Used in the Design of Machine

Elements 301
1 Stress at a point and combined stress 302 / 2 The maximum normal stress
theory 306 / 3 The maximum shear theory 309 / 4 Strain energy and the
distortion energy theory 3/2 / 5 Fatigue failure and the Soderberg
criterion 3/4 / 6 Cumulative damage 325 / Problems 326 / References 329

Shafts, Keys, and Couplings 331

1 Shaft materials 333 / 2 Bending moments and torque acting on a
shaft 334 / 3 Shaft design subjected to fluctuating loads based on the
maximum shear theory of failure 338 / 4 Design of shafts subjected to
fluctuating loads based on the distortion energy theory of failure 340 /
5 Comparison of maximum shear and distortion energy theories of failures
as applied to shaft design 340 / 6 Shaft design equations for fluctuating and
shock loads 34/ / 7 Shaft diameter for example 7-1 342 / 8 Shaft deflection
for example 7-1 345 / 9 Shaft design by computer 353 / 10 Critical speed of
shafts 357 / 11 Torsional stiffness 358 / 12 Torsion of shafts of various cross
sections 359 / 13 Keys 36] / 14 Splines 373 / 15 Couplings 378 / 16 Universal
joints 390 / 17 Flexible shafting 395 / Problems 396 / References 404

Journal Bearings and Lubrication 405

1 Introduction 406 / 2 Types of journal bearings 407 / 3 Newton’s law of
viscous flow (viscosity) 409 / 4 Hagan—Poiseuille law (flow through
capillary tube) 4// / 5 Viscosity measuring devices 414 / 6 Temperature
effect on viscosity 4/5 / 7 Petroff’s law 418 / 8 Hydrostatic lubrication 4/9 /
vill
9 Hydrodynamic lubrication 42/ / 10 Design charts 425 / 11 Heat
balancing of bearings 432 / 12 Methods oflubrication 437/ 13 Bearing
materials 438 / 14 Design considerations 440 / Problems 441 /
References 443

Rolling Bearings 444

1 Bearing nomenclature and types ofball bearings 446 / 2 Roller
bearings 460 / 3 Miscellaneous bearing configurations 468 / 4 Standard
dimensions for roller bearings 472 / 5 Bearing tolerances 478 / 6 Rolling
bearing materials 479 / 7 Rolling bearing friction 482 / 8 Basis for bearing
failure 483 / 9 Life, rating life, and basic load rating 483 /10 Equivalent
load 485 / 11 Bearing survival with probabilities higher than 90 °% 493 /
12 Life adjustment factors for materials 495 / 13 Life adjustment factor for
application conditions 496 / 14 Summary oflife adjustment factors 497/
15 Rolling element bearings subjected to variable loads 497/
16 Lubrication of rolling bearings 500 / 17 Speed limitations of rolling
bearings 507 / 18 Seals 509 / 19 Typical bearing assemblies 5/2 /
20 Selection of tapered roller bearings 5/5 / Problems 5/5 / References 5/8

Spur Gears 519

1 Spur gear terminology 520 / 2 Backlash 524 / 3 Fundamental law of

gearing and velocity ratio 525 / 4 Involute gear teeth 525 / 5 Gear tooth
action 526 / 6 Contact length and contact ratio 529 / 7 Interference 53/ /
8 Standard gear systems 533 /9 Common gear manufacturing methods 534 /
10 Other gear production methods 536 / 11 Gear finishing methods 539 /
12 Tooth loads 540 / 13 Beam strength of spur gear teeth 544 / 14 Stress
concentration 553 / 15 AGMA equation 554 / 16 Surface durability of
spur gears 564 / 17 AGMA gear equation 468/ 18 Gear design 58/ /
19 Backlash control 586 / 20 Gear lubrication 588 / 21 Gear materials 589 /
22 Nonmetallic gears 59/ /23 Gear blank design 592 / 24 Gear trains 594 /
Problems 595 / References 599

Helical, Worm, Bevel, and Other Gear Types 601

1 Helical gears 602 / 2 Helical gear tooth loads 603 / 3 Helical gear
terminology 607 / 4 Virtual or formative number of teeth 6/2 /
5 Dynamic load for helical gears 6/2 / 6 Bending strength ofhelical
gears 613 /'7 Surface durability of helical gears 6/6 / 8 Crossed helical
gears 620 / 9 Worm gears 623 / 10 Worm gear terminology 625 / 11 Worm
gear strength 627 / 12 Dynamic load 627 / 13 Wear load equation 628 /
14 Efficiency of worm gears 628 / 15 Thermal capacity of worm
gearsets 631 / 16 Worm gear standards and proportions 633 / 17 Bevel
gears 637 / 18 Straight bevel gears 639 / 19 Formative or equivalent number

ix
 of teeth 642 / 20 Strength of bevel gears by modified Lewis equation 642 /
21 Dynamic load for bevel gears 645 / 22 Allowable wear load for bevel
gears 645 / 23 AGMA method for designing bevel gears 646 / 24 Tooth
loads on straight bevel gears 651 / 25 ZEROL bevel gears 652 / 26 Spiral
bevel gears 654 / 27 Hypoid gears 655 / Problems 656 / References 659

Belt and Chain Drives 660

1 Belt drives 663 / 2 Chain drives 67] / Problems 675 / References 677

Brakes and Clutches 678

1 Positive contact clutches 678 / 2 Disc clutches 682 /3 Cone clutch 688 /
4 Other types of clutches 691 / 5 Brakes 695 / 6 Energy and power
considerations 695 / 7 Band brakes 697 / 8 Differential band brakes 699 /
9 Disc brakes 700 / 10 Short shoe block brake 700 / 11 Long shoe external -
block brake 702 / 12 Long shoe internal brake 706 / 13 Materials for
brakes 708 / 14 Electrical brakes 709 / 15 Brake actuation 709 / 16 Design
considerations 7// / Problems 7// / References 7/8

Springs 719
1 Torsion bars 7/9 / 2 Helical springs under static and dynamic loads 72/ /
3 Surging of helical springs 729 / 4 Bending and buckling of helical
springs 732 / 5 Extension springs 735 / 6 Leaf springs 737 / 7 Torsion
springs 740 / 8 Rubber mounts 743 / 9 Air springs 745 / Problems 746 /
References 748

Power Screws 749

1 Power screw thread forms 750 / 2 Some definitions 757 / 3 Torque

equation for power screws 758 / 4 Overhauling 76/ / 5 Screw efficiency 762 /
6 Stress considerations for power screws 765 / 7 Ball screws 770 /
Problems 772 / References 775

Fasteners 776

1 Axially loaded rivets 776 / 2 Methods of rivet failure 778 / 3 Lap and
butt joints 78/ / 4 Eccentrically loaded rivets 788 / 5 Screw fasteners 794 /
6 Standardized screw threads 794 / 7 Various types of screws, bolts, and
other fasteners 799 / 8 Stress analysis of screws and bolt 807 / 9 Materials
for screws and bolts 8/0 / 10 Preload and tightening torque on bolts 8/5 /
11 Elastic analysis and fatigue loading 8/7 / 12 Analysis when gaskets are
used with the connected parts 822 / Problems 823 / References 829
Xx
Welds and Adhesive Joints 830

1 Arc welding 832 / 2 Design of symmetrically loaded weldments 833 /

3 Eccentrically loaded joints 836 / 4 Adhesive joints 84/ / Problems 847 /
References 849

Axisymmetric Problems in the Design of

Machines 850
1 The thick wall cylinder 850 / 2 Interference fits 854 / 3 Stresses and
displacements in rotating discs 856 / 4 Energy storage in flywheels 858 /
5 Design based on plastic analysis 859 / 6 Initially curved members 862 /
Problems 864 / References 865

APPENDIXES 867

INDEX 919

xi
 =

a. <= = vam pate

~
oer é \b at a a”
vs oa _—
rng es

e
“S -
e -
7 . ©
tt pan’ Sab
a
ry
3 wv
-

Pt_s<
A 1 ee |: i
evpt
. oy: |:

Ti

ye
Introduction
SYMBOLS

N,, = factor of safety based on N, = factor of safety based on yield

ultimate strength strength

The object of engineering is to provide society with the requirements ofa

modern civilization. Thus, engineering becomes the link that connects and
converts nature’s gifts to man’s needs. It is apparent that it has been with us
since man first walked the face of this earth and struck flint to start a fire or
shaped his first arrowhead. Engineering is not directly concerned with seeking
nature’s truths. This is primarily the province of the scientist. Instead, it
concerns itself with understanding scientific principles and applying them to
achieve a designated goal. In this sense, engineering might be considered an
applied science.
As an applied science, engineering uses scientific knowledge to achieve a
specific objective. The mechanism whereby a requirement is converted to a
meaningful and functional plan is called design. In other words, design is the
formulation of a plan, a scheme, or a method to translate a need into a
satisfactorily functioning device that satisfies the original need. For example,
the builders of superhighways must follow the plans of design engineers.
The manufacturers of plastic extruding machinery likewise must follow the
plans of design engineers. In fact, virtually all technical functions depend
upon design for their successful operation.
Engineering school curricula tend to strong emphasis on courses in
engineering science and mathematics, and the student begins to look upon
these courses as ends in themselves. In reality, these courses are only tools in
the process of design. The student should be aware that just as “‘all roads lead
to Rome,” all engineering disciplines lead to design. Even those students
who will spend their working lives in some aspect of engineering other than
design must realize that they are in some way involved in design. It is therefore
advantageous for the neophyte to become acquainted with design engineering
to be a successful professional engineer.
 Zz
Chapter 1: Introduction

The remainder of the chapter will concentrate on the design process and
those aspects contributing to a satisfactory design.

SECTION 1-1
The Design Flow Diagram
Figure 1-1 is a typical form of design flow diagram. Other types may be
found in published material [1, 2, 3]’ too numerous to mention here. How-
ever, regardless of which reference is reviewed, most flow diagrams will
contain some or all aspects of Figure 1-1, depending upon the product to be
designed and the particular company procedure.
To understand fully all that must be considered in the design process, we
will proceed to explain the characteristics of each box of Figure 1-1.

Recognition of a need
This aspect of design can have its origin in any number of sources.
Customer reports on the product function and quality may force a redesign.
This source is emphasized by the feedback loop of Figure 1-1 that emanates
from the product release box. Business and industrial competition are
constantly forcing the need for new equipment, processes, and machinery
designs. For example, high-speed “‘hot-lead”’ type of printing machinery is
being replaced by higher speed photographic type printers that are computer
controlled. Another example is that of numerical tape control for machine
shop equipment when large quantities of complex parts must be manufactured
to exacting dimensions. Development of patents into engineered products or
incorporating them in overall designs is still another source of needs.
Among the largest generators of needs are the various agencies of the
government. Typical current needs are better mail handling services, which
will ultimately lead to automated post offices, development of a variety of
early warning defense devices and systems for national security, all kinds of
antipollution equipment for environmental improvement and control. This
last area in itself will foster a host of secondary and tertiary needs for years
to come.
Certainly, the reader can think of many other sources of needs that give
rise to engineering design problems. Regardless of the source, the important
thing is to recognize that a need exists, to use all of one’s senses and back-
ground experience to focus on the need, justify its gratification, and gather
as much information as possible concerning it.

Specifications and requirements

Once a need has been established, its requirements must be carefully
spelled out. This step in the design flow diagram is labelled specifications and
requirements. Many engineering organizations label this area design and
‘ All bibliographical material is listed at the end of each chapter.
 3}
Section 1-1: The Design Flow Diagram

Recognition
ofa
need

Specifications
and
requirements

Feedback

Feasibility study
Feedback

Creative design
synthesis

Preliminary design Figure 1-1 Design flow diagram with

and development = feedback loops.

E|Detailed design
Feedback

Prototype o
building and testing

Feedback

Design for
production

performance requirements. Very often the initial phase of a project is stymied

at this point because specifications are stated in such general terms as to
indicate that the client (for example, sales department, consumer, and so on)
has only a vague idea of what he really wants.
On the other hand, various local, state, and federal agencies (particularly
the military and the Atomic Energy Commission) and a variety of sophisti-
cated engineering clients (for example, companies that maintain their own
engineering staffs) submit specifications that are written in great detail,
 4
Chapter 1: Introduction

where both the design and performance requirements are carefully stated. In
fact, whenever health, safety, or any other legal or code requirement applies,
this requirement becomes part of the original specification. In addition,
governmental, commercial, and industrial standards (for example, Unified
Standard Thread, National Electrical Manufacturers Association Standards
for wiring, motors, and so on, American Iron and Steel Institute Material
Standards and many more) are invariably made part of a specification.
Quite often, specifications are tailored to satisfy the production competence
and capacity of the parent company.

Feasibility study
Once the specifications have been prepared, accepted, and submitted,
the next phase of the design flow is to make a feasibility study. The purpose
of this study is to verify the possible success or failure of aproposal both from
the technical and economic standpoint. Various questions must be answered.
(1) Is any natural law being defied? (2) Are somé ofthe specifications beyond
what is technically available at present? (3) Is there a dependency on scarce
materials? (4) Will the end product cost be so high as to completely eliminate
the product from further consideration?
The feasibility study should not be misinterpreted as having the demise of
a product as its purpose. However, it is quite likely that the ““enthusiasm”’ of
the sales department or members of management will, at times, override
potentially severe technical difficulties thereby wasting many man-hours of
design time. Many man-hours may also be wasted by overzealous design
engineers constantly seeking perfection at the expense of the overall project
cost. This does not mean that “anything goes” as long as the project is
completed within the allotted time and cost. It merely means that experienced
judgement must be exercised to determine that the design objectives have
been obtained via good engineering practice and that further effort will be
economically wasteful.
Itis apparent, then, that the individuals responsible for making a feasibility
study are going to be engineers with strong design backgrounds, knowledge
of the engineering sciences, and a good grasp of material usage, production
methods, and sales department requirements. In fact, it is likely that the men
responsible for the feasibility study will be the same engineers who will
ultimately inherit the design responsibility for the overall project.
Quite often, as a result of the feasibility study, changes are made in the
specifications and requirements in order to provide a greater likelihood of
project success. This action is indicated by the feedback loop in Figure 1-1
that goes from the feasibility study back to the specifications and requirements
box.

Creative design synthesis

Once the feasibility of the design is established, the design flow diagram
continues to the next step in Figure 1-1 labelled creative design synthesis.
 ©)
Section 1-1: The Design Flow Diagram

This phase is perhaps the most challenging and interesting part of design.
Here, unless one is restricted from doing so, the designer can act as engineer,
inventor, and artist all in one—for now he is called upon to create.
Creativity can be defined as the synthesis of various new and/or old ideas
and concepts in such a way as to produce an overall new idea or concept (at
least new toits creator). Psychologists and educators have not, as yet, devised
a method for teaching creativity, although methods to stimulate it have been
tried. Creativity is an aspect of human behavior that is still being investigated
by psychologists, and an in-depth understanding is still lacking. It is generally
agreed, however, that all ofus have the ability to create at various competence
levels. Creating involves thinking, and certainly we all think—to a greater or
lesser extent. In this regard, proper educational preparation should enhance
creative processes. It must be remembered, nevertheless, that psychological
handicaps tend to lessen the capacity for logical thought and creative effort
[see references 4 to 10 inclusive].

Preliminary design and development

After the process of creative design synthesis is complete, there will be one
or more possible designs that satisfy the given set of specifications and
requirements. It then becomes necessary to decide which of the “‘solutions”’
to choose for the preliminary design and development stage that is the next
step in the design flow diagram. The bases for making decisions are many
and varied. Techniques range from the “seat of the pants” approach to
complex procedures involving matrix tables, probability theory, and so on.
Anin-depth analysis of decision making techniques can be found in references
[9] through [15].
Once a choice has been made we enter the domain entitled preliminary
design and development. At this stage, layout drawings showing machine or
system interfaces are made to determine the overall configuration and to
establish functional relations between various parts of the machine or
system. These layout drawings will contain important dimensions and
notations as well as auxiliary and sectional views completely explaining the
proposed design. In addition, kinematic studies are made that include
complete machine layouts and machine cycle diagrams. Note that this stage
is primarily concerned with detailing the results of creative design synthesis
on drawing paper in order to validate the functional and overall size require-
ments ofthe specifications.
Rarely are all specifications and requirements achieved during this phase.
Therefore, referring back to Figure 1-1, we see a feedback loop from the
preliminary design and development box to the specifications and require-
ments box, indicating that it is necessary to reduce the stringency of some
specifications (if possible) in order to carry the design to completion.
In addition, while making the preliminary design layout, further develop-
ment work may be required to prove an idea, to determine some material
properties, to evaluate a device, or to determine some unknown quantity
 6
Chapter 1: Introduction

based on substantive technical information or experience. Thus, certain areas

ofthe preliminary design might be delayed or halted pending further develop-
ment data. In fact, the need for development work may also arise during the
next step of the flow diagram.

Detailed design
Detailed design is concerned with the actual sizing and dimensioning of
all the individual components, both fabricated and purchased, that go into
making the total product, device, or system. Here individual orthographic
drawings are prepared for each component, showing the necessary views and
all the dimensions and tolerances, the material and heat treatment (if any),
the quantity of such components per assembly, the name of the components,
and perhaps the assembly drawing number where the component part is
used. Many companies and government agencies follow a set of drafting
standards and procedures that may include more data than 1s stated above.
Nevertheless, the primary criteria are that the information given on the
drawing must be such that the shop knows specifically how the part is
going to be fabricated. It is apparent that dimensions cannot be in any way
redundant, for then the machinist making the part would have a choice
of dimensions and the decision he made might be wrong. At the same time,
no dimensions or information should be missing that make it impossible
to fabricate the component. A short discussion of dimensioning will be
found in Chapter 4, Sections 4-6 to 4-14 inclusive.
Usually a draftsman or junior designer makes these drawings under the.
supervision ofthe design engineer. The design engineer, in turn, must provide
the necessary sketches, data, and back-up information. In order to obtain
this information, the design engineer, working with the preliminary layouts,
must size the parts, choose the materials, specify the commercial components,
and so on, based on the analytical techniques available to him and his store-
house of past experience. This means that he must draw upon his back-
ground’ of mathematics, mechanics, strength of materials, fluid mechanics,
kinematics, vibrations, metallurgy, and shop processes. At times, he may
have to seek assistance from experts in specialized fields. As stated earlier,
the major portion of this text will be devoted to the techniques for selecting
the machine elements comprising a design.

Prototype building and testing

After completing all the details, the subassemblies and assembly drawings
including the materials and parts list, the completed design is sent to the
prototype or model shop for fabrication. According to Figure 1-1, the
completed design is now ready for prototype building and testing.

* Note that the design engineer is not only involved in innovative processes; he is also required
to have and to use almost all of his engineering education. It is apparent that design engineering
can be considered the zenith of engineering work. The individual at work in this area is expected
to be well prepared both in mathematics and engineering science.
 ys
Section 1-1: The Design Flow Diagram

At this juncture, the parts are fabricated, commercial components are

purchased, and the machine or system, after having been assembled, is ready
for evaluation and testing. The output of this test period may or may not
yield data that requires changes or modifications in the preliminary design
or detailed design area. This possibility is indicated by the feedback loop of
Figure 1-1. After the changes and/or modifications are made, when necessary,
the new components are incorporated into the prototype assembly for
continued testing and evaluation. The procedure of making continuous
revisions and improvements to the design is repeated until the design engineer
is satisfied that the performance specifications have been met. At this point,
all the drawings, parts, and materials lists are sent to the production
engineering department where the drawings are modified so that the com-
pleted project can be designed for production.

Design for production

Consideration is now given to design changes that would be compatible
with the best (often the most economical) methods of production. In modern
day nomenclature, this is sometimes called value analysis [see references 11
and 12] and is becoming an ever-growing and more important consideration
in design. For example, the production engineer may consider a part suitable
for manufacture by stamping, casting, or, perhaps, forging. For large
quantity manufacturing, any of these processes would be more economical
than individual part machining. Of course, in making a decision, considera-
tion must be given to the necessary tooling costs and whether they can be
amortized over the quantity produced.
Another example of what the production engineer may look for is the
possibility of combining several fabricated parts into one or replacing some
parts with commercially available equivalents. Still another consideration of
the production engineer is the possibility of replacing some of the materials
with equivalently satisfactory but less expensive materials. When the pro-
duction drawings have been completed, they are sent to the production
department for product release.

Product release
Production prototypes are usually made and tested, and any malfunction
that cannot be easily corrected is generally referred back to preliminary
design and development or to detailed design for alteration. This process is
indicated by the feedback loop originating in the product release box of
Figure 1-1.
By no means is the previous description complete in all its detail nor
is it the only path that is followed in the design of a product, device, or
system. However, to appreciate fully the aforementioned description of
the design process, it is necessary to be involved with the procedure on a
daily basis.
 8
Chapter 1: Introduction

SECTION 1-2
Design Analysis
Once a choice of a design has been made, preliminary and subsequent
detailed design takes place as shown in Figure I-1. At this time it is necessary
to draw the layouts, provide the details, perform the tests, make the support-
ing calculations, and so on, that will ultimately result in prototype design. This
is the area where the designer must specify dimensions, select components
and materials, and in general consider such things as methods of manufacture,
cost, reliability [see references 13 and 14], serviceability, and safety. The
designer must rely upon his analytical ability and training in the engineering
sciences to accomplish these objectives.
At this point in the discussion it is very necessary to understand that the
mathematical model chosen and the subsequent calculations that are made
merely approximate reality. One must, therefore, be fully aware of the
various assumptions and limitations (for example, oflinearity, homogeneity,
and so on) that were made in deriving the equations used in the study of the
engineering sciences. The designer, in his anxiety to apply the appropriate
equations to his mathematical model, may have so oversimplified his model
that it no longer represents the real case.
Therefore, it is very important to keep in mind that good design is based
upon good theory while emphasizing that the numbers that result from
applying the theory are merely “ball park’’ figures, which give the design
engineer a rational basis for substantiating his work. Unfortunately, not all
topics in design havea firm analytical base from which to work. In those cases,
we must depend upon a semirational or empirical approach to solving a
problem or selecting a design element.

SECTION 1-3
Factor of Safety
In the light of what has been said in the aforementioned paragraph
concerning analytic models, it seems reasonable to provide for the uncer-
tainties associated with any design based upon such models. In addition, a
designer may at various times also have to consider the following additional
uncertainties :
1. Variations in material properties. Because no two furnace melts are
exactly alike and some materials may have inclusions, and so on, the
strength properties given in materials tables are usually average values.
Ifthe value is a stated manufacturer’s value, it probably is the minimum
value.
2. Effect of size in stating material strength properties. Property tables,
unless otherwise stated, list strength values based upon a }-in. test
 9
Section 1-3: Factor of Safety

specimen. Yet larger components generally fail at a lower stress than a

similar smaller component made of the same material.
3. Type of loading. A simple static load is easy to recognize, but what of
those cases that fall between impact and suddenly applied loads’?
How does one take account of infrequently applied fatigue loading
mixed with some shock loads (for example, cams, links, or feeding
devices)?
4. Effect of machining or forming processes. These production operations
may, and usually do, introduce stress concentrations and residual
stresses.
5. Effect of heat treatment upon the physical properties of material. I[m-
proper heat treatment cancause residual stresses and cracks. Inaddition,
the actual yield stress may be quite different from that used in making
the design calculations.
6. Effect of wear upon the functions and life of a machine member. Constant
rubbing without proper lubrication can appreciably reduce the working
life and must be considered.
7. Effect of time and environment in which the device is expected to operate.
Certainly great caution must be taken when components are operating
in radioactive regions or corrosive atmospheres. When a material is
expected to be subjected to creep (that is, loaded member at an elevated
temperature for a long period of time), one must provide for this
contingency. Operation at lower than normal temperatures is also a
condition that often must be considered.
8. Specific requirements for life and reliability. For example, a machine gun
must be reliable but usually has a finite life. Certain machines, however,
may have an almost infinite life through which they are not wholly
reliable but are expected to be periodically repaired.
9. Overall concern for human safety. All designs must consider safety of
the operator and other persons who may be near or in contact with the
machine or device. Unexpected, sudden overloads may cause breakage
and considerable bodily harm.
In order to take these listed uncertainties into account in design, engineers
have introduced what is familiarly called the factor of safety. (Some engineers
think it should be called “‘the factor of ignorance.’’)
Weare sure that the reader has, in the past, employed or calculated a factor
of safety. For example, dividing the yield point stress by the calculated stress
results in the factor of safety. This appears to be a simple and straightforward
procedure to follow; unfortunately, things are never quite that simple. The
designer must be fully aware of what he means when he calculates such a
factor or when he bases his design on sucha factor. Improper use ofa “‘safety
factor’’ may result, in some cases, in a needless waste of material or, in other
cases, in physical or operational failure. Thus, one must define what he means
when he uses the factor of safety.
 10
Chapter 1: Introduction

For ductile materials, assumed to have the same ultimate and yield stress
in both tension and compression, we say
ultimate stress
N (design)
. =

working or design stress 1-1

va
Ne desione yield stress (1-2)
ae working or design stress

Most often, equation (1-2) is used because mechanical equipment is fre-

quently considered nonfunctional if some important component has yielded.
A typical example of such a failure would be the local permanent yield that
might be caused by a cam follower on a cam surface.
If a machine or mechanical component has already been sized (that is,
its dimensions are known), then the factor of safety is defined as

ultimate stress
N,(actual) = (1-3)
calculated stress
yield stress
tual) = (1-4)
SEOs calculated stress
For nonlinear types of problems, such as columns or rods subject to
failure by buckling, the yield or ultimate stress can no longer be used.
Instead, the actual failure load is used as the basis for a factor of safety. Thus,
we have
Nace failure load (1-5)
pee calculated load

Mechanical components subject to a continuously varying load have their

factor of safety based upon the endurance limit of the material. However,
because of the cyclic and static load that may exist, the definition depends
upon the Soderberg fatigue analysis, which is covered in Chapter 6, Section
6-5.
Joseph P. Vidosic [15] suggests the following factors of safety as being
reasonable. These factors are based on yield strength.
1. N = 1.25-1.5 for exceptionally reliable materials used under con-
trollable conditions and subjected to loads and stresses that can be
determined with certainty. Used almost invariably where low weight is
a particularly important consideration.
2. N = 1.5—-2 for well-known materials, under reasonably constant
environmental conditions, subjected to loads and stresses that can be
determined readily.
3. N = 2-2.5 for average materials operated in ordinary environments
and subjected to loads and stresses that can be determined.
4, N = 2.53 for less tried or for brittle materials under average conditions
of environment, load, and stress.
 IL
Section 1-4: Reliability

5. N = 3-4 for untried materials used under average conditions of

environment, load, and stress.
6. N = 3—-4should also be used with better known materials that are to be
used in uncertain environments or subjected to uncertain stresses.
7. Repeated loads: the factors established in items | to 6 are acceptable
but must be applied to the endurance limit rather than the yield strength
of the material.
8. Impact forces: the factors given in items 3 to 6 are acceptable, but an
impact factor should be included.
9. Brittle materials: where the ultimate strength is used as the theoretical
maximum, the factors presented in items | to 6 should be approximately
doubled.
10. Where higher factors might appear desirable, a more thorough analysis
of the problem should be undertaken before deciding upon their use.
In some cases, the selection of the factor of safety is stipulated by code or
contract requirements. For example, the ASME? Unfired and Fired Pressure
Vessel Code, the ASME Pressure Vessel Code for Nuclear Vessels, various
building codes and specific values that are stipulated in contracts for both
civilian and governmental designs.
It is apparent that the selection of an appropriate factor is rather empirical
and very much dependent upon an individual’s or industry’s accumulated
experience. Where a product or device has a long history of use, the factors
based upon such a history are reliable. In fact, one may still depend upon
such data even though modifications in design and materials have been made.
Statistical methods [16, 17] have also been employed in establishing a
factor of safety. Here, accounts taken ofthe variance in both the dimensions
and strength of a mechanical component. This approach results in a factor
of safety that, in general, is smaller than that based upon pure judgement.
However, this method also requires estimation of possible load and strength
variations, thereby making the method somewhat less than rational. Never-
theless, the statistical approach should be more than just of passing interest
to the modern designers, particularily in those areas where experience data
for certain components have been accumulated. Moreover, this method
permits the use of a relatively low factor, if a small percentage of failure is
acceptable. Unfortunately, space limitations do not permit a detailed
discussion, but the reader 1s urged to investigate the given references.

SECTION 1-4

Reliability
Consumer products, industrial machinery, and military equipment are
intently evaluated for reliability of performance and life expectancy.
Although the “military” and particular industrial users (for example, power

3 American Society of Mechanical Engineers.

 WY,
Chapter 1: Introduction

plants—both fossil fuel and nuclear fuel) have always followed some sort of
reliability programs, consumer products have of late received the widest
attention and publicity. One of the most important foundations for product
reliability is its design, and it is apparent that the designer should at least be
acquainted with some of the guidelines.
The article entitled ‘“‘A Manual of Reliability” [13] offers the following
definition of reliability: ‘‘Reliability is the probability that a device will
perform without failure a specific function under given conditions for a given
period of time.’ From this definition, we see that a thorough and in-depth
analysis of reliability will involve statistics and probability theory. Because
of space limitations, we cannot enter into a detailed study of this subject. We
do, however, recommend references [13] and [14].
As a guide to help the design engineer in producing a reliable product,
Tangerman [13] suggests the following considerations :
A. Product Requirements
1. Are all functional, reliability, and other requirements specified?
2. What are the environmental requirements? Are they reasonable,
based on experience? Are they based on measurement or conjecture?
3. What are the reliability requirements? Are they too severe or too
lenient? Consistent?
B. Preliminary Design
1. What proven design can meet functional requirements?
2. What standard components and assemblies can be used?
3. Does this environment differ enough to affect factors | and 2?
4. How much environmental extrapolation is necessary?
5. Is expert advice available?
C. Design Analysis
1. How does each component and material behave under these
environments?
2. How dependable are available life data?
3. Canreliability be calculated from available data? Do gaps exist and,
if so, can they be filled?
4. Can complete units be built for test?
5. What are the weakest links in the design?
6. Is reliability high enough or is redesign indicated?
D. Corrective Action
1. Will expert technical assistance help?
2. Can manufacturing or quality control give advice?
3. Is the reliability set by one or two components? If so, can they be
redesigned or derated, or is redundancy the answer?
4. Can the environment be changed—heating, cooling, shock
mounting, shielding?
5. Is redesign indicated?
(Steps C and D may have to be repeated several times)
 13
Section 1-5: Cost

E. Final Design
1. Can production, inspection or purchasing help in writing the
specifications?
2. Can specifications be written to assure 100% test and inspection?
3. If component characteristics cannot be 100% tested, are there
suitable manufacturing and quality control procedures?
4. Which component will be subcontracted or purchased? Is there a
list of approved vendors?
5. Can procedures for inspection and test be inserted to reject defective
parts early in manufacturing?
6. What minimum number of tests and inspections must be made at
each stage? Must all characteristics be tested?
7. Will a “shakedown” test eliminate substandard units more easily?
8. How much testing can be done safely without cutting product life
appreciably?
F. Redesign After Pilot Run
Must be carried back to Step C.

SECTION 1-5
Cost
Without doubt, cost is an extremely important factor in most (if not all)
designs. Achievement of an economical design depends on the designer’s
experience, knowledge, ingenuity, and ability to “trade-off” one design
parameter against other design parameters.
In fact, the desire to get one’s ““moneys worth” has developed into a new
engineering methodology called value analysis. Its origin is ascribed to L. D.
Miles, and the various techniques employed are explained in his book
Techniques of Value Analysis and Engineering [11]. Briefly, value analysis is
the organized procedure of cost reduction covering the phases of design,
production, materials, and distribution while maintaining product reliability.
Although we cannot here enter into the detailed aspects of value analysis,
an illustrative example should serve to convey its objectives and what may
be accomplished by the application ofits techniques. Figure 1-2 of a bracket
shows the before and after results of value analysis and was taken from a
paper by R. N. Mooney [12].

The cast iron bracket, part of a signaling device, was found to be one of
the more expensive parts of the assembly. It was being used as part of a
drive sprocket sub-assembly. The contributing factor to the high cost was
the finishing required for certain internal and external surfaces. In fact, the
unfinished casting by itself was only five per cent of the total cost of the
part.
Value Analysis showed that a simpler bracket design could be used if it
were relocated at a different point in the assembly.
 14
Chapter 1: Introduction

Figure1-2 Before and after value analysis. [R. N. Mooney: Savings through
use of value analysis techniques in engineering. ASME Paper No. 68-DE-44,
1968. }

As a result of using the simpler bracket, a saving of fifty per cent was
achieved for the bracket. And, as a result of its relocation to a new drive
position, some parts, including a gear, shaft and two bearings were
eliminated. A one hundred per cent saving was effected by the elimination
of these parts.
The overall saving for the entire project was thirty per cent.

The type of objectivity involved in value analysis should be included in a

good designer’s ‘“‘tool box.’ This means that in achieving a functionally
satisfactory design, where possible, the least expensive materials that are
consistent with the life and wear of the part as well as its operating environ-
ment should be used. In addition, other cost considerations such as method
of manufacture, tolerances, use of standard or commercially available
components, method of assembly, tooling, quantity to be produced, and,
finally, simplicity of maintenance should always be evaluated for every
design. Lastly, an attempt should be made to keep to a minimum the number
of components that constitute an assembly.

SECTION 1-6
Safety
As in the case of reliability, equipment and machinery must be designed
so that they are safe both for the operator and the surrounding community.
For fired and unfired pressure vessels, the ASME code is usually the mini-
mum standard of safety required. In the nuclear power field and related
areas, the AEC* demands that specific safety requirements be followed both
for design and operation. In mining, the Bureau of Mines has rigid safety
requirements. These requirements are, at times, loosely enforced for lack of
adequate inspection. Recently, the automotive industry has been required to
design safer automobiles although specific government standards are still in
* Atomic Energy Commission.
 3)
Section 1-7: Closure

a state of flux. Also, we should not forget that the military services require
that definite specifications concerning safety be followed in supplying their
equipment. The American National Standards Institute in New York, The
National Safety Council in Chicago, The Bureau of Mines, The Bureau of
Labor Statistics, and The National Bureau of Standards all located in
Washington, D.C., are good sources of safety information and statistics.
The vast American industrial complex, however, has, in the past, taken a
rather pedestrian view of the general problem of equipment and machinery
safety as it relates to the operator and the plant employee. The designer can
play an extremely important role in providing adequate safety protection
for the worker by providing the necessary safeguards at the earliest possible
design stage.
A brief list of items that the designer should consider in designing equip-
ment for operator safety follows.
1. Covers or enclosures should be provided for moving components that
are in close proximity to the operator.
2. Parts, which may cause injury to the operator (for example, clothing
getting caught and so on), should not project from the equipment.
3. The design should be such that any adjustments, lubrication, or general
maintenance can be performed with little difficulty or hazard.
4. The equipment or machine should be inoperative as long as the
operator’s hands, feet, arms or other parts of his anatomy are in a work
zone (for example, the working area of a press).
. Sharp corners and edges should be avoided.
— . Electrical equipment should be properly enclosed and grounded.
ne)
7. Natural or forced ventilation (if required) should be provided where
the atmosphere is contaminated with fumes, dust, or other particles.
8. Provision should be made to avoid exposure to various forms of
radiation (that is, x rays, ultraviolet rays, radioactive materials, and so
on).

This list indicates only a small portion of the number of hazards that a
designer must provide against for adequate safety. For each specific design,
the engineer must educate himself concerning the singular peculiarities of
his problem by referring to the appropriate codes and/or standards.

SECTION 1-7
Closure
In this chapter, we have given a brief description of design methodology
and several important associated aspects. As with all endeavors, profes-
sionalism and maturity are attained by practice and experience—some of
which will be realized in a machine design projects course. Unfortunately,
because of space limitations, we cannot here become involved in the broader
 16
Chapter 1: Introduction

aspects of design. We can only hope that this introduction and the references
cited will aid the student to think more logically about design procedures.
The remainder of this text will therefore concentrate on that phase of
machine design that will provide the student with sufficieat background to
select various machine elements properly as well as enable him to perform
design analysis. However, past experience in teaching machine design has
indicated that most students welcome a review of engineering materials,
methods of manufacture, and strength of materials. Thus, parts of Chapters
2, 3,4, and 5 are presented with this view in mind.

REFERENCES
[1] M. Asimow: Jntroduction to Design. Prentice-Hall, Inc., Englewood Cliffs, N.J.
1962.
2)
Joseph P. Vidosic: Elements of Design Engineering. The Ronald Press Co., New
York, 1969.
| D.H. Edel, Jr. (ed.): Introduction to Creative Design. Prentice-Hall, Inc. EARN eS
Cliffs, N.J.,1967.
| N. L. Munn: Psychology. Houghton Mifflin Co., Boston, 1946.
M.S. Allen: Morphological Creativity. Prentice-Hall, Inc., Englewood Cliffs, N.J.
1962.
| Eugene Raudsepp: Forcing ideas with synetics. Mach. Des., pp 134-139 (October
16, 1969).
John E. Arnold: Useful creative techniques. Creative Engineering Seminar,
Stanford University, 1959.
[8] A. F. Osborn: Applied Imagination. Chas. Scribner’s Sons, New York, 1963.
[9] Leo Spector (ed.): Denovate. Mach. Des., pp 20-28 (April 3, 1969).
)} D. L. Marples: The Decisions of Engineering Design. The Institution of Engineering
Designers, London, England, July 1960.
L. C. Miles: Techniques of Value Analysis and Engineering. McGraw-Hill Book
Co., New York, 1961.
R. N. Mooney: Savings through use of value analysis techniques in engineering,
ASME Paper Number 68-DE-44, 1968.
| E. G. Tangerman (ed.): A manual of reliability. Prod. Eng., pp 65-96 (May 16.
1960).
Robert Lusser: A study of methods for achieving reliability of guided missiles.
Technical Report No. 75, U.S. Naval Air Missile Test Center, July 10, 1950.
| Joseph P. Vidosic: Machine Design Projects. The Ronald Press, New York, 1957.
| Ray C. Johnson: Optimum Design of Mechanical Elements. John Wiley and Sons,
Inc., New York, 1961.
7) Hsuan-Loh Su: Design by quantitative factor of safety. Trans. ASME. J. Eng. Ind.,
p 387 (Nov. 1960).
 Metallurgical Properties
of Engineering Materials
SYMBOLS

S, = ultimate strength, psi ASTM = American Society for

S;, = endurance limit for polished Testing Materials
specimen, psi BHN = Brinell hardness number
AISI = American Iron and Steel SAE = Society of Automotive
Institute Engineers
ASM = American Society for Metals

This chapter will review ferrous materials and nonferrous materials, the
fundamentals of heat treatment, and plastics from the designer’s point of
view. In addition, the reader will be reacquainted with the various properties
of materials that are so important in making strength and dimensional
calculations for machine elements.
The young designer often favors a more analytical approach to design,
often overlooking the significance of selecting a material from both its
functional and economic standpoint. Thus, a review of the subject matter
presented in this chapter will serve to emphasize how a viable as well as an
economic design can be achieved.

SECTION 2-1
Brittle and Ductile Fracture
Metals fracture in one of two ways, (1) brittle fracture and (2) ductile
fracture. Brittle fracture occurs with virtually no plastic flow or reduction of
area. Separation takes place along cleaveage planes and appears as bright
granular surfaces. Brittle fracture will show the fracture planes to be perpen-
dicular (or nearly so) to the acting tensile force. Furthermore, the normal
stress on the plane of fracture will be higher than any other plane. In addition
Iv
 18
Chapter 2: Metallurgical Properties of Engineering Materials

to the type of material, dynamic loading and low temperatures can also be
causes of brittle fracture.
In ductile fracture, there is plastic flow with separation taking place in the
direction ofthe highest resolved shearing stress. Tensile testing of such metals
produces cup and cone fractures, and the appearance at the point of fracture
is silky and smooth. Fractures ofthis kind are called transcrystalline fractures
because slip takes place within the grain.

SECTION 2-2
Improving the Strength of Materials
There are basically five ways in which the strength of metals may be
increased: (1) growth of metal whiskers, (2) smaller grain size, (3) strain
hardening, (4) alloying, and (5) heat treating.
1. The larger the specimen, the greater the propensity for crystal defects
to exist. Thus there is a greater likelihood of failure at a lower strength.
Presently much research is being carried out with whisker metals that are
made in the laboratory. These metal whiskers are needlelike crystal filaments
only a few millimeters in length and from | to 10 microns (2) in diameter.
Being small in size, they are free of defects and their strength is nearly that of
their theoretically calculated values. For example, the tensile strength of
commercially produced iron is about 100,000 psi (100 ksi). Laboratory
made iron whiskers, however, have been tested to 2,000,000 psi (2000 ksi).
The results produced to date in the case of many metals made in laboratories
have been quite amazing. The concept, nevertheless, requires much further
study and investigation before it will evolve as a commercially feasible
method.
2. A smaller grain size in a metal improves its strength by providing a
longer and more complex path for slip lines to traverse. The most important
method for controlling grain size is by controlling the rate at which a metal
is cooled. Rapid cooling produces a fine grain. A slow rate of cooling results
in a coarse grain. There are a number of ways to control the cooling rate. For
example, pouring a melt into a sand mold (sand casting) and allowing it to
air cool will produce relatively coarse grains. On the other hand, pouring a
melt into a metal mold and allowing it to cool produces a finer grain. Still
more rapid cooling can be provided by coolants like water or oil, which are
circulated through the mold (die casting). This procedure produces a very
fine grain structure.
3. Strain hardening (also called work hardening or cold working) is a way
of producing a small grain size by mechanical means. Typical of such an
operation is the cold rolling of carbon steel bars to a specific diameter in a
rolling mill. The material is forced to deform in accordance with its slip
system as earlier explained. However, its resistance to slip increases as the
process of deformation continues. The result is a smaller grain size, an
 19
Section 2-3: The Iron-Carbon Diagram

increase in hardness and yield strength, but a loss in toughness and ductility.
The loss of ductility and toughness due to strain hardening can be recovered
by a process called annealing, which is discussed in Section 2-9.
It is possible for hot working to reduce the grain size of a metal and still
avoid the ductility and toughness loss attributed to cold working. In this
process, a metal is shaped and formed when it is above a certain temperature,
known as the recrystallization temperature. A typical example of hot work
is the forging of a steel component (for example, axles, shafts, cutlery,
horseshoes, and so on). Proper control of the forging loads and temperature
produces a metal that is both strong and tough.
4. The intentional addition of small amounts of ametal or metals to a base
metal is called alloying. These additions are made not only for purposes of
increasing the strength of the base metal by heat treatment but also toimprove
corrosion resistance, electrical properties, ductility, machinability, tough-
ness, andso on. When analloyed metal is not heat treated, it is called a simple
alloy.
There are three forms of alloy compositions. These are mechanical
mixtures, solid solutions, and intermetallic compounds. (a) When two
constituent metals are not soluble in their solid states, they form a mechanical
mixture. This type of alloy is one in which each component metal retains its
own properties and crystal structure. (b) When two metals are soluble in
their solid states, they form an alloy that is a solid solution. In forming a solid
solution, the alloying atoms may randomly replace lattice atoms ofthe base
metal. This is called a substitutional solution. It is also possible for the
alloying atoms to randomly locate themselves within the lattic structure of
the base metal. This type of solution is called an interstitial solution. (c) When
the alloying atoms replace the base metal atoms by proportional quantities
and are located in a regular manner rather than randomly, the alloy is called
an intermetallic compound. Alloys of this kind are generally higher in tensile
strength, less ductile than their constituents, and tend to be more resistant
to deformation than their constituents.
5. The most important way to increase the strength of a metal 1s by heat:
treatment. Heat treatment is defined as the controlled heating and subsequent
cooling of a metal or alloy. It is employed to obtain properties that are
desirable and appropriate for a particular application. The subject of heat
treatment is so important that we shall devote Sections 2-6 to 2-10 inclusive
to its detailed discussion.

SECTION 2-3
The Iron-Carbon Diagram
The iron-carbon equilibrium diagram is the most important representation
of an alloy system the designer will encounter. It provides a complete picture
of phase relations, microstructure, and temperature for the knowledgeable
 20
Chapter 2: Metallurgical Properties of Engineering Materials

cc
1600

Molten alloy
ASN ea ——
1500

(0.10%) 7]Mio 18
Delta iron Liquids, primary austenite a
2600
and austenite begins to solidify a 1400
r 2552 F 1 | Cementitie] +
2400 (1400'C) _| begins to £
Solidus, freezing is complete ox Mushy stage solidify 1300
| | burning’ occurs on reheating BS ustenite in
| liquid)
2200 eal : r al E (1.98%; 2075°F or 1135°C) eat
® 7 I 1 Tt + b =
2 Austenite \ 3° © =]
& ’ ; i /| 2066 F(1130 C) ! C(4.3%) 1100
3 Z0ue Reale Road - Caste (/ Solidus; ledeburite freezes | at
Ee in gamma iron “yy i : Ms
5 ee sy t a E(2.0%) t
Graphite in 7 1000
1800 ese ++ gamma iron A. : 7
1670 F 7) cm Austenite, ‘
gh 19105 Limit of solubility, ee 1 Guabieene =. ao
1600 > carbide in gamma iron7 Jl CeMmentite et
7 J Austenite, [ x
Magnetic ) +4 and cementite =: T L, | a 200
point A> ~ lron-graphite line | ‘
1414 F N f eS (Sper SS ee ee eS ee eee
(768°C) 7K y Sa1360° F (738°C) A A, 33 1333°F : (725°C)
.
a= ° tS)
Rear) | Ay ee “a6 (i: 700
Solid solution of - ie Austenite transforms to pearlitel 2 |}
lage carbon in alpha iron 2 s\ 3

Pearlite a
5 =
Pearlite
pete
Cementite, pearlite iF
Sb 3|_
2
gis 600
PS
1000 T and rz and + =f and iG wee r =] S|
ferrite 5 Cementite transformed ledeburite =| 2 2] 500
7 + 2=J aS { et ee
ial te :| EL §
7]
Ww 3) o o

Ag, magnetic change in cementite, 410 F(210 C) a pa Ss

ee pp a a — es eS i

1 |_| I. gl
34
(0.008%) sie |
0% 0.5% 1% 2% 3% 4% 5% 6% 7%
% Carbon
Hypoeutectoid Hypereutectoid
Irons Steels Cast irons

Figure 2-1 The iron-carbon equilibrium diagram.

heat treatment of steel. In addition, it clearly indicates the division between

steel and cast iron, depending on carbon content. Figure 2-1 shows an
iron-carbon diagram complete with the accepted standard nomenclature and
lettering for various points. To be strictly correct, the diagram should be
called the iron-iron carbide diagram because the carbon in equilibrium does
not appear as free carbon but in the form of iron carbide (Fe,C) known as
cementite. Common usage, however, terms it the iron-carbon diagram.
On close inspection of Figure 2-1, we see that pure iron (that is, 0% carbon)
has several allotropic’ forms. At room temperature, the iron has a body-
centered cubic structure that has magnetic properties. This form is called
alpha («) iron. As the temperature is increased to 1414°F the crystal structure
' The property that certain chemical elements have of existing in two or more different forms.
 Z)
Section 2-3: The lron-Carbon Diagram

remains unchanged, but the iron loses its magnetic properties.? A further
increase of temperature to 1670°F changes iron to a face-centered cubic
structure, which remains nonmagnetic. This form is stable to a temperature
of 2552°F, at which point another change takes place. The iron again takes
on a body-centered cubic structure and is called delta (6) iron. Finally, when
the temperature is increased to 2802°F, the delta iron melts and becomes
liquid.
Pure iron is important because each form has a different capacity for
maintaining carbon in solid solutions. These different capacities for retaining
carbon are the basis for the heat treatment of steel. Industrially, delta iron
is of little importance and, therefore, can be dropped from further
consideration.
As the reader will observe from Figure 2-1, carbon is soluble in alpha iron
to a maximum of 0.025 % at 1333°F (point P) and only to 0.008 % at room
temperature. The result is an interstitial solid solution with dissolved carbon.
Alpha iron is commonly called ferrite.* It is the softest of all materials in
the diagram.
At a temperature of 2066°F, carbon is soluble in gamma (jy) iron to a
maximum of about 2.0% (point £, Figure 2-1). The name commonly given
to this interstitial solid solution is austenite.
Under equilibrium conditions, the carbon is in the form of iron carbide
(Fe,C) called cementite. This material, containing 6.67 % carbon, determines
the right hand boundary of the equilibrium diagram. It is brittle, weak in
tension, strong in compression, and is the hardest of any material in the
equilibrium diagram.
Point C, in Figure 2-1 is a eutectic point* containing 4.3°% carbon and
consists of a mixture of austenite and cementite known as ledeburite. This
material is not observable because austenite is unstable at room temperature
(except under special conditions) and continues to change on cooling.
Of special significance is point S, at which iron contains 0.8 % carbon at
1333°F. This point is called a eutectoid® and is the lowest point on the diagram
at which austenite will disappear when slowly cooled. The material formed
at this eutectoid is called pearlite. Pearlite is a mechanical mixture of ferrite
and cementite. When viewed under a microscope, the mixture appears in
? Some metallurgists call this form of iron beta (f) iron, but we shall continue to refer to the
region (i.e., OPG in Figure 2-1) as alpha iron.

3 In reality, the alpha solution should be called alpha ferrite to distinguish it from delta ferrite
(i.e., delta iron at 2552°F). However, since delta ferrite has no role in engineering metallurgy,
alpha ferrite is simply called ferrite.
4 A eutectic point is a point on the equilibrium diagram at which two constituent metals solidify
simultaneously at the same temperature. Also, it is the point at which the alloy has the lowest
freezing (or melting) temperature of any other alloy combination (point S in Figure 2-1).
> A eutectoid as defined in Vol. 1 of the Metals Handbook, is ‘‘an isothermal reversible reaction
in which a solid solution is converted into two or more intimately mixed solids on cooling, the
number of solids formed being the same as the number of components in the system.”
 Hig
Chapter 2: Metallurgical Properties of Engineering Materials

lamellar layers of cementite within a field of white ferrite. The name pearlite
has its origin in the fact that the microstructure resembles that of mother-
of-pearl.
The iron-carbon diagram is divided into two major sections (see Figure
2-1). Alloys having less than 2 % carbon (some ofthe literature specifies 1.7%)
are called steels; alloys containing over 2°% carbon are called cast iron. In
addition, the classification for steel is again divided into two sections. Steels
having less than 0.8 % carbon (some literature specifying 0.83%) are called
hypoeutectoid steels; those above 0.8% carbon are called hypereutectoid
steels. Figure 2-2 shows the effect of different carbon contents on the micro-
structure of slowly cooled steels.

SECTION 2-4
Wrought Iron
Wrought iron is a mixture of pure iron and 1-3% slag. It also contains
traces of carbon, manganese, silicon, phosphorus, and sulfur. Wrought
iron is made by pouring molten slag from the open-hearth furnace into
vessels containing iron. The final mix is then squeezed in a press to remove
excess slag and reduced into billets by a rolling mill. The billets can be
reheated to form bars, tubing, plate, structural shapes, pipe, forgings,
nails, rivets, barbed wire, fittings, and so on. Wrought iron is ductile and
soft and is most readily forged and forge welded. In addition to its ductility,
it tends to resist corrosion by forming an oxide film quickly when exposed
to a corrosive environment. Figure 2-3 is a photomicrograph of a transverse
and longitudinal section of wrought iron.
Due to rolling, wrought iron has stronger mechanical properties® in the
longitudinal direction (that is, the direction of rolling) than in the transverse
direction. The strength of wrought iron can be further increased by alloying.
Typical of such alloying is the addition of 1.5-3.5°% nickel. The ultimate
strength of wrought iron can also be increased by cold working’ and sub-
sequent aging. Properties of unalloyed and 34% nickel wrought iron are
listed as follows:

Unalloyed Wrought Iron Nickel Wrought Iron

(longitudinal properties) (3 % Ni)

Tensile strength, psi 42,000-5 2,000 $5,000-60 ,000

Yield point, psi 26,000-35 ,000 45,000-50,000
Elongation (8 in.), % 25-40 25-30
Reducation in area, % 40-55 35-45
Sc
ee

° Various mechanical properties are defined in Chapter 3.

7 See Section 2.2.
Figure 2-2 Change of microstructure of steel with change in carbon content.
(a) Practically carbon-free iron, x 100. (b) Steel with 0.25% carbon. Ferrite (white) and
pearlite (black), x 100. (c) Steel with 0.45% carbon. Ferrite (white) and pearlite (black), x 100.
(d) Steel with 0.85 % carbon. All pearlite, x 100. (e) Steel with 1.10% carbon. Pearlite (dark)
and cementite (white), x 100. (f) Pearlite highly magnified showing lamellar structure.
Ferrite (white) and cementite (black), x 2500. [Courtesy of U.S. Steel Corporation. ]

23
 24
Chapter 2: Metallurgical Properties of Engineering Materials

(a) (b)

Figure 2-3 The microstructure of wrought iron. Slag in a ferrite matrix. (a) Traverse
section. (b) Longitudinal section. Etched in 2% Nital, x 100. [Courtesy of Buehler Ltd.}

In reversed bending and reversed axial loading, the endurance properties

are 19,000—23,000 psi and 11,000—16,000 psi respectively. Impact tests
indicate a Charpy value of 5 ft-lb for transverse “‘granular”’ sections and
18 ft-lb for longitudinal “‘granular”’ sections. At subzero temperatures,
nickel wrought iron displays a strong propensity for maintaining its impact
strength. At 700°F, wrought iron drops in tensile strength, which is reduced
to one half its room temperature value at 900°F.

SECTION 2-5
Cast Iron
As indicated in Figure 2-1, when the carbon content of the iron-carbon
system is in the range of 2-6.67% carbon® the resultant alloy is cast iron.
Most commercial types of cast iron contain between 2.5 and 4% carbon.
It is obvious that, with such a high carbon content, the resulting material
will be brittle and have a low ductility. Consequently, cast iron cannot be
cold worked. Because cast iron flows readily when liquid, it is easily cast
into intricate shapes that can be machined after cooling and aging. Although
relatively weak in tensile strength, cast iron is very strong in compression.
8 Some sources consider a carbon content of 1.7 % and above to be cast iron.
 OB
Section 2-5: Cast lron

Furthermore, the properties of cast iron can be varied extensively with the
addition of alloying metals and proper heat treatment. Cast iron alloys have
been developed for use as crankshafts in engines, gearing, dies, and so on.
There are four primary types of cast iron. These are (1) white cast iron,
(2) malleable cast iron, (3) gray cast iron, and (4) nodular cast iron.” In
addition, one may consider two other forms, which are chilled cast iron and
alloy cast iron.

White cast iron

White cast iron consists of pearlite (that is, transformed austenite) in a
matrix of white cementite (Figure 2-4a). Such a microstructure has virtually
all of its carbon in the form of cementite. White cast iron is difficult, if not
impossible, to machine because it is extremely hard and brittle. This cast
iron is termed “‘white”’ cast iron due to its white appearance when fractured.
The uses of white cast iron are quite limited, although it is employed in
such applications as ball mills, extrusion dies, and liners for cement mixers.
Machines of this kind need the strong wear resistant properties of white cast
iron.

Malleable cast iron

By far the greatest use of white cast iron is in the manufacture of malleable
cast iron. If white cast iron is maintained at 1600°F for an extended time
period and then cooled slowly,'° the cementite loses carbon (Fe,;C = 3Fe
+ C). The resulting microstructure when this substance is cooled to room
temperature consists of free carbon nodules in a sea of ferrite (Figure 2-4b).
Malleable cast iron has mechanical properties that are superior to gray
cast iron (discussed next) except for wear. The material is readily machined
and is best employed for relatively thin sections (for example, 4—2 in.). Wide
use of the material can be found in making equipment for automotive, oil,
agricultural, and railroad industries. Specifically, malleable cast iron is used
to make gear housings, brake pedals in cars, spring hangers, tractor parts,
and so on.
By adding manganese, malleable cast iron can be made to retain more
combined carbon and thus form pearlite malleable iron when slowly cooled
to normal temperatures. Pearlite malleable iron can also be formed by first
heating a ferritic malleable cast iron above its lower critical temperature
followed by rapid cooling in a quenching medium (for example, air, oil).
The resulting microstructure is shown in Figure 2-4b. Tempering of pearlite
malleable iron after it has been quenched will spheroidize'’ the pearlite.

? Also called ductile cast iron.

10 The procedure of heating to and holding a metal at a designated temperature for a period of
time and then cooling it at a controlled rate is called tempering. When the process is applied to
white cast iron, it is called malleablizing for obvious reasons.
'! Spheroidizing is heating and cooling to produce a globular form of carbide in a steel or iron.
Figure 2-4 (a) White cast iron, x 100.
(b) Malleable cast iron, x 100. (ec) Grey cast
iron, x 150. [Courtesy of U.S. Steel
Corporation. ]
 27
Section 2-5: Cast Iron

It is also possible to add alloying metals to malleable cast iron to influence

its mechanical properties. Those alloys most often used are copper or both
copper and molybdenum at the same time. The addition of copper improves
corrosion resistance, tensile strength, yield strength, and reduces ductility.
The combination of copper and molybdenum further improves corrosion
resistance and strength. The following table shows a comparison of the
mechanical properties of ferrite, pearlitic, and copper-moly malleable iron.

a SS EE

Tensile Strength, Yield Strength, Elongation

psi psi (in 2 in.), % BHN

Ferritic 50,000-60,000 32,000-39 000 10-20 110-156

Pearlitic 65,000-1 20,000 45 ,000-100,000 2-16 163-269
Copper-moly alloy 58,000-65,000 40,000-45 ,000 15-20 135-155

Two popular grades of malleable cast iron are ASTM A47-52 grade
32510 and A47-52 grade 35018. The mechanical properties of these two
materials are given in Table A-1 of Appendix A.

Gray cast iron

Gray cast iron is the most widely used ofall cast irons. In fact, itis common
to speak of gray cast iron just as cast iron. It contains between 2.5 and 4.00 %
carbon and usually somewhat more than 2% silicon. Upon solidification,
the cementite, being unstable, breaks up into austenite and graphite called
graphitic carbon. The silicon acts as a graphitizer. The graphite shows up
as irregularly shaped graphite flakes, which give the gray cast iron its gray
appearance when fractured.
The structure ofthe resulting alloy at room temperature is greatly affected
by the amount ofgraphite in the matrix. If, on the one hand, the proeutectoid
cementite!’ and eutectic cementite separate, the resulting matrix will consist
of pearlite and graphite. On the other hand, a large quantity of silicon will
cause the cementite to separate from the pearlite, reducing the strength of
the resulting material. The strength ofgray cast iron is determined by varying
the composition of its matrix from pearlite to different mixtures of pearlite
and ferrite to virtually pure ferrite. The strongest and hardest form 1s
pearlitic gray iron. The weakest and softest is the graphite-ferrite mixture
of gray iron having the lowest carbon content. Strength and hardness
increase with increasing carbon content. Figure 2-4c shows a photomicro-
graph of gray cast iron.
The ASTM specification A48-46 classifies gray cast iron into seven
classes: Numbers 20, 25, 30, 35, 40, 50, and 60. Each class number indicates
the minimum tensile strength. For example, class 20 gray iron has a minimum
'2 The cementite was formed prior to the formation of the eutectoid and therefore is termed
proeutectoid cementite. It is the cementite formed in the region between lines A,,, and A; ,.
See Figure 2-1.
 28
Chapter 2: Metallurgical Properties of Engineering Materials

tensile strength of 20,000 psi, class 25 gray iron has a minimum tensile
strength of 25,000 psi, and so on. A complete listing of the mechanical
properties for gray cast iron can be found in Table A-1 of Appendix A.
Gray cast iron is often used for foundation supports for machinery and
structures due to its high compressive strength and good damping character-
istics. Also, because of its capacity for damping, high torsional shear strength
and low notch sensitivity, gray cast iron has been used in the manufacture
of engine crankshafts. The ease with which it may be cast and excellent wear
properties allow it to be used extensively (in some cases exclusively) in the
manufacture of engine blocks, brake drums, sliding surfaces of machines,
gearing, gear housings, and so on.
In those instances where the use of gray iron may be appropriate yet not
considered because of significantly higher tensile stresses, the higher classes
of gray iron may be used. For example, in fabricating valves, fittings. or
piping and their connections, one might specify a class 40, 50, or 60 gray
iron. However, as the cost of class 60 iron is about 2.5—3 times that of class
30 iron, the designer must consider his selection with attention to this cost
factor.
The designer should be careful when making stress calculations before
selecting a value of E (the modulus of elasticity). Because the stress-strain
curve for cast irons is not linear, the modulus of elasticity is not constant.
Thus, an arbitrary value is established by drawing a line from the origin of
the stress-strain curve to a point corresponding to one quarter of the tensile
strength. The slope of this line is taken as the value for E.
Cast irons (all kinds) are not easily welded. By proper preparation,
however (that is, preheating, surface preparation, choice of welding method
and welding rod) successful welds are possible. Care must be exercised to
make certain that the heating and subsequent cooling of nonuniform sections
do not cause the casting to crack.

Nodular cast iron

Nodular cast iron (also called ductile iron) consists of spheroidal shaped
graphite within a steel matrix. The spheroidal shaped graphite is due to the
addition of small quantities of magnesium (for example, in the form of
nickel-magnesium alloy) or cesium to a desulfurized iron melt just before
casting. In contrast with malleable iron, nodular cast iron is formed on
solidification and requires no tempering.
The matrix can be predominantly ferritic or mixtures of ferrite and
pearlite iron. Nodular cast iron containing no more than 10% pearlite is
called a ferritic iron. It is a tough, ductile iron with good machinability.
Those matrix structures that are predominantly pearlitic are called pearlitic
nodular iron and can be obtained by casting or by heat treatment.!? It is also

'S The heat treatment used here is called normalizing. This procedure consists of heating to
above the A, , line and then cooling in air. (See Figure 2-1.)
 29
Section 2-5: Cast lron

possible to produce a matrix that is martensitic'* by heating and then

quenching (that is, rapid cooling) in oil or water. Martensitic structures are
stronger and harder than either the pearlitic iron or ferritic iron. How-
ever, the pearlitic iron, although not as ductile as ferritic iron, is much
stronger.
With the addition of certain alloying elements, an austenitic matrix may
be produced. Such structures display good resistance to corrosion and creep
at high temperatures. :
Nodular cast iron is stronger, more ductile, tougher, and less porous than
gray cast iron. It is used in the fabrication of crankshafts, pistons, cylinder
heads, forming rolls, pulleys, forming dies, and so on. Figure 2-5 shows the
microstructure of a ferritic nodular iron and a pearlitic nodular iron. The
mechanical properties of nodular cast iron can be found in Table A-1 of
Appendix A.

Chilled cast iron

Chilled cast iron is made by placing metal “‘chill bars’’ inside the mold
but near its surface. As the molten metal is poured, that part of the melt
coming in contact with the chill bars is cooled more rapidly. This produces
a surface of white cast iron which is predominantly cementite and therefore
very hard. Below this surface the material is gray cast iron. Chilled cast iron
can only be machined by grinding and is used in making stamping dies, mill
and crushing rolls, railway wheels, and so on.

Alloy cast iron

In addition to the aforementioned common types of cast irons, there are
various kinds of alloy cast irons. These cast irons contain different alloying
elements.!° By controlling the rate of graphitization, these elements develop
special capabilities, such as better mechanical properties, improved resistance
to heat, corrosion, wear, or brittle fractures. Also, alloying can improve
both the castability and machinability properties of castings. Common
alloyingelementsare nickel, copper, chromium, molybdenum, and vanadium.
Many alloy cast irons are proprietary materials of different companies.
For example, Ni-Resist, an alloy of nickel, chromium, and manganese, is
an International Nickel Company product. Meehanite cast iron, produced
under patent protection, is made with the addition ofa calcium-silicon alloy.
Various grades of Meehanite are produced. See Table A-1 in Appendix A
for the properties of some ailoy cast irons.

‘4 Randomly arranged needlelike crystals that are hard and strong. Such a matrix is a super-
saturated (with carbon) solution of cementite in a body-centered tetragonal iron.

'S The reader should not mistake impurities such as manganese, phosphorus, silicon, and sulfur
for alloying elements. The processes for mass producing steel, cast iron, nonferrous materials,
and so on, are not designed for complete deletion of all impurities.
Figure 2-5 Microstructure of (a)
ferritic nodular iron, x 125 and (b)
pearlitic nodular iron, x 500. [Courtesy
of The International Nickel Company,
Inc.]

30
 oy
Section 2-6: Heat Treatment of Steels

SECTION 2-6
Heat Treatment of Steels
Heat treatment is defined as the heating and cooling of a metal alloy to
alter its mechanical properties. These alterations are varied and have
different purposes. For example, an intended heat treatment may be for the
purposes of strengthening and hardening a metal, relieving its internal
stresses, hardening its surface only, annealing a cold-worked piece, or
improving its machinability by spheroidizing.
The most important part of the iron-carbon diagram in the heat treatment
of steel is that portion involving the transformation from austenite to the
region below the lower critical temperature (see Figure 2-1). Transformation
of austenite, upon slow cooling, begins at the upper critical temperature
line A, or A,,, (actual transformation probably begins at a somewhat lower
temperature as earlier indicated in Section 2-3). At this line the gamma iron
is transformed to alpha with a small quantity of interstitial carbon (that is,
ferrite). As the temperature continues to fall through the transformation
region, more carbon is precipitated out of solution forming cementite (Fe,C)
because alpha iron cannot hold as much carbon as gamma iron. Further
cooling past the lower critical line produces a hypoeutectoid steel (pearlite
and ferrite), a hypereutectoid steel (pearlite and cementite), or a eutectoid
steel (pearlite). Which of the three steels is eventually formed depends solely
on the carbon content.
However, if the cooling rate is such that the transformation temperature
(Ar, or Ar,3) 1s much lower than that for equilibrium transformation, the
time available for the carbon to diffuse out of solution is greatly reduced.
As a result, the gamma iron cannot readily transform to alpha iron because
the carbon is still in solution. Consequently, a supersaturated solid solution
of carbon in iron is formed. This structure is called martensite (also known
as fresh martensite) and is shown in Figure 2-6. A martensite structure has
a white needlelike random arrangement that is very hard and unstable.
It is important to note that the change from austenite to martensite involves
an increase in volume. '°
It is obvious that the rate at which the transformation from austenite
to martensite takes place is significant in preventing the formation of softer
constituents. Thus, for a steel of a fixed chemical composition and known
austenite grain size (that is, temperature of the austenite), there is a cooling
rate that will transform all of the austenite to martensite. This is called the
critical cooling rate. If for some reason cooling during transformation is
stopped, then transformation cannot continue. The temperature at which
martensite first begins to form is designated by M,. The temperature at
which martensite is completely formed is designated by M,. However, for

16 This is so because the atoms ofaustenite are more densely packed than the atoms of martensite.
 a2.
Chapter 2: Metallurgical Properties of Engineering Materials

Figure 2-6 The structure called martensite, x 500. [Courtesy of Buechler, Ltd.]

any one alloy, the cooling rate has no effect on temperature M, or in prevent-
ing the formation of martensite.
Martensite is the agent that contributes to the hardness to steel. It is
apparent that sufficient carbon must be present to produce enough marten-
site. Thus, the extent to which a steel may be hardened depends solely on
its carbon content. Steels with less than 0.35% carbon are not considered
commercially hardenable.'’ Steels with a carbon content of 0.35-0.70%
(these are called medium carbon steels) display the greatest percentage
increase in hardness. As a result, the medium carbon steels are usually
considered for machine elements because they can be hardened while
retaining their ductility.
Higher carbon steels are also readily hardenable but, due to distortion,
warping, or cracking as a result of rapid cooling, are considered for use in
special cases only.
The process of cooling is called quenching and the rapidity with which
heat is removed depends upon the quenching medium. A good quenching

‘7 It should not be overlooked that alpha martensite will nevertheless be formed with a very
high cooling rate.
 38)
Section 2-6: Heat Treatment of Steels

medium should initially permit very rapid cooling and then reduce the
cooling rate at lower temperatures to minimize distortion. In order of
decreasing severity of distortion and cracking, the following quenching
media are most often used: (1) a 10% salt solution in water, (2) plain tap
water, (3) liquid salts, (4) soluble oil and water, (5) oil, and (6) air.
After quenching, the martensite is so hard and brittle that it is of little
practical use. To relieve the internal stresses brought on by quenching,
reduce the hardness, and improve both the toughness and ductility, steel
is subjected to a process called tempering. The tempering process involves
reheating the unstable martensite to any temperature below Ac,, holding
it at this temperature for a fixed time period (for example, } hr) and then
cooling the steel in air to room temperature. Reheating to 150—400°F, for
example, will relieve internal stresses but will reduce the hardness only by
a small amount.'® Further increases in tempering temperature produce
more strain free steel with lower hardness and greater ductility.
Figure 2-7 shows schematically the various micrographic structural
changes possible for tempering at different range levels of temperature.
Figure 2-7 also shows that, if a hardened steel is tempered to a point just
below Ac, and held there for an extended time, the cementite forms globular
or spheroidal shapes. Such a structure is known as spheroidite. It is produced
not only as a sequence in the hardening-tempering cycle described above
but is also an established method of annealing (still to be discussed). For
illustrations ofthe effect of tempering a steel, refer to Table A-4 of Appendix
A. From Table A-4, we can readily see that, as the tempering temperature
increases, the tensile strength, yield strength, and hardness decrease while
the reduction of area, elongation, and Izod impact values increase.

Austenite

Heating above critical

p Spheroidite
Slow cooling Fast cooling
Heating 1100F-250°F
Fine pearlite

Coarse pearlite a Martensite Heating 750F-1100°F

Very fine pearlite

Heating 400F-750°F

B Martensite

Heating 150F-400°F

Figure 2-7 Schematic flow diagram for hardening and tempering steel. Alpha («)
martensite is unstable and beta (f) martensite is stable.

18 At 200°F, the unstable martensite (that is, fresh martensite) becomes stable martensite. This
is sometimes referred to as tempered martensite.
 34
Chapter 2: Metallurgical Properties of Engineering Materials

SECTION 2-7
Isothermal Transformation Diagrams
The discussion in Section 2-6 briefly reviewed the phase change that takes
place when austenite is transformed to martensite by very rapid quenching
under equilibrium conditions. However, the time-temperature combination
at which the austenite is transformed can markedly affect the structure
resulting from this temperature. Thus, the main objective of heat treating
(to develop desirable mechanical properties) may not be achieved. In other
words, we may ask, “What will the final structure and its mechanical
properties be if we allow transformation to begin at temperatures lower than
those at equilibrium (that is, at subcritical temperatures) and then quench
the metal?”
This question is best answered by the use of the isothermal transformation
(I-T) diagrams (also referred to as the TTT or S diagrams).'° Figure 2-8 is
typical of an I-T diagram. These diagrams are obtained by cutting small
specimens from a bar of steel with a specific carbon and/or alloying content.
The first S curve from the left is where transformation begins and the
extreme right curve is where the transformation ends. The form of all
diagrams is very much alike, except for a eutectoid carbon steel from which
the middle heavy line in the upper portion of the diagram is missing.
With few exceptions the diagrams shift to the right (that is, increased
time) with increasing alloy content or grain size of austenite at temperatures
higher than about 900°F. This is the point where the curve bends back on
itself and is called the “‘knee”’ or the “nose’’ of the beginning curve. The
dotted portion of the curves indicates some uncertainty as to the exact
curvature. M,, M;,. and Mo, respectively signify the beginning of the
formation of martensite, the 50% formation of martensite, and the 90%
formation of martensite. Note that not all I-T diagrams have calculated
values for the per cent of martensite formed. Many diagrams have values
for M based upon actual experiment.
Starting from the top of Figure 2-8 we see that, as the subcritical tempera-
ture falls, the time at which transformation begins is also reduced along with
the time of transformation. The shortest time of initiating and completing
transformation is at the “‘knee”’ of the curve. With a further reduction of
the subcritical temperature, both the start of transformation and time for
transformation increase. The right hand side of Figure 2-8 shows the
structure to which the austenite has changed and the temperature range at
which these changes occurred. This information does not normally appear
on I-T diagrams but was put here for clarification purposes only. Actually,
the diagrams indicate each phase constituent by an appropriate letter and
'° TTT stands for time-temperature-transformation diagrams. S diagrams is a term also used
because the shape of the curves resembles the letter S. A most excellent and complete set of such
curves may be obtained from the U.S. Steel Corporation.
 40} “dwar LuNtqiINby ‘Alay JO UONEWUOY
JOY “dua, WuNHqiInby “ployoayna e JO VONeWHOJ

5
Shes,
D:;s

OY-SSANQGYVH
004 — —

iy
O0¢L
OOOLa loey
YW + +
006 —{ Fr |

008 —- f + LS
o
—00r = SW

ainjesadwa|
o®
@

(| jo09 I 0S
{Sn Pe = —+- }——+- SS
eal
*K \
—100z 00b 05), a | ai 2901 e215
06) . D v9°0-9
x 1-| weibeig _ a L-UWEL:
fe an os | payeynojeg T Ir 1 z
i a | Lgerpasnieny
ia | Ulu 4Y | Aep
ules
Uleso)
:9ZIS
L:92Z/S

Ne | | if
GO lt ¢ OS zOl oO pOl gOL 90 QN3931

aunyeiaduuay
ae ee vy = ayluasny W = alisuayey|
' | RACE q = auleg
=9 apiqieg d = euesd
JO

JO}

Q-zZ
‘S'A

TOO]
]901s

[aa§
sanbi4
WieISeIP

30
Asaqinogd]

[PUIOYIOS]

[uoneiodi0
UONRUTIOJsSUR
 36
Chapter 2: Metallurgical Properties of Engineering Materials

legend as can be seen in Figure 2-8. Notice that the I-T curves are asymptotic
to the 1333°F line. This implies that it would require an infinite time to
convert austenite to pearlite. Such a concept is in full agreement with the
transformation criteria of established transformations ‘n the iron-carbon
diagram, namely, that equilibrium conditions exist.
Keep in mind that austenite, once completely transformed at a particular
temperature, will not form another phase when cooled to a lower tempera-
ture. If we wish, for example, to form bainite*® or martensite with the 1062
steel in Figure 2-8, it is obvious that the austenite must be rapidly cooled to
about 500°F within 0.75 sec and held at this temperature for about | hr
to make certain of complete transformation. Any time less than | hr would
end the transformation in the A + F + C region. Cooling, then, to room
temperature would change a percentage (depending on time in A + F+ C
region) of austenite to complete martensite, with the remainder being lower
bainite.
Two questions become apparent in this example. (1) ““How can cooling
from, let us say, 1500 to 500°F be accomplished within 0.75 sec?” (2) “Is 1 hr
too long a period oftime to wait for complete transformation?’ The answer
to these questions is primarily a design decision. If the material is very thin
(for example, needles of a needle bearing) and a very rapid quenching
medium such as ice water or iced brine is used, it might be possible to
quench to 500°F in 0.75 sec. Such rapid quenching (for 1062 steel) is necessary
to avoid touching the *“‘knee”’ or entering the I-T curve at 950°F. Should the
cooling curve touch the “‘knee’’ (or enter the A + F + C region), some
pearlite would be formed. This is what we do not want if we wish to form
only lower bainite and martensite.
If a large component is to be hardened to the bainite-martensite range,
it is therefore unlikely that it can be accomplished without the formation of
some pearlite. Whether the resulting mechanical properties are satisfactory
to the designer is a decision only he can make after the consideration of
many factors.*! The answer to the question of maintaining 500°F for | hr
depends upon the equipment available and whether the time will affect the
rate of production. One aspect the designer can investigate to mitigate such
problems is to seek out another carbon steel or, if necessary, an alloy steel
where the “‘knee”’ of the I-T diagram would be further to the right on the
time scale. Thus, the I-T diagrams provide the designer with a vast choice
of steels having both the strength and toughness to satisfy the requirements
of his design.
Because we can readily measure the temperature-time coordinates of a
steel specimen as it is being cooled (regardless of the quenching medium),

?° Bainite (formerly called troosite) is formed between 900 and 700°F and resembles pearlite. In
this region it is called upper or feathery bainite. At a lower range of 700 to 500°F, the transforma-
tion is called lower or acicular bainite and begins to resemble martensite.
*! Strength, cost, application, environment, quantity produced, and so on.
 37
Section 2-7: lsothermal Transformation Diagrams

Austenite (stable)

\ \\LS é Se NS = ee = = Ka 7
11, @ @
Id \ a Coarse pearlite Ro 10
6| \X5 25) SS eS
\ \ Medium pearlite Rg 20
SS
| \
S

Temperature
\ \ Austenite
\ Munetatta) a Lower bainite R- 60

%y hy CCR = critical cooling rate

Re = Rockwell C hardness (see sect. 3.22)

%1
\ Martensite

Time, log scale

Figure 2-9 A hypothetical I-T diagram of a eutectoid steel with superimposed

cooling curves. The circled numbers distinguish the different rates of cooling and
the cross-hatching indicates the transformation. [S. H. Avner: Introduction to Physics
Metallurgy. McGraw-Hill Book Co., Inc., New York, 1964.]

a more meaningful and descriptive I-T diagram is possible. All that is

necessary 1s to superimpose the temperature-time cooling values on the I-T
diagram. Figure 2-9 shows such a superimposition on a hypothetical I-T
diagram. The cooling curves are numbered | to 7 and each represents an
increased rate of cooling.
Curve | has the slowest cooling rate, beginning its transformation at x,
and completing it at x,. Although the temperature change during trans-
formation was small, the time of transformation was relatively long. The
transformed material structure is coarse pearlite (a soft and ductile material)
and it will retain this structure as it is cooled to room temperature from x’;
regardless of the cooling rate. Curve | is typical of conventional annealing.
Curve 2 is also an annealing process. However, the material is first cooled
more rapidly to a selected temperature, x,, and held within the transforma-
tion area with little temperature change. The result is a uniformly distributed
and somewhat finer pearlite structure.
Curve 3 is typical of what is termed normalizing”? and is performed at a
somewhat faster cooling rate than annealing (curve 1). The resulting structure

22 This term is explained in Section 2-9.

 38
Chapter 2: Metallurgical Properties of Engineering Materials

is a medium pearlite. Curve 4, having a faster cooling rate than curve 3,

produces a fine pearlite structure.
Curve 5 is one of great interest since it results in a mixed microstructure
and is tangent to a 25% A + F + C line within the transformation region.
At x, the fine pearlite begins to form and completes its transformation at
x,. At this point, only 25% of the unstable austenite has been transformed
to fine pearlite. This leaves 75% yet to be transformed. Inspection of the
cross hatching below point x‘ indicates that the cooling curve is cutting
lines (not shown in Figure 2-9) with less than 25 % pearlite. Thus, no further
transformation will take place below point x, until the cooling reaches x‘
at which point the 75 % austenite begins to convert to martensite completing
its change at line M,. The resulting structure consists of 25% pearlite
within a 75% martensite matrix.
Curve 6 represents an extremely rapid quenching which, upon completion,
shows the whole microstructure to be martensitic. Curve 7, which is tangent
to the “knee”’ of the transformation curve, is the limiting cooling rate for
producing a completely martensitic structure. This limiting cooling rate is
the critical cooling rate mentioned earlier in this section. A well-explained
and detailed discussion of I-T diagrams, cooling rates, and so on, can be
obtained from reference [2]. It is also of importance to mention that cooling
curves 4, 5, 6, and 7 are usually followed by tempering to relieve internal
stresses.

Martempering and austempering

With the advent of I-T diagrams, it has become possible to ““map out”
new heat treating paths. Two such paths are known as martempering and
austempering. These processes are shown schematically in Figures 2-10 and
2-11.
In martempering, the steel is quenched (usually in a molten salt bath) to
a temperature slightly above M,. It is held in the bath until all ofthe steel’s
cross section 1s at the bath temperature. The steel is then taken from the
bath, allowed to cool in air to form martensite, and is subsequently tempered
as shown in Figure 2-10. The purpose of martempering is to prevent dis-
tortion and any tendency towards quench cracking. Large steel components
that must be quenched to a high hardness are very prone to these undesirable
characteristics. If the quenching rate is not fast enough, the cooling curve
may not miss the “knee”? completely, resulting in some bainite in the final
structure. Nevertheless, distortion and the tendency towards cracking will
have been greatly minimized.
Austempering is a hardening process where the resulting microstructure
is both hard and tough (bainite) and is accomplished without the need of
tempering. Referring to Figure 2-11, we see that the steel is again cooled
from above the Ae; line by means ofa rapid quench in a molten salt bath to
a temperature just above M,. While still in the bath, the steel is isothermally
transformed into bainite, after which it is allowed to air cool. The resulting
 of)
Section 2-8: Hardenability

Martempering

Temperature

diagram
Tempered to
desired hardness

Room
temperature
Time, log scale

Figure 2-10 Schematic chart illustrating relationship of martempering to a

typical I-T diagram. [Courtesy U.S. Steel Corporation. ]

bainite is, of course, hard but tougher than martensite. By selecting the
isothermal temperature, it is possible to achieve a variety of hardness-
toughness combinations without tempering. The process is usually limited
to small diameter parts because the temperature of the bath retards the
cooling rate.

SECTION 2-8
Hardenability
Hardenability can be defined as the characteristic of asteel that determines
its depth and uniformity of hardness resulting from quenching. As we
learned in the previous section, isothermal transformation can be retarded
(that is, there is a tendency to move the I-T curves to the right) by the addition
 40
Chapter 2: Metallurgical Properties of Engineering Materials

Austempering

KX ii
Ae3

jer Six

4a}UuaD

KS Isothermal
Temperature he
M transformation
diagram

A ss COLL OLS b
“6.0.
O OCC CCCCCE CELE.

AWW Product:
bainite

Room
temperature
Time, log scale

Figure 2-11 Schematic chart illustrating relationship of austempering to a

typical I-T diagram. [Courtesy U.S. Steel Corporation.]

of an alloying agent(s) or by increasing the grain size of the austenite. Thus,

even relatively slow cooling rates can produce a completely martensitic
structure at room temperature. Large grain size is generally not recommended
because the hardened steel looses toughness upon subsequent quenching.
Also, large grains are prone to cracking or distortion when quenched.
To this point, hardenability has been discussed (including prior sections)
in terms of the various transformations required to achieve different levels
of hardness. But what of the depth of hardness in a steel specimen? All
things being equal, can we expect the same hardness throughout the cross
section of two steel bars of different diameters when both are similarly heat
treated? The answer, as expected, is no! In fact, because each grade of
carbon or alloy steel can range within a specified chemical composition,
many steel bars with the same diameters and subjected to like heat treatments
will have different hardness values both for their surfaces and interiors.
 4]
Section 2-8: Hardenability

Other factors affecting hardenability are the size-shape relationship of the

component and its surface condition. Quenching operations are, in reality,
heat transfer phenomena; therefore, the rate at which the heat is extracted
depends upon mass, surface area, metal conductivity, surface heat transfer
coefficient, and temperature difference between the metal. and quenching
medium. Obviously, heat will be more rapidly extracted from those com-
ponents having high surface to mass ratios and no scale to act as an insulator.

Jominy test [3]

For a designer, the task of specifying a hardness is problematical, con-
sidering the complex shapes and variety of steels he encounters. He has,
at least, to be reasonably sure that his requirements will be met. Therefore,
various tests for measuring hardenability have been devised. One test,
among many, has emerged with the widest acceptance. It is called the
Jominy test [3] (also referred to as the end-quench hardenability test) and
has been standardized by the ASTM (for example, A255—48T), the SAE
and the AISI. In the test, all parameters affecting the depth of hardening
are constant from specimen to specimen. Because the only variation in the
specimens is the chemical composition, hardenability can be indicated
solely by the depth of hardening.
The test specimen is a l-in. diameter bar, 4 in. in length, having a larger
diameter shoulder at one end in order that it may be hung vertically in a
test fixture.?? The bar is heated to a proper austenizing temperature in an
inert atmosphere (to avoid scaling), removed from the furnace, and hung
vertically in the test fixture. A stream of water at 75 + 5°F is then impinged
on the lower end of the hanging test specimen. The nozzle diameter is 4 in.
and the pressure head is such that the water column is 24 in. with the nozzle
in a vertical position. When the apparatus is assembled, the end ofthe speci-
men is 3 in. above the nozzle opening. As a result, upon hitting the end of the
specimen, the effluent water forms a spherical surface. Figure 2-12 shows a
test bar in its fixture as it is being quenched.
After a 10-min quenching, the bar is removed and two parallel longitudinal
flats are ground on opposite sides. The grinding operation is performed to
a depth of 0.015 in. per side; Rockwell hardness measurements (see Section
3-22) are made on these flats in ;¢-in. spacing from the quenched end. The
hardness readings versus the distance from the end ofthe bar are then plotted.
Representative curves for several steels are shown in Figure 2-13.
A steel with good hardenability is one in which the hardness does not fall
too rapidly with the distance from the quenched end. Such steel is called a
‘““deep-hardening”’ steel. Figure 2-13 shows 4340 to be a deep-hardening
steel, whereas 4620 obviously cannot be so classified. Thus, 4620 is called a
“‘shallow-hardening”’ steel. Because the test specimen has experienced a
gradation of cooling rates along the bar from rapid quenching at the water
?3 Notice that all specimens have fixed surface to mass ratios since every specimen has the same
dimensions.
 42
Chapter 2: Metallurgical Properties of Engineering Materials

Figure 2-12 End-quench hardenability test

bar in quenching fixture. [Courtesy U.S. Steel
Corporation. ]

end to air cooling at the other end and because the cooling rate is also
independent of the type (that is, the composition) of the steel, it is possible
to convert the distance and hardness from the quenched end to the same
hardness at the center of an equivalent diameter steel bar subjected to
different quenching media.** As a result of the Jominy test data,*> we can
confidently specify the hardness of a large component or make a choice in
selecting an economical steel.
Because Jominy testing has proved reliable, it has become arene to
select a steel by hardenability with parameters like chemical composition,
grain size, and so on, being of lesser importance. The AISI and the SAE
jointly have established hardenability bands for most alloy steels. These
bands are minimum and maximum Jominy hardenability curves and have
been determined from hundreds of heats of each grade of steel. Steels that
are specified and purchased on this basis are designated by the suffix letter H.
Figure 2-13 shows 4140 H to be a steel that may be purchased on the basis
of a hardenability band. The H steels are guaranteed to meet minimum and
maximum hardenability. These steels should not be confused with the
minimum-maximum hardenability ranges possible with steels that are
defined by chemical composition limits. Figure 2-14 shows this distinction
for 4140 steel.
To use the curves in Figure 2-14, it is necessary to know the hardness at
some interior point of a particular bar of steel. Such curves may also be
used to find the carbon content (by using the relationship between carbon
and hardness) of a satisfactorily quenched and tempered steel. There are
24 The method of conversion is discussed in reference [3].
*° See U.S.S. Carilloy Steels, The United States Steel Corporation, Pittsburg, Pa.
 Equivalent hardness at center—water quenched

1 2 3 4 5
| | it | iL
Equivalent hardness at center—oil quenched
1 2 3 4
70 | | Mt

C,
Rockwell
hardness

Vin, L
| 2in.
4

ees 8 12 16 20 24 28 32
Distance from quenched end, 16th in.

Figure 2-13 Hardenability curves. The materials shown were

selected to show differences. Notice that the hardness of AISI 4340
holds up well with depth, that the higher carbon 4063 has the hardest
surface, that the same carbon steels, 4340 and 1340, have practically
the same surface hardness (at ;4 in.), and that the low carbon 4620
(actual 0.17 °% C) has lower hardness all the way. The individual
lines are typical actual test values. The H band shown for 4140 is
suitable for specifications. The top coordinates are defined by this
example: a 2-in., oil-quenched piece of 4063 has a hardness at its
center of R, = 42. [Courtesy Bethlehem Steel Corporation. ]

Rockwell
C,
hardness

—
"eee
Seen Sees
2 a Se ames ace ee

I a 1 it at 1 =k: it 1 1 4
O 2 A st DGS) a, BEE He) ake Slo)
Distance from quenched end, 16th in.

Figure 2-14 Solid lines are for the H band steel. The dotted lines
are 4340 steel based on the limits of chemical composition. [Metals
Handbook, 8th ed., Vol. 1. American Society of Metals, Metals Park,
Ohio, 1961.]

43
 44
Chapter 2: Metallurgical Properties of Engineering Materials

fop)(=)

an=

oSoO

WwoO

CscaleN oO
Rockwell
hardness

= (>)

220 .30 40 .50 .60 70 .80 90 1.00

Carbon, %

Figure 2-15 Minimum as-quenched hardness for obtaining

best properties after tempering as a function of carbon content.
(Courtesy U.S. Steel Corporation. ]

available a whole series of curves”°® for materials other than those shown in
Figure 2-14. Also, there are curves of Rockwell hardness versus carbon
content for different percentages of martensite formations. Figure 2-15 is
representative of these curves and is perhaps the most helpful. It can be used
by the inexperienced designer in the selection ofa steel that will be of sufficient
hardness after quenching and tempering to yield optimum properties.
To see how one might employ these curves, we refer to Example 2-1.

Example 2-1. Because all the other round bar stock in a metal wire straightening
and cutting machine is already 1340, the material of the main drive shaft is
to be changed to 1340 steel. A saving could be realized if the shaft were made
of the same material. The shaft is subjected to a bending load that varies
continuously from 0 to 50,000 psi.
Specify. the maximum diameter of a through hardened shaft and its hardness.
Solution: In order to find a solution, we must anticipate some ofthe equations
and data of Sections 3.22 and 3.24. We therefore have
S, 2~ 500 x Brinell hardness number (BHN) (3-12)
S;, = 0.5S,, (3-13)
therefore

Sj, = (250)(BHN)

where S;, = the endurance limit in psi, and S, = the ultimate strength in psi.
Since it is anticipated that the shaft size will exceed 4 in. in diameter, a size
correction factor of 0.85 will be used (see Section 3-26). Thus we now have
S', = (0.85)(250)(BHN) = 215 BHN

7° U.S.S. Carilloy Steels, United States Steel Corporation, Pittsburg, Pa.

 45
Section 2-9: Heat Treatment Definitions

If we assume a factor of safety of 50% (see Section 1-3), the endurance limit
becomes
1.58, = 215 BHN
We can now solve for the BHN value

1.5 x (50,000)
BHN = = 349
215
From Figure 3-17, we get a Rockwell C hardness number of about 38 for
a BHN = 349. Turning now to Figure 2-15, we note that a steel with a carbon
content of no less than 0.18% carbon would produce the desired hardness
after it has been quenched and tempered. From Figure 2-13, it can be seen
that, if we use the 1340 steel, it exceeds our minimum requirement of 0.18 %
carbon. Also, an oil-quenched shaft 14 in. in diameter or a water-quenched
shaft 13 in. in diameter can be through-hardened to a Rockwell C of 38. @7’

It is important to state at this point that the problems of shaft diameter and
stress have in no way been treated in Example 2-1. All that was demon-
strated was the feasibility of achieving a through-hardness of Rockwell
C = 38 for two possible diameter shafts made of 1340 steel. Interestingly,
according to Figure 2-13, the alloy steel 4620 would not have satisfied the
requirements of the problem.

SECTION 2-9
Heat Treatment Definitions?®
For the purpose of continuity, the following heat treating terms have been
assembled in this one section. Some of the terms have already been men-
tioned, but are presented once again for completeness of definition. Other
terms are defined and explained here for the first time.
Aging. The change in a metal by which its structure recovers from an
unstable condition produced by quenching (quench aging) or by cold
working (strain aging). The change in structure is due to the precipitation
of one of the constituents from a saturated solid solution and results in a
material that is stronger and harder but usually less ductile. This type of
aging takes place slowly at room temperature; it may be accelerated by a
slight increase in temperature. The process is then called precipitation
hardening. (Also see stress relieving.)
Age hardening. The same as aging.
Annealing. A process involving heating and cooling applied usually to
induce softening. The term is also used to cover treatments intended to

27 The symbol @ designates end of Solution.

28 Some of the definitions and descriptions are based on material in the Machinery’s Handbook,
15th ed., The Industrial Press, New York. All of these terms have been adopted by the American
Foundrymen’s Association (AFA), the American Society for Metals (ASM), the American
Society for Testing Materials (ASTM), and the Society of Automotive Engineers (SAE).
 46
Chapter 2: Metallurgical Properties of Engineering Materials

remove stresses, alter mechanical or physical properties, produce a definite

microstructure, and remove gases. Certain specific heat treatments of iron-
base alloys covered by the term annealing are black annealing, blue annealing,
box annealing, bright annealing, full annealing, graphitizing, malleablizing,
and process annealing.
Annealing, black. A process of box annealing iron-base alloy sheets
after hot rolling, shearing, and pickling.*? The process does not impart
a black color to the product if properly done. The name originated in the
appearance ofthe hot-rolled material before pickling and annealing.
Annealing, blue. A process of softening iron-base alloys in the form of
hot-rolled sheet, in which the sheet is heated in the open furnace to a
temperature within the transformation range and cooled in air. The
formation of a bluish oxide on the surface is incidental.
Annealing, box. A process of annealing that, to prevent oxidation,
is carried out in a suitable closed metal container with or without packing
material. The charge is usually heated slowly to a temperature below,
but sometimes above or within, the transformation temperature range
and cooled slowly. It is also called close annealing or pot annealing.
Annealing, bright. A process of annealing that is usually carried out in
a controlled furnace atmosphere so that surface oxidation is reduced to
a minimum and the surface remains relatively bright.
Annealing, flame. A process in which the surface of an iron-base alloy
is softened by localized heat applied by a high temperature flame.
Annealing, full. A softening process in which an iron-base alloy is
heated to a temperature slightly above Ac, and, after being held for the
proper time at this temperature, slowly cooled below the transformation
range. The parts are ordinarily allowed to cool slowly in the furnace,
although they may be removed from the furnace and cooled in some
medium that assures a slow rate of cooling.
Annealing, inverse. A heat treatment, analogous to precipitation
hardening, usually applied to cast iron to increase its hardness and
strength.
Annealing, process. A heat treatment that is applied for the purpose of
softening the component for further cold working in those cases where
previous cold work has caused severe stresses and brittleness. The iron-
base alloy is heated to a temperature close to, but below, the lower limit
of the transformation range and is subsequently cooled. This process
produces less scaling than full annealing.
Austempering. This is a hardening process consisting of quenching an
iron-base alloy from above the transformation range in an isothermal salt
bath ranging in temperature from 300 to 1100°F. The bath temperature
depends upon the material and is maintained slightly above the M, tempera-
ture forming bainite. The process is generally reserved for small components.
°° Pickling is the removal of oxides from metal surfaces by dipping in chemically reactive
solutions or by electrochemical methods.
 47
Section 2-9: Heat Treatment Definitions

Bluing. A treatment of the surface of iron-base alloys, usually in the form

of sheet or strip, on which, by the action ofair or steam at a suitable tempera-
ture, a thin blue oxide film is formed on the initially scale-free surface as a
means of improving appearance and resistance to corrosion. This term is
also used to denote a heat treatment of springs after fabrication in order to
reduce the internal stress created by coiling and forming.
Burnt steel. A term applied to a permanently damaged metal by heating
close to its melting point or by intergranular oxidation. This damage is
sometimes caused by incorrect grinding, which results in the discoloration
of the workpiece due to the heat.
Carbonitriding. A surface hardening process (see case hardening).
Case hardening. A process where the surface layer (or case) of an iron-base
alloy is made substantially harder than the metal’s interior core. Case
hardening can be separated into two classifications. The first, which includes
carburizing, nitriding, carbonitriding, and cyaniding, is based upon the
diffusion of carbon or ammonia gas to some depth of the steel surface.
The second class consists of flame hardening or induction hardening. Here
the change of surface temperature as compared to the core, without the
addition of other materials and the subsequent cooling, hardens the case
and not the core of the steel. Figure 2-16 compares the range of case depths
possible by the methods of diffusion.
Carburizing. A process where carbon is added to a solid iron-base
alloy by heating it to about the 1600—1700°F range while it is in contact
with carbonaceous material, which may be a solid, liquid, or gas. The
depth of case depends on the time of exposure to the carbon atmosphere.
After a sufficient time period, the metal is quenched and then tempered
(for example, 300—-450°F) to produce the desired case hardness and core

Gas carburizing (0.3 to 0.4-in. possible)

of
Type
case

| | | | | valle sal |
0.020 0.040 0.060 0.080 0.100

Depth of case, in.

Figure 2-16 General ranges in case depth possible by different

methods of case hardening. [R. W. Bolz: Production Processes, Their
Influence on Design. The Penton Publishing Co., Cleveland, Ohio,
1956.]
 45
Chapter 2: Metallurgical Properties of Engineering Materials

toughness. Not all steels can be carburized. Steels with no higher than
0.20°%% carbon (there are one or two with about 0.30-0.32% C) are so
treated. Such steels are known as carburizing grades. (See Table A-3 in
Appendix A for a partial list of carburizing grade steels. The table also
indicates the case and core properties.) The case structure is hyper-
eutectoid, whereas the core changes from a eutectoid steel to a hypo-
eutectoid steel as the depth increases.
Solid pack (or box) carburizing places the steel part in contact with
carbonaceous materials such as charcoal, burned bone, burned peach
pits, coke, coal, and so on.
Gas carburizing takes place when the part is heated in a gaseous atmos-
nhere like methane (natural gas), propane, ethane, or butane. The process,
which takes place in a horizontal rotary oven, is usually used for quantity
production. The method is well adapted to small parts such as ball and
roller bearings, chain links, small axles, bolts, and so on.
Liquid carburizing requires that the part be placed in a molten salt bath
(for example, sodium cyanide) where the steel absorbs carbon in its
surface; the depth of absorption again depends upon the time of im-
mersion. The advantages of liquid baths include rapid action, uniform
carburization, minimum distortion, and elimination of the packing
and unpacking required when carbonaceous mixtures are used. In
selective carburizing, the portions of the part not to be carburized are
copper plated.
Carburization is generally used on those parts, such as gears, cams,
gages, and so on, that are subject to wear. The biggest disadvantage
resulting from carburizing is that the part usually tends to warp. However,
the endurance limit is usually positively affected (see Section 3-30,
Carburizing).
Nitriding. A process in which the machined and heat treated part is
placed in a nitrogen atmosphere (ammonia gas) in an air tight box and
heated to about 950—1000°F. After a period of exposure to the ammonia,
the steel is slowly cooled. The case is hardened by iron nitride, which is
formed by the released nitrogen combining with the iron. As a result,
quenching is not necessary and therefore nitrided steels are not prone to
warping.
Nitriding is mostly applied to certain alloy steels that contain elements
such as aluminum, chromium, molybdenum, or vanadium. These alloy
steels nitride better than carbon steels. Nitralloy steel is the name given
to a special group of such steels (for example, Nitralloy N, a product of
the Nitralloy Corporation). However, some carbon steels can also be
treated (for example, 4340). The core properties of the part can be imparted
by previous heat treatment and are not affected by drawing temperatures
up to 950°F. The Nitralloy steels are readily machinable in both the heat
treated and annealed state. Prior to nitriding, the part must be annealed
to relieve stress due to rolling, forging, or machining. Those components
 49
Section 2-9: Heat Treatment Definitions

not requiring heat treatment should be machined to the dimensions

required.
Nitriding produces an extremely hard surface (for example, Nitralloy
N nitrided at 970°F has a hardness of 780 on the Brinell scale) and is
excellent in resisting wear. The process also has beneficial effects on the
endurance limit of the component (see Section 3-30, Nitriding) and
improves corrosion resistance and the tendency to gall (that is, seize).
Nitriding does have some disadvantages: the higher cost of material,
the procedure is more critical than others, the annealing of parts necessary
before nitriding, and the limitation of machining to grinding only (a
procedure not recommended unless very carefully performed).
Typical applications for nitriding are cams, camshafts, cylinder liners,
gears, king pins, piston pins, pistons, pump sleeves, pump shafts, valve
seats, splines, wear plates, wire forming rollers, and so on.
Carbonitriding. A process of case hardening an iron-base alloy in which
both carbon and nitrogen are simultaneously absorbed by heating in a
gaseous atmosphere of suitable composition (for example, hydrocarbons
and nitrogen or molten salt mixtures containing carbonaceous materials
and nitrides). This is usually followed by either quenching or slow cooling
as required. For example, an atmosphere containing about 15% ammonia
in the carburizing gas can be hardened without quenching. With small
amounts of ammonia (for example, | %), quenching is required.
Cyaniding. A process of case hardening a machined steel part by the
simultaneous absorption of carbon and nitrogen by heating in a cyanide
bath. The bath temperature is between 1450 and 1650°F and the bath
consists of sodium cyanide (NaCN) in liquid salt mixture. The percentage
of cyanide varies over a wide range, depending upon the steel used and
the properties required. However, the most frequently used composition
consists of 30% sodium cyanide, 40% sodium carbonate (Na,CO,) and
30% sodium chloride (NaCl). This mixture is most stable and is used for
continuous operations. After immersion in the cyaniding bath for the
proper time period, the part is quenched in a suitable medium. The process
is intended to produce a thin but very hard outer surface on low carbon
steels. The hardness is known as superficial hardness and, because the
surface becomes quite brittle, it does not have high shock resisting qualities.
Although a thin layer is produced, the surface is quite wear resistant.
Flame hardening. A process of surface hardening a steel by heating to
above the transformation temperature range by means of a high tempera-
ture flame and then quenching in caustic, brine, water, oil, or air. The
quenching medium is usually sprayed on the surface but, in some cases,
the part is immersed in the medium. This hardening method is especially
used for large parts (for example, steel forgings and castings, which must
be finish machined prior to heat treatment) or for parts that are irregular
in size or shape and cannot be readily placed in a furnace or bath. The
source of heat is an oxyacetylene torch, which rapidly heats the steel
 50
Chapter 2: Metallurgical Properties of Engineering Materials

surface. The torch may have multiple flame tips and also be equipped
with holes for spraying the quenching medium.
This method is not case hardening in its truest sense, since no diffusion
takes place. Medium carbon and many low alloy stecls as well can be
hardened in this manner. Steels with a carbon content of between 0.40
and 0.45% have excellent core properties after such treatment. Higher
carbon steels (for example greater than 0.60 %) can also be flame hardened,
but great care must be taken upon quenching to avoid cracking. For large
quantity production such as the hardening of gears, splines, shafts, and
So on, special machines are used.
Induction hardening. This process of hardening involves placing the
part in a high frequency electrical field that can vary between 500 and
15,000,000 cycles per second (cps) and that causes the surface of the
component to be heated above the transformation range to a fixed depth.
The part is subsequently quenched in oil or water, but caustic brine or
air are also used. Induction hardening is particularly suitable to parts
that require localized hardening or controlled depth of hardening and is
also suitable to irregularly shaped parts such as cams, gear teeth, or shafts
that require uniform surface hardening around their contours. The method
is principally applied to 0.35—0.55 % carbon steels. Cast irons and malleable
irons can also be surface hardened by this method.
Advantages of induction hardening are(1)a short heating and quenching
cycle, (2) absence of the tendency to produce oxidation or decarburization,
(3) control of hardening depth and area, (4) close regulation of the degree
of hardness, (5) minimization of warping and distortion, and (6) the
possibility of replacing higher cost alloy steels with carbon steels. Figure
2-17 shows how the depth of hardness of a gear tooth may be controlled
by induction hardening.
Controlled cooling. A term used to describe a process by which a steel part
is cooled from an elevated temperature. Cooling is usually from the final
hot-forming operation and is performed in a predetermined manner to
avoid hardening or cracking externally or internally.
Cores. (1) The interior portion of a steel which, after case hardening, is
substantially softer than the surface layer or case (see case hardening). (2)
The term core is also used to designate the relatively soft central portion of
certain hardened tool steels.
Cyaniding. See case hardening.
Decarburization. The loss of carbon from the surface of a steel because of
heating in an atmosphere (for example, hydrogen, oxygen) that reacts with
carbon. This loss can occur during heat treatment, hot rolling, or forging.
Decarburization can be deleterious in two ways. The first is that a maximum
hardness cannot be achieved in heat treatment. The second, and of greater
importance, is that a steel with a decarburized surface has poor fatigue
resistant properties and is likely to develop cracks on its surface.
 oy|
Section 2-9: Heat Treatment Definitions

Vex
36
325
30 --
29--
ee
13) 2) 2B-¢
26-¢
aS
3 Sry
aoc
160
wr
Sy
Wee

Figure 2-17 Hardness ofa gear tooth, induction hardened. Observe the hardness
close to the surface and near the middle of the base of the tooth (where the lower
hardness indicates a tough core). Hardness readings are Rockwell C. [Courtesy
Tocco Div., Park-Ohio Industries. }

The biggest source of decarburized steel is that processed at the mill

during fabrication. To correct for this deficiency, many steel mills will supply
carburized bars in which the surface carbon 1s restored as part of the final
processing. This remedy can also be instituted by the end user. Sometimes,
the mill ““bark”’ is removed by grinding, but subsequent heat treatment may
again cause decarburization unless the treatment is performed in a controlled
nonoxidizing atmosphere. In addition, the effects of grinding may be
equally damaging to steels subjected to fatigue. Nevertheless, some improve-
ments have been cited. For example, automotive leaf springs that retained
decarburized surface failed when subjected to repeated loading between
50,000 and 75,000 psi, depending upon the surface finish. On the other hand,
springs that had been ground down below the decarburized surface and
then carefully polished withstood a repeated stress of 125,000 psi indefinitely
[6]. This same reference cited 30,000 psi as the endurance limit for decar-
burized springs in torsion and 115,000 psi when the decarburized layer had
been removed.
 D2
Chapter 2: Metallurgical Properties of Engineering Materials

Drawing. Drawing is synonymous with tempering, which is the preferred

terminology. Drawing can be confused with drawing a metal through a die.
The term tempering, therefore, is preferred to avoid confusion.
Flame hardening. A surface hardening process (see case hardening).
Graphitizing. An annealing process applied to some iron-base alloys, such
as cast iron, and some steels with high carbon and silicon contents, in which
the combined carbon is wholly or partially precipitated from solution in
rounded nodules as free or graphitic carbon. This is also called temper
carbon.
Homogenizing. A high temperature heat treating process intended to
eliminate or decrease chemical segregation by diffusion.
Malleabilizing. A process of annealing white cast iron in which the com-
bined carbon is wholly or partially transformed to graphite or free carbon.
In some cases, the carbon is completely removed.
Martempering. A hardening process where the steel is quenched in a
molten salt bath to a temperature just above that at which martensite begins
to form. It is held at this temperature until temperature equalization occurs
throughout the part, and then is followed by air cooling.
Nitriding. A surface hardening process (see case hardening).
Normalizing. A softening process in which a steel is heated to about 100°F
above the upper critical temperature and then is allowed to cool in still air
at room temperature. The procedure is intended to produce steel ina uniform,
unstressed condition of small grain size and refinement that permits proper
response to further heat treatments. Normalizing is particularly important
for steel forgings that are to be subsequently heat treated. Depending upon
the material composition, normalizing may or may not leave the steel
sufficiently soft for machining with the tools available. Sometimes, annealing
for machinability is preceded by normalizing. This combined treatment,
frequently called a double anneal, produces a result superior to a simple
anneal.
Overheated. A metal is overheated if it has been exposed to an unduly high
temperature and develops an undesirable coarse grain but is not damaged.
The grain structure can be corrected by suitable heat treatment, by me-
chanical work, or by a combination of the two. In this respect, it is different
from a burnt metal.
Preheating. (1) A general term used to describe a heating preliminary to
some further thermal (for example, welding) or mechanical treatment. (2) As
applied to tool steel, the term preheating describes a process in which the
steel is heated slowly and uniformly below the hardening temperature and is
then transferred to a furnace where the temperature is substantially above
the preheating temperature.
Precipitation hardening. A process of hardening an alloy in which a
constituent precipitates from a supersaturated solution. Also see aging.
Quenching. The rapid cooling of a metal from an elevated temperature by
injecting or spraying the metal with a suitable quenching medium for the
 OB)
Section 2-10: Supplementary Metallurgical Terms

purpose ofincreasing hardness. (Also see austempering and martempering.)

If only certain portions of a componentare to be hardened, then a differential
quenching is used.
Spheroidizing. Any process of heating or cooling that produces a rounded
or globular form of carbide. Spheroidizing of low carbon steels is performed
to meet certain strength requirements before subsequent heat treatment.
High carbon steels are spheroidized to improve machinability, especially in
continuous cutting operations, such as with lathes and screw machines. Tool
steels may also be spheroidized. This is done by heating them to slightly
above the critical temperature, holding them at this temperature for a period
of time, and then letting them cool in the furnace. Spheroidizing also tends to
improve abrasion resistance and is desirable when the material is to be
severely cold worked (for example, extruding, bending, drawing, or cold
upsetting).
Stress relieving. A heat treatment designed to relieve internal residual
stresses induced by cold working, machining, welding, casting, normalizing,
or quenching. Stress relieving 1s a subcritical anneal, where the steel is heated
to a temperature below Ac, and held at this temperature from | to 3 hr. The
partis thenallowedtocoolslowlyinair. To obviate distortion, itis particularly
effective after machining operations.
Tempering. A process of reheating hardened or normalized steel to a
temperature below the transformation temperature range followed by any
desired rate of cooling.
Transformation range. For ferrous alloys, the transformation range within
which austenite forms on heating or disappears upon cooling.

SECTION 2-10
Supplementary Metallurgical Terms
The following terminology, some of which has been previously mentioned
in the text, is listed separately from Section 2-9, which defines heat treating
terms, because the terms to be listed cannot be strictly classed as heat
treating nomenclature.
Alloy. A material with metallic properties consisting of two or more
elements, one of which is a basic metal.
Alloying elements. An element(s) deliberately added to a metal for the
expressed purpose of altering the mechanical or physical properties of the
metal.
Anistropy. That characteristic of a material which exhibits different
property values in different directions with respect to a set of reference axes.
For example, property values “with the grain” as compared to the value of
the same properties “transverse to the grain.”
Cold shortness. The brittleness that exists in some metals at temperatures
below their recrystallization temperature.
 54
Chapter 2: Metallurgical Properties of Engineering Materials

Damping capacity. The ability ofa metal to dissipate the energy of vibratory
or cyclical stresses by means ofinternal friction. Some materials such as lead
have a high damping capacity ;cast iron also has good damping characteris-
tics. Steel, however, has poor damping characteristics.
Embrittlement. Embrittlement is the loss of ductility of a metal. This loss
may be caused by physical or chemical changes.
Flaking. After hot working steel, stresses caused by local transformation
and reduced hydrogen solubility while cooling produce internal microcracks
and “‘flakelike’’ fractures. This adverse phenomenon is known as flaking and
can be corrected by cycle cooling to at least 600°F before air cooling.
Homogeneity. A material is said to be homogeneous when it has the same
properties throughout. For example, steel, with its various size crystals and
different material in the grain boundaries, is not a homogeneous material.
Hydrogen embrittlement. The low ductility of a metal caused by the
absorption of hydrogen.
Isotropic. A material having the same properties in all directions is said to
be isotropic.
Killed steel. A steel that has been deoxidized with a strong deoxidizing
agent such as silicon or aluminum. Such oxidation reduces the free oxygen
content to a point where the carbon cannot oxidize as the steel solidifies.
Killed steel contains few gas holes; this characteristic contributes towards
better forgings and rolled sections.
Red shortness. A brittleness of the steel and tendency towards cracking at
high temperatures caused by the formation of iron sulfide.
Rimmed steel. A low carbon steel that has not been completely deoxidized.
The ingot solidifies with a sound surface (that is, rim), but the core contains
blowholes that are eliminated in the subsequent rolling of sheet and strip
steel.
Semikilled steel. A steel that has been partially deoxidized but still contains
enough oxygen to reduce shrinkage in the ingot by forming carbon monoxide.
Temper brittleness. A brittleness resulting from holding certain steels
within a particular temperature below the transformation range or by
cooling slowly through this range. The brittleness appears at or below room
temperature.

SECTION 2-11

Alloying Elements in Steel and Their

Functions
An alloying element is deliberately added to a metal to alter its physical
‘or mechanical properties. Plain carbon steel contains small quantities of
manganese, phosphorus, sulfur, and silicon, but the amounts are not sufficient
to alter the basic effect of the carbon. Thus, various alloying elements are
added to carbon steel in smaller or larger quantities to produce one or more
 2p)
Section 2-12: Classification of Steels

of the following effects: (1) increase strength, (2) improve hardenability,*°

(3) improve low or high temperature properties, (4) increase resistance to
corrosion, (5) improve machinability, and (6) improve resistance to wear.
Table 2-1, adopted from the 1948 edition of the Metals Handbook
[reference 4] is a summary of the effects of the major alloying elements in
steel. In addition to the elements listed in the table, five other elements are
added to some alloys to promote particular properties. These are copper,
boron, lead, tellurium, and sulfur.
Copper. An element used as an alloying agent primarily to improve the
corrosion resistance of steel when exposed to the atmosphere. The addition
of copper also improves the fluidity of steel when poured. It is usually added
in quantities from 0.10 to 0.40%. When more than 0.75 % copper is added,
steel can be precipation hardened.
Boron. A nonmetallic element that, when added to steel in quantities not
exceeding 0.003 %, is very effective in increasing the hardenability of low and
medium carbon steels. Boron has no effect upon the steel’s tensile strength.
Lead. Additions oflead to steel do not form alloys. Other than improving
the machinability, lead has no effect upon the steel’s mechanical properties.
The self-lubricating properties of lead, which reduce the cutting tool friction,
are said to improve machinability. The amount of lead depends upon the
particular steel considered and varies from 0.15 to 0.35%.
Tellurium. When this element is added to a leaded steel, it will further
improve the machinability ofthe steel. The amount oftellurium depends upon
the lead content and varies from 0.03 to 0.05 %.
Sulfur. A nonmetal, sulfur is considered to be an undesirable impurity in
a steel because of the formation or iron sulfide, which can result in cracking.
However, in the presence of the proper amount of manganese, it forms
manganese sulfide, which improves the machinability of steel. The sulfur
content can vary between 0.06 and 0.30%.

SECTION 2-12
Classification of Steels
Steels are categorized as (1) plain carbon steel, (2) high-strength, low
alloy steel, (3) low alloy structural steel, (4) cast steel, (5) stainless steel,
(6) tool steel, and (7) special purpose steel.

Plain carbon steel

Plain carbon steel is divided into three groups. (1) Low carbon steel has a
carbon content between 0.05 and 0.30%. (2) Medium carbon steel has a
carbon content between 0.30 and 0.50%. (3) High carbon steel has a carbon
content over 0.50%.

3° This term is not to be confused with hardness (see Section 2.8).

 e1qeL1
L-2 BulAO]|\V7 S}UsWAa|Z
UI }891S Pue WY] $}99}JF

uy aouan poexg YaNosY]

Spique):

7%,
Uy dduUANT
uo uy aduon
uo

POS

56
auaysny
7 JUdU9] oa jediouug suoljoun
)

AGHIOS
(Aqyiqeuapivy -Opiqiry uoloy
Uy PuUIWIey uy eydyy SUIWIOY SULING
uod] uol] AQuapua| wd Sud

Vl 9€ suapsep] -Pisuoo soseasouy

-piry DATJESON SA9PIXoIq A]JUSIDIJJa

yy
poseaiout) Aqsa
Aq os Ayyiqeua ) sozu1ydess $]DII]SAy
UIIB YIMOIS
AQ)
(9
Aq uolNjos ATP]
Jt Suro] pasiadsip saptxo
JO

WnUruIn
PaAajOssip
UL (sapliqu
ajluajsne
SulAOT[LY JUdWUa[a
Ul SUIPLyIU
[903s

uintuloly9, STI poqwiypug SudpseY

YSI[SSA saseoiouy
-piry tayeary
ueyy
| A[PIYA S}sIsat

0}
07) YUM SASPIIIUI
-O1109 Ayiqeuap [UJ] SSof SUIUAJOS UOISOIpure UOLRPIXO

-109
(OS°0 UOIS 9DULISISAL pow Aja}e19 uURY)
A soseaiouy Ayyiqeuapsey

SaSPaIOUT
30URjsISAI
sppy awWos ySuaijs
3e ysly
saimnjerodwa}
SISISAY UOISPIGE
PUK IVIM
Ysty
YqA\)

3eQo) poywuntypuy Gh suopiey -pIsuoo sasroiooq

-piry 1eyTUIS
0} a4 sulejsng] saynquyUoD
Oj Pal ssoupseY
(uoqies

Ajqeie
Aq
prjos Aytpiqeua
se ssaupiey
Aq Aq Suruspsey
Soy
uonn[os PoA|Ossip pros WOLNLOS

asouRsUPA, utyur)
poy € soseosouy
-pivy 10]ea15
uey) Aaa] ‘ajqqy] s}oR1aJUNODSs9Ud]}}LIG
WOJJ
Ayyiqeua 94} [NS ANY
ul

pow

S94
Ajaye19 saseaiou] Aj[Iqeuapsey
-ul

sudpie}y
yensn

SSO]4)

poxseuu
Ajaaisuadxa

SA] -seyd
ueyy

saonpar
A}I919
te SLE SS9]) SOPIAOIg
-950 sasvoiouy
-piry Sastey SUUASIROD-UI
-WWd} eIS
8)

JeYMoUWOS
Ayt[iqeua
saseqjudosad

YIM YIM PoLOMO| SuLUOpseY ainjesiad

Jo

Ajo;
d}Usysne
Aq

(Oise uwia4 (o4nzesod A[BUOI}S susdsaq Suluapiey

uinuapgq
sosoddg

dQ

<oOW)(4) S]De10}UNOD AQUIDUA}PseEMO}

Joyeois
{BU01)Suey)

1odua} ssoua[}jiq
“BuLud}JOs

sasiey
joy pur dadio “yysuars
Arepuosas

pot SsauUpseY
Suluopsey,

SoOURYUA UOISOLIOD SDUBJSISAL

Ut Sso]Ureys
[99}S8
suo] Bulsisos-uoiseig
-led e
ay
sajon
 [219IN Uy PopUnyOl SudY}BUdIIS dAI}eBaN | -p1vY Saseasou] Ul gf)q] Ala, susyj}suens payouanbun
Jo
datqoodsai) suoysno} pue Aytiqeua (soziqryde.s ) [yeuis
Jo
uoqies

sjoajs
prjos
Aq yng ‘ATP sade} udoied

ONL

7B

0}
uonnyos

payeouue
(Quaqyuos

MO]
-Wd}

spud}
(ainjesod

sjaaqs

ule}al
suaysno
J-o1T,1ead

UIIM
UOI! WINIWOIYO-Ysty siopudy

| AT[eIoadsa)

IOYysty
aztuaysne

uogied
shore
snioydsoyg foare) cra sudp1ey] saseo1ouy
-piey TIN ae SUIYIBUIIS UOGIeD-MO]
[99}$
Aytyiqeua SASedIOU] 9dUPISISAI
0} -O1109

o1jtuaysne
aatqoadsait) Aq Ajsuods
uoqieojo UOT]N]OS pITos uols
(qua}u09 wy] saaoid Ajt[iqeuryorut
ur
$]99}S BUI}}ND-d91J
UOdITIS
#7 YUM suapiey sasvolouy
-Piey
| aaljesan surejsng
6)
JOU) $°Q|
UIA yonuw yonse|d ur ssoy Aqyiqeua |(saziqrydess)ssoupsey
Aq
(SEO Ajayesopow pros Onnyjos
pas—)
se asodind-[esoued
-ap
pesuryo (d > IS >UW)
(uoqies Aq I9ZIpIxo
BUIAOTLY JUDWITA
JOJ eol4j99]9
yooys
pue
orjousew
WI]
SoAoId
UOT]RPIXO
QOUL}SISOI
Jo
saseoiouy
Ayiqeuapsey
sjaajs BurArsed Sulzijrydewsuou
syuatuaya
sudYySUIIIS
AO]]P-MO]
§[99}$

WISE slo #9 SSo]) -98 SOPIAOIG A[qeqoig

-ul ysaqyeain qusqsisiog UOgIed
ul youl sajolqiedsoxIy
#1) YIM
paramo]
Suruapsey saseaio |-Uuapsey sapiqieo saprqied ssonpay“7 dIjIsudjrew
“prey
UyTK |
(anjeraduiay ur ysry
waysAs AyigeAton %Z)IL “udt Alqeqoid ssou pue Aj[iqeuopsey
ul
(9.070 O4-1L SAOTE |
“payoajjeun
Ajsuoss
%OS'O
Slap
-Sip
se pour WNTWOIYS-Wint
$]99}S
“padjos24 uoqueo{jaa}s
= auiog “Gg
S]UdAIIg
UORULIOJ
JO
-SNB
9prqieo $J99j}Ja -uapieyun Arepuoses 9}1U9}
Ul UWINTWOIYO-Ysry
sonper -piey (2198 Buruspsey
Aqyiqeua s}udaolg
pozijeoo]
uonaldap
‘2
s]9a}s

JO WINTWOIYS
UL ssayureys
[9e}s BuLINpBUC] Suljeoy

| uajssun 9 €¢ ssa) SAPIAOIg

-988 saseasou]
-piey
| suons sosoddg
TT) yum YM Mo] Pola Suluapiey Aytpiqeua Sutuajos
Aq
ul
‘prey

st0(0 so}OWOIg
ssaupiey
SWIC]

WwWdysks
(ainjeraduray ur Ysiy Ajsuols
ul
[00}

Alepuosaspue
$[99}$

94-A\ SAOTTP |[eBus sjunowr Suruapiey YSUdI)s

Je Po}eAd]a -P1OdWI9}
Sa[dIqIed any

uinipeue, I payunpug suapsleH -19POU! sosvosouy

-paey
| Ara Suons
| unuxey
JUejSISAI-UOISeIQR

Sa}eAd|q BUlUasIeOD
Ua} -b1dd
p)
YUM Ayiqeua
Aro >A) JOIL 10j ain}Jo ayiuajsne Sa}JOWOId)

Aq
070(9 ‘A]Buos
sv (qo Arepuooas

pros
ouly (ureds

Ajaqe
PoAlossip suluapsey saseosouy Aj[Iqeuopiey
usym)
a

(Pealossip

uonnyos
sjsisay BuLIaduia}
puke sesneo
paxieui Arepuosas Suluapiey
:ADUNOSsjolaW ‘yoogpuny
TOA‘| yIg “pa UroAWY Maloog
JO “SIEIOW
STEW “seg “OlUO “1961

57
 58
Chapter 2: Metallurgical Properties of Engineering Materials

Low carbon steel is used in a wide variety ofindustrial products and also in
the construction industry. Typical uses are for pipes and tubes, buildings,
storage tanks, railroad cars, automobile frames, nuts, bolts, automobile
bodies, and galvanized sheet steel. When this steel has a high sulfur content,
it is known as free-cutting steel and is used extensively in automatic screw
machines. Many industrial applications use this steel directly in hot- or
cold-rolled form. Cold rolling improves the steel’s strength, machinability,
and stock dimensions. Where a wear resistant surface is required, this steel
can be case hardened.
Medium carbon steel can be quenched and tempered by conventional heat
treatment methods. Thus it can be used for applications requiring greater
strength and wear. Typical products made from medium carbon steel are
forgings, castings, axles, shafts, crankshafts, connecting rods, and any
machined part that requires a strength above that which can be supplied by
a low carbon steel. .
High carbon steel is used extensively where a product must have high
strength, hardness, and good resistance to wear. It is obvious that this steel
must be heat treated to achieve the desired properties. It is often purchased in
an annealed condition, and the finished product is then heat treated to its
proper hardness. High carbon steel is used in making forgings and a wide
variety of tools, such as drills, taps, reamers, dies, and hand tools. It is also
used for making products requiring cutting edges (for example, cutlery,
chisels, shear blades, planer tools, and so on) for spring wire, and for cable
and wire rope.
The use of high carbon steel must be carefully considered due to its loss of
strength and hardness at high temperatures. These are undesirable charac-
teristics for certain cutting tools and for hot-forming operations. Also, this
steel, when quenched, is prone to distortion, warping, and even cracking.
Lastly, high carbon steel has the disadvantage of being shallow-hardening
unless the part is relatively thin. To this extent, the full benefit of heat
treatment may not be attained from high carbon steel.

High strength, low alloy steel

High strength, low alloy steel is used in the “‘as is’’ condition (other than
annealing or normalizing) or heat treated to develop and improve its
mechanical properties. For use “as is,” the alloying elements are added
mainly to strengthen the ferrite, in which the mechanical properties are not
fully developed. When subjected to various heat treatments, high strength,
low alloy steel improves in such properties as strength, through-hardness,
ductility, toughness, and so on.

Low alloy structural steel

Low alloy structural steel was developed mainly for use in the transporta-
tion and construction fields. This steel is not heat treated and, consequently,
to a large extent the desired properties depend ona proper blending of alloy-
 59
Section 2-12: Classification of Steels

ing elements and carbon content. A typical low alloy structural steel has a
yield strength of about 50,000 psi and an ultimate tensile strength of about
70,000 psi. This steel is easily welded and does not harden in air. To attain
a higher strength steel, the carbon content is increased to about 0.30%.
However, the higher strength is obtained at the expense of ductility, forma-
bility, and weldability.

Cast steel
Cast steel has a chemical composition similar to wrought steel except for
the addition oflarger quantities ofsilicon and manganese as deoxidizing and
degassing agents. These additions are made as the steel is being molded.
Steel castings offer a means of fabricating complex components with
mechanical properties approximating those of wrought steel at a cost lower
than other methods of manufacture. Although cast steels have mechanical
properties superior to the cast irons, they are inferior to equivalent wrought
steel compositions. However, standard heat treating procedures can be
employed to adjust the mechanical properties to satisfy specific requirements.
Since the mechanical properties ofcast steel are similar to those of wrought
steel, cast steel can, with few exceptions, be used for the same applications.
Thus, one will find steel castings used in railroad equipment, rolling mills,
mining machinery, electrical machinery, automotive equipment, marine
equipment, machine tools, and so on.
There are five classes of commercial steel castings. These are (1) low carbon
steel, where the carbon content is below 0.2°%; (2) medium carbon steel,
containing from 0.20 to 0.50 % carbon; (3) high carbon steel, with a carbon
content above 0.50%; (4) low alloy steel, with a total alloy content below
8%; and (5) high alloy steel, with a total alloy content over 8%. Those high
alloy steel castings with compositions similar to wrought stainless steel
constitute a group that is of the heat resistant or corrosion resistant type.
By far the greatest number of steel castings produced are of medium carbon
steel. The low and high carbon steel classifications are considered special
because the compositions of many are formulated for particular products
and applications.
Most steel castings are purchased on the basis of ASTM specifications,
which cover virtually all classes and types of steel castings. Table A-1 of
Appendix A lists the mechanical properties of plain carbon steel castings.
The lower strength grades, such as class 60000 and 70000, are specified by
ASTM A27-58. The high strength cast steels, such as class 80000, 85000, and
so on, are specified by ASTM A148-58.

Stainless steel
Stainless steel consists of three types: (1) austenitic, (2) ferritic, and
(3) martensitic. All of these steels possess corrosion resistant properties to
various degrees, depending mainly on the chromium content. Stainless steel,
particularly the austenitic type, polishes to a high luster and finish. As a
 60
Chapter 2: Metallurgical Properties of Engineering Materials

result, it also has many decorative uses in, for example, architecture. In
addition stainless steels are also used for heat resisting applications.
The type of stainless steel is specified by an AISI numbering system. The
SAE employs the same system except for the addition of two prefix numbers.
The AISI numbering system employs three digits: the first identifies the
particular series of stainless steel and the last two serve as identification of
the steel. Any suffix letters that follow the three digits represent a modification
in that series, The series designations are shown in the following table.

AISI SAE General Group Classifications

2XX 203XX Wrought chromium-nickel-manganese austenitic

steel, non magnetic and not hardenable by
thermal treatment

3XX 303XX Wrought chromium-nickel austenitic steel,

non magnetic and not hardenable by thermal
treatment

4XX S14XX Wrought chromium-iron martensitic stainless

steel, magnetic and hardenable by thermal
treatment

4XX 514XX Wrought chromium-iron ferritic stainless

steel, magnetic and not hardenable by thermal
treatment

Austenitic stainless steel is the chromium-nickel group and is known as the

300 series. The chromium-nickel-manganese group, consisting of types 201
and 202, was developed during World War II as a result of the shortage of
nickel. Manganese was used as a substitute for some of the nickel. The
outcome produced stainless steel comparable to types 301 and 302 except for
a slightly reduced ability to resist chemical corrosion. Due to continued use,
types 201 and 202 were added to the “regular”’ list of AISI designations.
The 300 series generally is superior to the martensitic and ferritic types of
stainless steel in resisting corrosion. Of all the stainless steels, it possesses
the highest resistance to scaling and has superior high temperature strength
characteristics. Type 302 is the basic type in this series and is often called
18-8" stainless steel. It is the “work horse” of the 300 series and is used
extensively in food processing and handling machinery, kitchen equipment,
architectural trim, dairy plants, some textile machinery, and so on.
A most important field of application for the stainless steels is nuclear
engineering. Stainless steels are used as cladding for fuel elements, reactor
vessels, piping systems, valves, fittings, and so on. They are used extensively
in the nuclear field because they sustain minimum damage from nuclear

>! Some other types of 300 stainless steels are also known as 18-8, but these are followed by
suffix letters. For example, type 305 is also known as 18-8FS. Effort is being made to use only
the type numbers in order to avoid confusion.
 61
Section 2-12: Classification of Steels

radiation, have high corrosion resistance, reasonably good fabricating

characteristics, good cold and hot ductility properties (that is, are thermally
shock resistant), low neutron capture cross section,*?* and are readily
available at a reasonable cost. The most common types of stainless steels
used are 304, 316, 346, and 347.
The austenitic stainless steels are not hardenable by quenching but respond
readily to cold working, followed by a quick annealing for severe cold
working. Austenitic stainless steels are difficult to machine due to work
hardening, and they have machinability ratings of about 50 % of the reference
steel B1112. The 300 series is very ductile, but due to its propensity towards
work hardening, it does not possess the most desirable formability properties.
Austenitic stainless steels are readily forgeable and weldable by all fusion
methods. The best welds are made when the joint to be welded is protected
from the atmosphere during welding. Also, chromium-nickel steels have
low coefficients of thermal conductivity and high coefficients of thermal
expansion, which results in distortion while cooling. Thus, proper jig design
and use of cooling bars may be required to avoid distortion and warping.
Annealing after welding is required for grades other than the 300 series.
Ferritic stainless steel (part of the 400 series) is not hardenable by thermal
treatment and does not therefore work harden to an appreciable degree.
Being ductile, this steel can readily be drawn, formed, coined, or bent. Cold
working will increase its yield strength about 30 % with a lesser increase in
ultimate strength. Ferritic stainless steel is readily forged or rolled but its
machinability is poor and requires that the cutting tools always be kept
sharp.
This steel can be arc or resistance welded but must be annealed in order to
avoid embrittlement and to improve toughness. The strongest types of welds
are obtained with austenitic welding rods. At elevated temperatures, ferritic
steel has low notch toughness properties, poor creep properties, and poor
rupture strength.
Martensitic is the third type of stainless steel. Like the ferritic type,
martensitic steel belongs to the chromium-iron group and is part of the 400
series. The basic martensitic stainless steel is type 410 and it is the most
expensive of this type. Martensitic type stainless steel has good impact
properties and is hardenable by quenching in oil from 1800°F, followed by
tempering.
The applications for the type 400 martensitic steels are wide and varied.
For example, 410 1s used for valves, screens, pump shafts, cutlery, bolts, nuts,
and various parts in the chemical and petroleum industries. Type 403 is used
to make steam turbine blades, jet engine compressor blades, and, in general,
parts that are highly stressed. Type 416 is used in the manufacture of car-
buretor parts, instrument parts, valve trim, shafts, fishing reels, and golf
clubs. Type 420, as a result of heat treatment, has a high hardness and is used

32 A term used to indicate the probability of a nuclear reaction taking place.

 62
Chapter 2: Metallurgical Properties of Engineering Materials

in manufacturing cutlery, surgical instruments, and so on. Type 440C,

because of its good wear resistance, is used in the manufacture of ball
bearings, bushings, valve parts, valve seats, and more expensive cutlery.
With high carbon grades, machining can only be performed at slower
cutting speeds and feeds. Types of these steels suitable for cold forming are
limited mainly to 403 and 410. All of the martensitic types are readily hot
forged or rolled at temperatures between 1900 and 2250°F.
Both arc and resistance welding of martensitic stainless steel are restricted
mostly to types 403, 410, and 416. For satisfactory welds (that is, without
resulting embrittlement or cracking), the parts should be preheated to
between 150 and 300°F and air cooled to between 1200 and 1350°F after
welding.
Martensitic stainless steel exhibits excellent creep and rupture strengths up
to 1000°F.
All three types of stainless steel can be soft and hard soldered. Soft
soldering (that is, using filler materials of tin-lead mixtures) presents no
problem since it is performed at a relatively low temperature and does not
produce unwanted carbides. Hard soldering, more properly called silver
soldering and/or silver brazing, is performed at higher temperatures (1145°F
being the lowest). It may produce unwanted carbides in the austenitic types
of stainless steel. It is, therefore, preferable to use low carbon or stabilized
grades of steel when hard soldering is required. It is possible, also, to copper
braze stainless steel, but this method requires high purity copper and a
protective atmosphere while brazing. Temperatures of about 2000°F are
required, and the effect upon any previous heat treatment must be care-
fully considered. This method of joining is restricted to relatively small
parts.
There is a class of stainless steel called precipitation hardening stainless
steel, that is a modification of the 18-8 type austenitic steel. This steel
possesses corrosion resistant properties much like austenitic stainless steel,
but it can be hardened by relatively low temperatures to a high yield strength
without any distortion or scaling. It possesses high strength to weight ratios,
much like aluminum, magnesium, and titanium.
Hardening of this steel is accomplished (1) by the transformation of
austenite to martensite by thermal treatment or cold working, and (2) by
precipitation of carbon and low temperature aging, which tempers the
martensite. The types that respond to direct thermal treatment are the
martensitic grades. These include 17-4PH, 15-5PH, and Ph13-8Mo, which
are products of Armco Steel Corporation; AM362, which is made by
Allegheny Ludlum Steel Corporation; and Custom 455, which is a product
of the Carpenter Steel Corporation. The second type is classed as semi-
austenitic grade and includes 17-7PH, PH15-7Mo, Ph14-8Mo (Armco
products) and Am350 and Am355 (Allegheny Ludlum products). These
types are austenitic when solution treated, and the resulting ductility permits
them to be worked much like the ordinary 18-8 stainless steel. For parts that
 63
Section 2-12: Classification of Steels

do not require good ductility, the austenite can be transformed to martensite

by cold working and low temperature aging which is performed after
fabrication. Such treatment can result in an ultimate strength of about
300,000 psi.
The semiaustenitic grades are best for cold forming, whereas the martensitic
grades are best for forging. All grades machine with the same difficulty as
ordinary stainless steel, so speeds and feeds are relatively low and cutting
tools must be sharp. Because the steels are precipitation hardened, welding
poses no problem as in other stainless steels.
Table A-6 in Appendix A contains properties of some representative
wrought stainless steels. Also see Table A-1 in Appendix A for properties of
some cast stainless steels.

Tool steel
The chemical compositions of tool steel make it possible through heat
treatment to produce characteristics essential for cutting tools, shearing tools,
forming dies, drills, punches, gages, and so on. This steel is not intended for
the manufacture of hand tools but is sometimes used for making machine
components which may profit from its special mechanical properties.
In the broadest sense, tool steel should satisfy the following requirements:
(1) The ability to retain its hardness and strength at the elevated temperatures
caused by machining (that is, “hot hardness’’). (2) The ability to withstand
shock and suddenly applied loads without chipping or breaking (that is,
toughness). (3) The ability to withstand the wear and abrasion of continuous
use in order to minimize resharpening or changing tools.
No one tool material can completely satisfy all the above requirements.
Thus, some are sacrificed or modified, as the situation warrants, in order to
obtain desired objectives. Tool steel has been classified? in accordance with
AISI specifications (also used by the SAE), which include quenching methods,
applications, special properties, and steels used in particular industries.
There are six major groups some of which have subgroups. Table 2-2 lists
these groups. A tool steel is identified by the group letter and a suffix number
(for example, W3, D3, and so on). The AISI has completely tabulated the
chemical compositions of each type and number. To discuss the detailed
properties of each group is beyond the scope of this text. The interested
reader is directed to reference [7] and the Tool Engineers Handbook of the
American Society of Tool Manufacturing Engineers, published by McGraw-
Hill Book Company, Inc.
It should be noted, however, that the water hardening group is the least
expensive and will satisfy most requirements. Its major fault is that it is
subject to red hardness and warping due to quenching. The oil hardening
steels, although more expensive, are not as subject to red hardness or warping
after being quenched.
33 Many users persist in keeping the various manufacturer’s names and designations for tool
steel even though most are identifiable by the AISI classification.
 64
Chapter 2: Metallurgical Properties of Engineering Materials

Table 2-2 AISI Tool Steel Designation

TE
RE CET ES eS SE A AES SEES

Group and Symbol Type

ed

Water hardening (W)

Shock resisting (S)

Cold work (O) Oil hardening

(A) Medium alloy air hardening

(D) High-Carbon high-chromium

Ilot work (H) H1-H19 inclusive, chromium base

H20-H39 inclusive, tungsten base
H40-HS59 inclusive, molybdenum base

High speed (T) Tungsten base

(M) Molybdenum base

Special purpose (L) Low-alloy

(F) Carbon-tungsten

(P) Mold steels (P1 to P19, low-carbon,

P20 to P39 inclusive, other types)

Special steel
Special steel may be needed when design requirements call for special
material properties. Special steel is sometimes necessary for high or low
temperature applications, unusually high strength requirements, and
so on.
High temperature service. Power plant equipment, gas turbines, jet
engines, petroleum refineries, chemical plants, and so on, all have com-
ponents requiring steels that will resist oxidation and have good creep
properties (see Section 3-31) for high temperature service. Also, such steel
will not change its crystal structure or suffer embrittlement under long
exposure to high temperatures. Some austenitic stainless steels (for example,
302, 309, 310, 316, 321, and 327) are used for continuous service for tempera-
tures between 1700 and 2000°F, but their creep strengths drop sharply above
1100°F. Some martensitic and ferritic stainless steels (for example, 405, 410,
418, 430, and 446) can also be used for continuous temperature exposure
between 1300 and 2000°F. Their creep strengths, however, are relatively poor
even at 1000°F, and are almost nil at higher temperatures. Thus, they
cannot be used where strength is an important design factor. Stainless steel
has good resistance to scaling. For example, type 440 can resist scaling in
continuous service up to a temperature of 1400°F. All of the other types can
withstand scaling at temperatures higher than 1400°F.
 65
Section 2-13: AISI-SAE Designations for Wrought Steel

Table 2-3 Typical Tensile Strengths of

Ultrahigh-Strength Steel

Type of Steel Yield

i Strength, Ultimate
Ultimat Strength,
ksi ksi

Medium carbon alloy steels 250 300

Modified hot work tool steels 240 290
Martensitic stainless 235 245
Cold rolled austenitic stainless 180 200
Semiaustenitic stainless 220 235
18% Ni Maraging steel 350 355
High-strength, low-carbon
hardenable steels 245 285
High-alloy Q and T steels 290 350
High-carbon steel wire 580 600

The mechanical properties listed are approximate and represent typical maximum
strength levels used in current applications. In many applications, higher strength
levels are possible, but mill practice or manufacturing technology does not alow
these levels to be considered commercially feasible at this time.

SOURCE: 1970 Metals Reference Issue, Machine Design. Penton Publishing Co.,
Cleveland, Ohio.

In addition to stainless steel, alloy steel has been developed expressly for
high temperature service. A representative listing of such alloys is given in
Table A-13 in Appendix A.
Low temperature environments. Special properties for steels subjected to
low temperature environments have become increasingly important. Food
processing, dewaxing of petroleum, liquefaction of gases, synthetic rubber
manufacture, hydrocarbon polymerization, high altitude aircraft, military
requirements, and so on, are but a few ofthe industrial areas that require low
temperature steels.
The largest classification of steels possessing properties suitable for low
temperature applications is the austenitic stainless steel group. (Also see
Section 2-18.) Since the components of a low temperature system are likely
to be welded, the welded joint should also show both ductility and good
notch sensitivity. Generally, welds made with austenite stainless steel
welding rods meet these requirements.
Ultrahigh strength steel. There is also a group of steels known as ultrahigh
strength steel. It can develop very high yield strength and ultimate strength
values. Table 2-3 provides a list of the nine broad categories that are classified
as ultrahigh strength steel.

SECTION 2-13
AISI-SAE Designations for Wrought Steel
The most widely accepted method of specifying a steel is based on the
SAE four- and five-digit system developed in the earlier part of this century.
With passing time and the creation of new alloys, the SAE and the AISI
 66
Chapter 2: Metallurgical Properties of Engineering Materials

worked out a system known as the AISI-SAE, which retains the basic
concept originated by the SAE. The system is structured as follows: The first
digit indicates the type of steel (that is, the principal alloying element), the
second specifies the approximate percentage of the predominant alloying
element, and the last two (or three) digits indicate the average points of carbon

Table 2-4 AISI-SAE Designation Numbers for Wrought Steel

SAE Type SAE Type

Designation Designation

1OXX Plain carbon (nonsulfurized 48XX Nickel 3.50%, molybdenum 0.25%

and nonphosphorized)
51XX Chromium 0.80%
11XX Free cutting (resulfurized,
eg screw stock)

12XX Free cutting (resulfurized S14XX Corrosion and heat resisting steels
and rephosphorized)

Manganese 1.60-1.90% SISXX Corrosion and heat resisting steels

Nickel 3.50% S2XX Chromium 1.50%

Nickel 5.00% 6O1XX Chromium 0.78%, vanadium 0.13%

Nickel 1.25%, chromium 0.60% 86XX Nickel 0.55%, chromium 0.50%

molybdenum 0.20%
Nickel 1.75%, chromium 1.00% 87XX Nickel 0.55%, chromium 0.50%,
molybdenum 0.25%

Nickel 3.50%, chromium 1.50% Nickel 0.55%, chromium 0.50%

molybdenum 0.35%

34XX Nickel 3.00%, chromium 0.80% Manganese 0.80%, silicon 2.00%

303XX Corrosion and heat resisting Nickel 3.25%, chromium 1.20%

steels molybdenum 0.12%

40XX Molybdenum 0.25% Manganese 0.95%-1.25%,

nickel 0.45%, chromium 0.40%,
molybdenum 0.12%

41XX Chromium 1.00%, molybdenum 0.20% 97XX Nickel 0.55%, chromium 0.17%
molybdenum 0.20%
43XX Nickel 1.83%, chromium 0.80%
molybdenum 0.25% 98XX Nickel 1.00%, chromium 0.80%,
molybdenum 0.25%
460XX Nickel 1.75%, molybdenum 0.25%

NOTES: 1. The percentages of the alloying elements are average

>) . In the AISI system, the prefix letters have the following meaning:A = basic open-hearth alloy steel, B = acid bessemer
carbon steel, C = basic open-hearth carbon steel (no prefix also means basic open- hearth carbon steel), D = acid open-hearth
carbon steel, and E = electric furnace steel (usually an alloy),
. When the letters B or Lappear in the middle of a number, it means that boron or lead have been added. For example,
94B40 and LIL41.
4. Suffix letters added to the end of the numbers indicate certain guarantees, restrictions and conformities as follows (a
partial list only): A = restricted chemical composition, C = guaranteed segregation limits, H=guaranteed hardenability
(see section 2,8), and | = guaranteed conformity to nonmetallic inclusions.
 67
Section 2-14: Wrought Aluminum Alloys

in the steel where one point is equivalent to 0.01%. When using the AISI
criteria, the four- and five-digit numbers are the same as for SAE designations
except for prefix and suffix letters. Table 2-4 lists the various designations
and associated types of steel.
As an example of how the system works, an AISI C1040 (SAE 1040) steel
is a plain carbon steel made by the basic open-hearth process containing 40
points or 0.40 % carbon. As another example, AISI 4340 (SAE 4340) steel is
a “nickel, chrome, moly”’ steel with 1.83% average nickel, 0.80°% average
chromium, 0.25 % average molybdenum, and 40 points or 0.40 % carbon.
Tables A-2 and A-3in Appendix A list the mechanical properties of various
plain carbon and alloy steels in the “as drawn,” “‘as rolled,”’ and heat treated
conditions. Also, Tables A-4 and A-5 in Appendix A show, respectively, the
effects of various tempering temperatures and masses on the properties of
some heat treated plain carbon and alloy steels.

SECTION 2-14
Wrought Aluminum Alloys
Second tosteel, aluminum is the most widely used metal and its applications
are manifold. It is available in all forms, such as wire, bars, extrusions
(including structural shapes), powder, sheets, plates, forgings, and castings.
Aluminum possesses a high resistance to most corrosive atmospheres because
it readily forms a passive oxide surface coating. Different alloying elements
affect the corrosion resistance of aluminum to varying degrees; however, all
alloys are essentially corrosion resistant. Halogen acids and strong alkalies
are two ofthe few types of solutions that will attack aluminum because these
substances can remove the oxide coating and expose fresh aluminum.
Aluminum has good electrical and thermal conductivity as well as high
light reflectivity. It is easily formed, drawn, stamped, spun, machined (with
the exception of “O”’ temper), welded, or brazed, depending upon the temper
of the particular alloy.
As a structural material, it can develop strengths comparable to steel.
Because the modulus of elasticity of aluminum is about one third that of
steel, it is not as stiff and will deflect more than steel when subjected to a
comparable load. This characteristic can also affect dimensional accuracy
during machining if too heavy a cut is taken without proper clamping or
support of the work piece. However, aluminum has a high strength-to-weight
ratio, which can be a very important consideration in design of, for example,
aircraft, missiles, trains, and so on. Due to its elastic modulus, cold forming
operations require that the aluminum be bent or shaped much further than
the final required shape (as compared to using steel) because of metal spring-
back. In fact, to assure proper forming, aluminum should be deformed
beyond its yield point.
 68
Chapter 2: Metallurgical Properties of Engineering Materials

The wear resistance of aluminum is not very good, and it should not be
used where this characteristic is a major design parameter. In addition, the
fatigue strength of aluminum is not very high, and careful thought must be
given to those situations where varying loads are an important consideration
before a decision can be made regarding its use.
The coefficient of thermal expansion of aluminum is about 1.5 times that
of steel, and the coefficient of thermal conductivity of aluminum is about
5 times that of steel. These two factors require that both machining and
welding be performed very carefully. When machining aluminum, the
energy ofcutting will cause the work piece to expand due to its high thermal
expansion. This results in dimensional inaccuracies. To minimize this
problem, cutting tools should be kept sharp, feeds and speeds should be
moderate, and a good coolant should be used.
A four digit system is used to designate and identify wrought aluminum
alloys (see Table 2-5). The first digit identifies the alloy group. The second
digit indicates modifications of the original alloy or an impurity limit. The
last two digits identify the alloy or indicate the aluminum impurity.
Some clarification of this system seems to be appropriate at this juncture.
The last two digits of the 1X XX group are the same as the two digits to the
right of the decimal point in the minimum aluminum percentage to the
nearest 0.01%. The second digit indicates the modification, if any, in the
impurity limits. These can be any number from 0 to 9, where zero indicates
that there is no special control of individual impurities, and the numbers | to
9 inclusive indicate a special control of one or more of the impurities. For
example, the aluminum alloy 1060 indicates a minimum of 99.60%
aluminum, which requires no impurity control. Aluminum alloy 1100,

Table 2-5 Wrought Aluminum Alloy

Designations

Alloy Group Number

Aluminum, 99.00% min. and greater 1XXX

Copper 2XXX

Manganese 3XXX

Silicon 4XXX

Magnesium SXXX

Magnesium and silicon 6XXX

Zinc 7XXX

Other elements 8XXX

Unused series 9XXX

 69
Section 2-14: Wrought Aluminum Alloys

however, indicates a minimum of 99.00% aluminum, which requires some

control of one or more of the impurities.
For the remaining alloy groups (that is, 2XXX to 8XXX inclusive), the
last two digits only identify the different aluminum alloys within the group.
For example, aluminum alloy 2017 is in the copper alloy group and nominally
contains 4.0% copper, 0.5 °% manganese, and 0.5% magnesium. Aluminum
alloy 2024 is also in the copper alloy group but nominally contains 4.5%
copper, 0.6% manganese, and 1.5% magnesium. The second digit of the
2XXX to 8XXX groups serves only to indicate alloy modifications. The
assigned numbers can be 0 to 9 inclusive, where zero signifies the original
alloy and the numbers | to 9 (consecutively assigned) indicate alloy
modifications.
The temper designation system indicates the strength, hardness, and
ductility for wrought and cast aluminum and aluminum alloys with the
exception ofingots. The system is based on letters that represent the primary
temper designations. Numbers following these letters indicate temper
subdivisions. Table 2-6 is a list of the temper designations.
Aluminum alloys harden and achieve a high strength by heat treatment in
a manner different from that of steels. The aluminum alloy is first solution
heat treated,** then quenched, and finally age hardened. In contrast to steel,
which hardens upon quenching, aluminum alloys become soft and ductile.
They harden and increase in strength by age hardening. This process
precipitates some of the hardening elements (mainly copper assisted by
magnesium, manganese, and sometimes silicon and nickel) throughout the
alloy structure. Some alloys age harden at room temperature; others age
harden more quickly with the application of heat. The process of heat
application to aluminum alloys is called artificial aging. Alloys that have
been treated by this process are listed under the “T”’ temper designation.
Another method of hardening aluminum alloys, known as strain hardening,
places the alloy in the ‘‘H”’ temper designation (see Table 2-6). For design
conditions requiring a high degree of corrosion resistance, a wrought
aluminum alloy called alclad is available. Alclad aluminum is manufactured
by cladding a thin layer of pure aluminum to the alloy surface while it is being
rolled. Such a layer may be applied to one or both sides ofthe alloy sheet or
plate.
Table A-7 in Appendix A lists the mechanical properties and manufac-
turing characteristics of some wrought aluminum alloys. This table also
indicates a variety of applications for these alloys.

34 As defined in the Metals Handbook, solution heat treatment is “the heating of an alloy to a
suitable temperature, holding it at that temperature long enough to allow one or more con-
stituents to enter into solid solution and then cooling rapidly enough to hold the constituents
in solution. The alloy is left in a supersaturated, unstable state and may subsequently exhibit
quench aging.”
Table 2-6 Temper Designations* and Subdivisions for Aluminum Alloys

—F _ As fabricated: Applies to products that acquire some temper from shaping or forming
processes (wrought products).
=0 Annealed, recrystallized: Safest temper of the wrought alloys.
-H _ Strain hardened: Applies to products that have their strength increased by cold
working only. The —H is always followed by two or more digits. The first digit
indicates the specific combinations of basic operations or as follows:
Asli Strain hardened only: The second digit following this designation
indicates the extent to which the alloy has been hardened by cold
working. The numeral 8 designates the “‘full-hard” temper, the
numeral 4 designates the “half-hard”’ temper, etc. Thus, —H18 is
“full-hard”, -H17 is “‘seven eights-hard”’, —H16 is ‘‘three quarters-hard”’,
etc. Although 8 is ‘“‘full-hard’’, 9 is used to indicate an “‘extra-hard”’
temper. When a third digit is used, it indicates the degree of control
of temper or slightly different mechanical properties than those for
the two-digit -H temper.
-H2 = Strain hardened and then partially annealed: Products are strain
hardened more than desired and then reduced in strength to the desired
level by partial annealing. The number following the designation in-
dicates the degree of strain hardening remaining after annealing, and
uses the same numerical coding as —H1.
—-H3 Strain hardened and then stabilized: Products are strain hardened and
then stabilized by low temperature heating to lower their strength
and increase ductility. This designation applies only to alloys con-
taining magnesium. The second digit indicating the degree of strain
hardening after stabilization, follows the same coding as —H1.
—W _ Solution heat treated (wrought products): An unstable temper applicable only to those
alloys that spontaneously age at room temperature after solution heat treatment. This
designation is specific only when the period of natural aging is indicated (e.g.,—W t hr).
-T Solution heat treated to produce stable tempers (other than —F, —-O, or—H with or with-
out supplementary strain hardening). The —T is always followed by numerals 2 through
10 inclusive, each of which indicates a sequence of basic operations as shown below.
Deliberate variations of the conditions resulting in different product characteristics
are indicated by adding one or more digits to the basic designations listed.
—T2. Annealed to improve ductility and to increase dimensional stability
(cast products only).
—T3 Solution heat treated and then cold worked for the primary purpose of
improving strength (wrought products).
—T4 Solution heat treated and naturally aged to a substantially stable con-
dition. Applies when the products is not cold worked after heat treat-
ment (wrought or cast products).
ae es) Artificially aged after cooling from being formed at an elevated tempera-
ture (wrought or cast products).
—T6 Solution heat treated and then artificially aged (wrought or cast
products).
—T7 Solution heat treated and then stabilized (wrought or cast products).
= 43} Solution heat treated, cold worked and then artificially aged (wrought
products only).
—T9 — Solution heat treated, artificially aged and then cold worked (wrought
products only).
—T10_ Artificially aged and then cold worked (wrought products only).
SS SR SS SS SS A RT ETD

*Based on material from The Aluminum Data Book, Reynolds Metals Co., 1959.

70
 71
Section 2-16: Copper and Copper Alloys

SECTION 2-15
Aluminum Casting Alloys
Aluminum casting alloys are versatile engineering materials and are fast
becoming more popular with the development of new alloy combinations.
Depending upon the particular alloy, aluminum casting alloys are usually
readily diecast, permanent mold cast, sand cast, plaster mold cast, premium
quality cast, investment cast, and centrifugally cast. They can be treated to
a variety of surface finishing processes (for example, shot peened, etched,
sandblasted, and so on), are machinable and with proper casting design (that
is, wall thickness, section thickness), and can be welded with ease. Brazing is
limited to a few alloys such as the sand casting alloy A712.0, the permanent
mold casting alloy C712.0, and alloy 443.0, which can be both sand cast
and permanent mold cast.
The disadvantages associated with aluminum alloy castings result from
their large shrinkage (for example, from 3.5 to 8.5% by volume), possible
hot shortness, and gas absorption. The effects of shrinkage can be minimized
by careful design of the casting, providing for gradual change in sections,
and observing the standards established for minimum wall thickness
depending upon the type ofcasting (for example, 7¢in. minimum thickness
for sand castings). Proper pouring rate, temperature, and so on, can reduce
or eliminate the effect of hot shortness and/or gas absorption.
It is apparent from the variety of casting processes that aluminum alloys
can be produced in singular units, small lots, or in large quantities.
Table A-8 in Appendix A lists the mechanical properties of some typical
aluminum casting alloys that can be used in different casting processes.

SECTION 2-16
Copper and Copper Alloys
Copper
Due to its high electrical and thermal conductivities, pure copper 1s used
almost exclusively in the electrical and electronics industries (although some
inroads into these industries have been made by aluminum wire and cable),
andin the petroleum, chemical, and power generating industries (for example,
in heat transfer equipment). Copper used for electrical conductors is either
tough-pitch copper, which has been electrolytically refined, or oxygen-free
copper, which has a high conductivity. Tough-pitch copper is subject to
embrittlement, whereas oxygen-free copper is not. Both tough-pitch copper
and oxygen-free copper contain 99.90+ % copper.
Copper as well as most copper alloys cannot be heat treated. The
mechanical properties are altered and strengthened by cold working.
However, beryllium copper is an exception because it 1s hardenable.
 72
Chapter 2: Metallurgical Properties of Engineering Materials

Copper alloys
There are approximately 250 copper alloys. Due to the various numbering
designations (ASTM, SAE, Federal Specifications, Navy Specifications, and
so on) and the continued use of alloy names by industry, identification of the
various copper alloys is rather confusing. Recently, the Copper Development
Association has classified copper alloys by number and type, but these
designations have been slow in “taking hold.”
Copper alloys are divided into brasses and bronzes. Brass is principally
an alloy of copper and zinc. The basic brass alloys are alpha brasses (for
example, red and yellow brass) and alpha-beta brasses. It is also possible for
basic brass to contain, separately, small quantities of lead or tin. These alloys
are called, respectively, leaded brass and tin brass (for example, admiralty,
naval, and tobin brass). When a brass contains a relatively large percentage
of nickel (no more than 20°), it is known as nickel silver. When a brass
contains no more than 2.75% lead and no more than 12.00% nickel, it is
known as leaded nickel silver. There is also a “brass” called cupronickel that
contains no zinc at all; instead it is an alloy of copper and nickel (30%
cupronickel also contains a small quantity of iron).
Bronze is principally an alloy of copper and tin. However, there are some
alloys classified as bronzes that contain little or no tin. Many of these alloys
are classified as bronze because they have a color similar to that of bronze.
The principal bronze alloys are phosphor bronze, silicon bronze, aluminum
bronze, and manganese bronze. Table A-9 in Appendix A is a selected list of
some copper alloys and their pertinent mechanical and manufacturing
properties. Table A-10 in Appendix A lists the properties of some cast copper
base alloys.

SECTION 2-17
Magnesium Alloys
Magnesium is the lightest metal (specific gravity = 1.74) known that is
used for engineering purposes. Alloying is necessary, however, to develop its
maximum usefulness and strength.*°> The alloying elements used are alumi-
num, zinc, zirconium, manganese, thorium, and the rare earths. The ASTM
classifies magnesium alloys according to the alloying elements as foliows:
aluminum is indicated by the letter A, zinc by the letter Z, zirconium by the
letter K, manganese by the letter M, thorium by the letter H, and the rare
earth metals by the letter E. The addition of thorium and rare earths radically
improves the elevated temperature properties of magnesium alloys.
The ASTM type designation is quite simple and is best explained by an
example. ASTM type AZ61A-T4 contains 6% aluminum and 1 % zinc. The

°° For forging extrusions, plate, sheet, and strips, the yield strength in tension is greater than
the yield strength in compression. Both strengths are the same for castings.
 73
Section 2-18: Nickel and Nickel Alloys

letter A signifies that this alloy was the first one of this composition (that is,
of aluminum and zinc) to become standard. The T4 is a temper designation
and is the same system that is used for aluminum. Thus, this alloy has been
solution heat treated as indicated by T4.
Magnesium alloys have a high strength-to-weight ratio and are ideal for
intermediate stage aircraft and missile components (for example, frame and
skin stiffeners, fairings, bulkheads, engine parts, wheels, and so on), trucks
and ordnance vehicles (for example, crankcases, transmission housings, fuel
pumps, roof rails, and so on), materials handling equipment (for example,
dockboards, hand trucks, gravity conveyors, platform trucks, and so on),
hand tools, optical equipment, office equipment, and die cast components.
These alloys are readily machinable (for example, machinability index =
500 based on 100 for B1112 free cutting carbon steel) and are virtually
adaptable to all forms of metal working and joining such as casting, forging,
extruding, inert gas-arc welding, resistance welding, and riveting. Caution is
advised when machining magnesium alloys because finely divided particles
are easily ignited. Large pieces of magnesium are difficult to ignite because
heat transfer is so efficient that the material cannot reach its melting point.
The usual practice is to moisten the areas where fine magnesium dust collects
and frequently sweep it into a container.
Magnesium alloys show good resistance to atmospheric exposure, although
salt water will attack them unless they are protected by surface finishing.
They also have good resistance to attack by chromic and hydrofluoric acids,
alkalies, solvents, and most organic compounds such as hydrocarbons,
aldehydes, alcohols, phenols, amines, esters and oils.
Table A-11 in Appendix A lists the properties of three popular magnesium
alloys.

SECTION 2-18
Nickel and Nickel Alloys
Nickel is used in a variety of applications, particularly those where
corrosion resistance and oxidation resistance is an important requirement.
In addition, some nickel alloys are extremely tough (that is, 200 ft-lb Charpy
V-notch test) so that they can be used at temperatures as high as 2000°F as
ultrahigh strength and superalloy structural materials. Other nickel alloys
are excellent for cryogenic applications and are strong, tough, and ductile
even at temperatures as low as — 400°F.
Wrought nickel alloys have good manufacturing characteristics; they are
readily machinable, cut, sheared, punched, cold worked, hot worked, and
weldable. The casting alloys can be machined, ground, welded, and brazed.
Although more costly than either steel or aluminum, nickel alloys are less
costly than the refractory metals for solving severe temperature-strength
problems. Also, they display magnetic, magnetostrictive, electrical, and
thermal properties that are important for particular applications.
 74
Chapter 2: Metallurgical Properties of Engineering Materials

There are some nickel and nickel alloy specifications based on ASTM,
ASME, AMS, and governmental standards, but, to date, the customary
practice is to refer to these alloys by their trade names with the exception ofa
few high nickels (that is, 94°% or more nickel—for example, Nickel 200,
Nickel 201, Nickel 210, Duranickel 301, and Berylco Nickel 440). Table A-13
in Appendix A lists the mechanical properties of some nickel and nickel
alloys at room temperature. Table A-14 in Appendix A lists the mechanical
properties of some nickel and nickel alloys at low temperatures. One need
only compare the corresponding mechanical properties in these tables to see
why a nickel alloy is considered a tough material, capable of satisfying the
severe requirements of high and low temperature environments.

SECTION 2-19
Plastics
The name plastics, used to describe a very large number of synthetic*®
organic compounds, is misleading in that such materials are only truly plastic
during a particular stage of processing. However, plastics represent a vast
and growing field of synthetic materials. These substances have such diverse
properties that there are few areas of modern civilization where they are not
in use. A design engineer’s interest in plastics concerns their use as structural
materials as well as their use for special applications.
The basic building blocks of plastics are carbon, hydrogen, and oxygen.
These are linked together in saturated*’ or unsaturated?* organic com-
pounds. Methane and ethane are typical saturated compounds. Ethylene and
acetylene are typical unsaturated compounds (see Figure 2-18). When heat,
pressure, and suitable catalysts are used, the unsaturated bonds of amolecule

Saturated Molecules Unsaturated Molecules

H HH H H
| kd ; |
F=CSn Pi—G~C—4n C=C H—C=C—H
| eel ie
nt tal H oH
methane ethane ethylene acetylene

Figure 2-18 Typical saturated and unsaturated molecules.

°° There are natural plastics such as resins (for example, copal, rosin, pitch, ambers, and shellac).
*7 Saturated means that the four valence bonds of carbon in a compound are satisfied (Gigs
methane, CH,).
°® Unsaturated means that the four valence bonds of carbon in a compound are not satisfied.
This situation arises when there are not enough hydrogen or oxygen atoms to link with the
carbon atoms present. In this case, the unsatisfied carbon valence bonds link with each other.
These are represented by double or triple lines (e.g., ethylene and acetylene).
 US)
Section 2-19: Plastics

allow additional similar molecules to join and form a long chain. This process
is called polymerization. The resultinglongchain moleculeis called a polymer,
and the component smaller molecule is called a monomer. Figure 2-19 shows
a long chain polymer made up of ethylene monomers. A large number of
other polymers can be produced by replacing the hydrogen with oxygen or
linking two or more different monomers, each with different properties.
New structural arrangements of the same elements of amolecule, which form
what are known as isomers, are also possible. This is similar to the existence
of allotropic forms of an element (for example, sulfur).

T Figure 2-19 Long chain molecule

of ethylene units.
ae
Qi ==
25
28)Bey
ere
G3
IEG ele
was
Oba
eee
(ee
Z—-o-,
|

When two or more polymers are crosslinked, the process is called copoly-
merization. It results in a class of plastics known as thermosetting plastics.
Thermosetting plastics experience a chemical change upon the application
of heat and pressure, and once formed cannot be resoftened. Polymers that
are not crosslinked result in a class of plastics known as thermoplastic
plastics. These materials soften with increasing temperatures, and harden
with decreasing temperatures. This cycle of heating and cooling can
be repeated as often as one wishes without destroying the material’s
properties.
Thermosetting and thermoplastic resins may be mixed. This results in a
compound having some of the properties of each type. These compounds
can be resoftened, but not to the same degree as the thermoplastic constituent.
The addition of different fillers, such as paper, cotton, linen, asbestos, or
glass, results in two groups of plastic materials known as reinforced plastics
and laminates. These plastics possess special characteristics (that is, high
toughness, wear resistance, low dielectric losses, and so on), depending upon
the particular laminate. Mention should also be made of the recently
developed filament wound composite, which is formed with an epoxy base
(for example, diglycidyl ethers of bisphenol A). It is being used for rocket
motor cases, chemical tanks, pressure containers, high strength tubing,
shotgun barrels, and missile bodies.
Table A-15 in Appendix A lists the properties of plastic materials for
applications that are of primary interest to the designer. The plastics are
identified in this table by their generic names. A particular plastic material
taken from this table is likely to be produced by several manufacturers, each
of whom labels it with his own trade name. Such a list would be extra-
ordinarily lengthy and, therefore, cannot be given here.
 76
Chapter 2: Metallurgical Properties of Engineering Materials

SECTION 2-20
Elastomers
Elastomers is a term used to describe materials that are resilient and possess
rubberlike qualities. They include natural rubber and many synthetically
compounded materials that have special properties. Elastomers are used to

Table 2-7 ASTM Designations and General Properties of Elastomers

Common Name or
ASTM D 1418 Representative
Designation Trade Name Chemical Designation General Properties

NR Natural Natural polyisoprene Excellent physical properties; good resistance to cutting,

gouging, and abrasion; low heat, ozone, and oil re-
sistance; poor resistance to petroleum-base fluids

IR Polyisophrene Synthetic polyisoprene Same properties as natural rubber: requires less masti-
cation than natural rubber

CR Neoprene Chloroprene Excellent ozone, heat, and weathering resistance; good

oil resistance; excellent flame resistance

SBR SBR Styrene-butadiene Good physical properties; excellent abrasion resistance;

not oil, ozone, or weather resistant

NBR Buna N Acrylonitrile-butadiene Excellent resistance to vegetable, animal, and petroleum

oils; poor low-temperature resistance

IIR Butyl Isobutylene-isoprene Excellent weathering resistance; low permeability to

gases; good resistance to ozone and aging, low tensile
strength and resilience

IIR Chloro-butyl Chloro-isobutylene-isoprene | Same general properties as butyl

BR Butadiene Polybutadiene Excellent abrasion resistance and high resilience; used

principally as a blend in other rubbers

it Thiokol Polysulfide Outstanding solvent resistance; other properties poor

EPM EPR Ethylene propylene Good aging, abrasion, and heat resistance; not oil
resistant

EPDM EPT Ethyl propylene terpolymer Good aging, abrasion, and heat resistance; not oil
resistant

CSM Hypalon Chlorosulfonated polyethylene Excellent ozone, weathering, and acid resistance; fair
oil resistance; poor low-temperature resistance

VMQ Silicone Methyl-viny] siloxane Excellent high and low temperature resistance; good
PVMQ Silicone Phenyl-methyl-viny! siloxane mechanical properties at high temperature; low com-
FMQ Fluorosilicone Trifluoropropy! siloxane pression set; fair oil resistance
Fluorosilicone rubber has excellent oil resistance

AU Urethane Polyurethane diisocyanate Exceptional abrasion, cut and tear resistance; high
modulus and hardness; poor moist-heat resistance

FPM Viton Fluorinated hydrocarbon Excellent high-temperature resistance, particularly in

air and oil

ACM Acrylic Polyacrylate Excellent heat, oil and ozone resistance; poor water
resistance
SOURCE: 197] Plastics/Elastomers Reference Issue, Machine Design. The Penton Publishing Co., Cleveland, Ohio
 77
Section 2-20: Elastomers

make products for commercial and industrial application. Tires, soles and
heels, footwear, floor covering, gloves, wearing apparel, electricians’ tape,
household utensils, and so on, are some examples of their commercial
application. Industrial applications, however, are of greater interest to the
design engineer. Examples of some industrial applications are V belts, hoses
for steam, chemicals, water, and oil, tubing for experimental and medical
purposes, gaskets, oil shaft seals, bearing seals, rubber springs, machinery
mounts, and vibration dampers. Table 2-7 lists the ASTM designations and
general properties of various elastomers.
Important properties of elastomers are evaluated by standard ASTM
tests. Tests have been standardized for hardness, tensile strength, elongation,
elastic modulus, tensile and compression set,*° abrasion resistance, resilience,
oil resistance, low temperature characteristics, and deterioration resistance
from exposure to oxygen, ozone, light, and heat. Table 2-8 indicates some
quantitative strength values for some general purpose elastomers. Reference
[8] includes an excellent elastomer selection guide and extensive information
about the properties of elastomers.

Table 2-8 Properties of General Purpose Elastomers

Hardness, Elactomiert Tensile Elongation

Shore Strength, psi %

Soft
20 EPR, NR, § 500-1000 400-700
30 CR 500-2000 400-800
Medium
40 EPR oSBRS 500-3000 400-600
50 NR, § 500-4000 250-500
60 IIR, CR, 500-4000 250-500
70 NBR 500-4000 150-300

Medium Hard

80 SBR, 500-3000 100-200

90 NR, 8 500-2000 75-125
50+ CR, 1000-2000 50-100
607 NBR 1000-3000 25-50
Hard
70+ SBR 3000-6000 2-20
807 {NR ‘5 } 3000-6000 2-10
90+ NBR 3000-7000 2-5
1007 3000-8000 2-5

“Elastomers in each group are listed in order of increasing cost.

t Shore D scale, (Hardness ratings are always A scale unless otherwise indicated.)
Recommendations for natural rubber, in general, include synthetic natural (IR). The current
trend is to replace the tree-grown varieties with the synthetic isoprene types.

SOURCE: 1/97] Plastics/Elastomers Reference Issue, Machine Design. The Penton Publishing
Co,, Cleveland, Ohio.

3° Tension set is the amount of deformation remaining after removal of the tensile load. Com-
pression set is the same as tension, except the load is compressive. The tests are performed at
different fixed temperatures and for different fixed periods of time.
 78
Chapter 2: Metallurgical Properties of Engineering Materials

PROBLEMS
1. What methods are available for improving the strength of metals? Which of these
methods are most commonly employed?
2. What effect does cold working have upon steel?
3. Explain why a smaller grain metal is stronger than the same metal with a larger
grain. How does ductility and toughness vary with grain size?
. What is the foremost reason for adding an alloying material to a steel? Are there
any other reasons?
5. What is the difference between wrought steel and cast iron?
6. Name the four types of cast irons and indicate the advantages and disadvantages
of each as a structural material.
. Define the following : (a) hypoeutectoid steel, (b) hypereutectoid steel, (c) eutectoid
steel.

. What are the constituents of the following: (a) austenite, (b) ferrite, (c) cementite,
(d) pearlite, (e) martensite.
. List in proper order the procedure involved in hardening a plain carbon steel by
heat treatment. Name the metallurgical phase that exists at each step.
. Why isn’t low carbon steel hardened by heat treatment? Explain your answer.
How is low carbon steel generally hardened?
. What is meant by tempering a steel, and why is this process employed?

. Distinguish between annealing and normalizing. What is the purpose of these

processes?
. From the heat transfer aspect, what is the difference between using oil or water
as a quenching medium? How do they affect the material being hardened?
- Explain how an isothermal transformation diagram is used for heat treating.

. In general, what effect does an alloying metal have on isothermal transformation

diagrams?
. Distinguish between hardness and hardenability.

. Name the seven different steel categories.

. Plain carbon steels are divided into three groups. Name these groups and indicate
some typical applications for each.
19. Identify the major alloying agent and per cent carbon content of the following
steels: (a) 1020, (b) 1095, (ce) 1112, (d) 2330, (e) 3140, (f) 4340, (g) 5120, (h) 8740,
(i) 9255.
20. Which of the steels in Problem 19 is known as a “free machining steel’’? Which
is often called “drill rod”’?
21. How does case hardening (also called surface hardening) differ from through-
hardening? What is its purpose? Name the various case hardening methods.
22. What are the advantages of nitriding over carburizing and cyaniding? What are
the disadvantages of nitriding over cyaniding?
23. What are the causes of decarburization and why is it deleterious?
 Ue)
References

24. What advantages can be had with high strength, low alloy steels?
25. Describe martempering and austempering.
26. Differentiate between age hardening and precipitation hardening. What is artificial
aging?
27. Name the three grades of stainless steel. What constituent in stainless steel makes
it particularly resistant to corrosion? List some typical applications for each grade.
28. Which of the three grades of stainless steel is hardenable by heat treatment?
How are the other two grades hardened?
29. What requirements must a tool steel satisfy?
30. What two methods are used to harden aluminum alloys?
31. For most atmospheres aluminum is known to resist corrosion. Explain why.
32. How are wrought aluminum alloys designated?
33. What is the difference between brass and bronze?
34. How are wrought copper alloys hardened? (See Table A-9 of Appendix A.)
35. List some advantages and disadvantages of wrought magnesium alloys as a
structural material.
36. What group of metal alloys are particularly well suited for high and low tem-
perature applications?
37. Name the two general classifications of plastic material. Explain their significance.
38. List some applications for each classification mentioned in Problem 37.

REFERENCES
[1] E. P. De Garmo: Materials and Processes in Manufacturing, 2nd ed. The Macmillan
Co., New York, 1962.
[2] S. H. Avner: Introduction to Physical Metallurgy. McGraw-Hill Book Co., Inc.,
New York, 1964.
[3] W. E. Jominy and A. L. Boegehold: Hardenability tests for carburizing steel. Amer.
Soc. Metals Trans., 26, (June 1938).
[4] Metals Handbook, Vol. 1, 8th ed. American Society for Metals, Metals Park, Ohio,
1961.
[5] R. W. Bolz: Production Processes, Their Influence onee The Penton Publishing
Co., Cleveland, Ohio, 1956.
[6] H. W. Gillet: The Behavior of Engineering Material. John Wiley & Sons, Inc., New
York, 1951.
[7] 1970 Metals Reference Issue, Machine Design. The Penton Publishing Co., Cleveland,
Ohio.
[8] 1971 Plastics/Elastomers Reference Issue, Machine Design. The Penton Publishing
Co., Cleveland, Ohio.
Mechanical Properties
of Engineering Materials
SYMBOLS
BHN = Brinell hardness number S, = fracture strength, psi
Cr = fatigue strength reduction S, = endurance strength for.a
factor due to surface finish fixed number of cycles, psi
Cr = fatigue strength reduction S;, = endurance limit, psi
factor based on reliability S,, = ultimate tensile strength, psi
Cy = fatigue strength reduction S,- = ultimate compressive
factor due to size strength, psi
Cy = fatigue strength reduction S,; = ultimate shear strength, psi
factor due to welding S,» = yield point, psi
D.M.F. = deviation multiplication U,, = modulus of resilience,
factor in.-lb/in.4
E = elastic modulus in tension, U,; = modulus of toughness,
psi in.-lb/in.*
G = elastic modulus in shear, psi VHN = Vickers hardness number
K, = fatigue stress concentration y = angular strain, rad
factor é = linear strain, in./in.
K, = theoretical stress €, = strain at fracture, in./in.
concentration factor = angular deformation,
K,, = theoretical stress radians
concentration factor for Tmax = Maximum shearing stress,
shear psi
q = notch sensitivity factor T) = nominal shearing stress, psi
R = Rockwell hardness number v = Poisson’s ratio
S. = working endurance limit, o, = compressive stress, psi
psi o, = tensile stress, psi

The primary purpose of this chapter is to review those mechanical

properties that the reader may have studied in more specialized courses
dealing with engineering materials. In addition, we wish to highlight and
supplement those material aspects that have a decided affect upon the design
and manufacture of an engineering device or structure.

80
 81
Section 3-4: lsotropy

SECTION 3-1
Mechanical Properties of Materials
Mechanical properties are those that indicate how the material is expected
to behave when subjected to various loads or combinations of loads. These
mechanical properties are determined by standardized test methods that are
outlined by The American Society of Testing Materials (ASTM). Knowledge
of these properties permits the designer to determine the size, shape, and
method of fabricating structural and machine elements. Since no two test
specimens, even from the same “melts” are exactly alike, the results given in
the tables of materials are often minimum values, average values, or minimum-
maximum values. Thus, one must consider the assigned table value of a
particular material carefully before making a decision. This is particularly
true where high strength-to-weight ratios are being sought. Often, it is wise
to verify the values of interest with a particular supplier before proceeding
further with a material selection.
One of the most complete compendiums of ferrous and non-ferrous
materials is the Metals Handbook, Volume | published by The American
Society for Metals (ASM). Volume | deals with the properties and selections
of metals and is periodically up-dated to reflect the latest data. Of particular
importance is the emphasis placed upon the statistical distribution obtained
for metal properties. Thorough understanding of the definitions and
meanings of the various mechanical properties is of paramount importance
to the intelligent selection of a material. Much ofthe discussion that follows
is concerned with the definition of some of these properties.

SECTION 3-2
Homogeneity
A material that exhibits the same properties throughout is said to be
homogeneous. Homogeneity is an ideal state that is not achieved by real
materials, particularly metals. However, the variation in properties is so
small that calculations for stress and deflection assume that a material is
homogeneous throughout.

SECTION 3-3
Elasticity
Elasticity is defined as the ability of abody, when subjected to an external
load, toreturn toits original size and shape when the external load is removed.

SECTION 3-4
Isotropy
A material that displays the same elastic properties in all loading directions
is said to be isotropic. The equations of elasticity and strength of materials
 8&2
Chapter 3: Mechanical Properties of Engineering Materials

are based upon this assumption. However, in more recent times, great
interest is being displayed in those materials that are not isotropic (for
example, filament wound plastic and other fibrous materials).

SECTION 3-5

Plasticity
When a material is subjected to an external load of such magnitude that
deformation continues with no apparent further increase in load, the material
is said to have become plastic. Once in this region, the material will have
experienced a permanent set (that is, deformation) and will not return to its
original size and shape when the load is removed. Thus, plasticity can be
considered as the opposite of elasticity. This definition is for a theoretically
perfect material. Actually, a material such as metal (except for a very low
carbon steel) will continue to deform with only a very small further increase
of load (see Figure 3-1).
80

7075-T6 Aluminum alloy

WO
LA
Va
/ Molybdenum
60 |— Uf
fh Mild steel
/ Le See Pele ha Sa ae St oo aaa Wade ha ee

sof /
/
fe
& { Strong Cl
4 40 —
a ie /
B I ~
| Magnesium alloy
30 =I
I Modulus
Molybdenum 40 million
20 Steel 30 million
Alurminum alloy 10 million
Magnesium alloy 6.5 million

10

| l i I 1
0 0.004 0.008 0.012 0.016 0.020 0.024 0.028 0.032
Strain, in./in.

Figure 3-1 Stress-strain diagrams of some metals.

SECTION 3-6

Tensile Strength
Tensile strength, S,, given in pounds per square inch and also called
ultimate strength, is the highest point plotted for a material on a stress-strain
 8&3
Section 3-8: Strain

curve when the material is subjected to a tensile test (see Figure 3-2). Non-
brittle materials display the same stress-strain curve for both tension and
compression, whereas brittle materials do not have this characteristic.

Figure 3-2 Stress-strain curve for a mild (low-

carbon) steel, showing the stress variation based on
the original cross-sectional area of the specimen and
the true variation of stress based on a reduced
cross-section area. Point A is the proportional
kpsi
Stress,
limit, B is the upper yield point, C is the lower
yield point, D is the tensile strength (ultimate
strength), and E is the fracture strength. Point F
is the true fracture strength.

Elastic region

Strain, % elongation

SECTION 3-7
Stress
Stress, S, is also given in pounds per square inch. There are three kinds of
simple stresses. These are tensile stress, compressive stress, and shearing
stress (this includes torsional stress). The numerical value ofthese stresses is
determined by dividing the load by the original area of the member. Although
it is common practice, a stress calculated in this way is really a nominal or
engineering stress as contrasted to the true stress. The true stress would be
that stress obtained by dividing the load by the reduced cross-sectional area.
This area must be that measured area associated with the measured load. As
a result, the true stress is much higher than the stress based upon the original
area. Figure 3-2 illustrates this point.

SECTION 3-8
Strain
In tensile testing, strain (¢, linear measured in inches per inch, and jy,
angular measured in radians) is defined as the change of the specimen length
divided by the original gage length. This strain is sometimes called the
nominal strain, engineering strain, or conventional strain. There is also a
term called true strain, which 1s defined as the logarithm ofthe ratio between
 84
Chapter 3: Mechanical Properties of Engineering Materials

the specimen length at the time of measurement and the original gage length.
For shearing stress there is an associated shearing strain. This value, a
dimensionless quantity, measures the angular change in radians between two
lines that were originally at right angles on the test specimen.

SECTION 3-9
Fracture Strength
Fracture strength, S,, given in pounds per square inch and also called the
breaking strength, is that value at which a test specimen actually separates.
To obtain this value, the load at separation is divided by the original cross-
sectional areas. Again we must recognize that the value so obtained is not
the true value. To obtain the true strength, the load at separation should be
divided by the cross-sectional area at separation (see Figure 3-2).

SECTION 3-10
Proportional Limit
The proportional limit, given in pounds per square inch, is defined as that
point on a stress-strain curve beyond which the stress is no longer propor-
tional to strain (see Figure 3-3). The actual value of the proportional limit is
quite difficult to establish because it depends greatly upon the sensitivity and
quality of the measuring instruments used. Except where it is absolutely
necessary to know this value, it is generally not used in engineering
calculations.

SECTION 3-11
Elastic Limit
Elastic limit, given in pounds per square inch, is that point on the stress-
strain curve where the material being tested begins to display a slight
deviation from a straight line. Most elastic materials, when strained to their
elastic limits, will return to their original size and shape when the applied
load is removed. Very often the elastic limit and proportional limit are taken
to have the same values.

SECTION 3-12
Proof Stress
Since the proportional limit and the elastic limit are such undeterminable
parameters, a term called proof stress, also given in pounds per square inch,
is employed. Thus proof stress is that stress which will cause a small but
permanent strain in a material. This permanent strain is specified as 0.01 %
 8&5
Section 3-13: Yield Point

ksi
Stress,

0 0.1 0.2 0.3 0.4 0.5

Elongation, % in 2 in.

Figure 3-3 Stress-strain curve defining proportional limit,

proof stress, and yield strength. [Technical Editor Speaks. The
International Nickel Co., Inc., New York, 1943.]

in 2 in. or called the 0.01 % offset. One can readily see that this type of stress
specification can become the basis for assuring a user that the supplier is
delivering a material with a consistent and reliable “‘yield property”’ (see
Figure 3-3).

SECTION 3-13
Yield Point
Also called yield strength, the yield point (S\,,, pounds per square inch) is
that point on the stress-strain curve where the test specimen experiences a
relatively large increase in deformation with no increase in load. This is an
ideal statement based upon what is considered a perfectly elastic and perfectly
plastic material. Actually, the location of the yield point depends very much
on the material being tested.
Figure 3-4 demonstrates this inconsistency. The curve for metal A is
typical of that for a mild steel (that is, low carbon). Note that this curve
 8&6
Chapter 3: Mechanical Properties of Engineering Materials

ksi
Stress,

Elongation, %

Figure 3-4 Stress-strain curve showing clearly defined yield points for
metal A and metal B, as displayed by the dip in those curves. Metal C shows
no such dip. [Technical Editor Speaks. The International Nickel Co., Inc.,
New York, 1943.]

shows a dip, indicating that the stress is actually reduced while the strain
continues. After straining to a certain point, the material recovers and
continues to increase in stress (at a much reduced rate) with increasing
strain. In fact, curve A is said to display an “upper yield point” and a “‘lower
yield point.”’ The curve for metal B is typical for hard steels (for example,
medium carbon) and some nonferrous metals. Here we see that the curve no
longer has a distinct dip but does display a break in what otherwise would
be a smooth curve. High carbon steels, alloy steels, and most nonferrous
metals have a stress-strain curve similar to that shown by the curve for metal
C. In this case, there is no dip, no break, but only asmoothcurveand therefore
a distinct yield point is nowhere indicated.
Because of the possible confusion that might result in specifying the yield
point, it has been agreed to use a 0.2 % offset line to establish the yield point.
Thus, if we draw a line parallel to the elastic part of the stress-strain curve
but offset 0.2 % in elongation, the point on the curve cut by this line is taken
to be the yield point. This is clearly shown on Figure 3-4. For copper base
alloys, the offset is usually taken as 0.5%. Other factors such as stress
relieving’ and cold drawing? can also have an appreciable effect upon the
yield point value. Figure 3-5 for a mild steel bar indicates these variations.

' See Section 3-30 (that part dealing with heat treatment).
See Section 3-30 (that part dealing with mechanical surface treatment).
 8&7
Section 3-14: Modulus of Elasticity
120

Ht

Lean bar

100 Af ii —

Yield strength at 0.2%

offset = 100,000 psi

80 i
=I
T e|
Yield of cold-drawn and stress

Lp fee relieved bar = 85,000 psi

L
= 60,000

60
ksi
Stress,
Drop of beam for as-rolled bar
ait fe

40
| Proportional limit of cold-drawn bar doubtful

[ace eal

20
0.2% offset

| J
0 0.2 0.4 0.6 0.8 1.0 (he 1.4 1.6
Strain, %

Figure 3-5 Effect of cold working and/or stress relieving on a

steel Bar. [Batelle Memorial Institute: Prevention of the Failure
of Metals Under Repeated Stress. John Wiley & Sons, Inc., New
York, March 1949.]

SECTION 3-14
Modulus of Elasticity
The modulus of elasticity (E for tension and G for shear, pounds per
square inch) is the ratio of stress to strain measured within the proportional
limit of the material. Thus, the modulus is really a measure of the stiffness
(for tension or compression) or rigidity for shear of a material. For ductile
material, the modulus of elasticity is the same for both tension and compres-
sion. In particular, it is called ““Young’s modulus” and is represented by
the letter E. For brittle materials such as cast iron, certain magnesium
alloys, and so on, this modulus is not the same for both tension and compres-
sion.
 88
Chapter 3: Mechanical Properties of Engineering Materials

For the case of shear (or torsion), the modulus analogous to that for
tension is called the shear modulus or modulus of rigidity. (It is also known
as the transverse modulus.) The symbol used to signify this modulus is G.
The modulus of rigidity and Young’s modulus are related by the equation
ns E
DA)
(3-1)

where v is Poisson’s ratio and is defined as the absolute value of the ratio
of the transverse strain to the axial strain when an axial load is being applied
to a body.

SECTION 3-15
Hooke’s Law
Hooke’s law is really implicit in the aforementioned definition of the moduli
of elasticity. Directly stated it says that stress is proportional to strain up
to the proportiona] limit of the material. Thus, for tension (or compression)
we have
S = Ee (3-2)
For shear (and torsional stress) we have
S, = Gy (3-3)
SECTION 3-16

Direct Shear Strength

Shear strength values are obtained by simple tests as shown in Figure 3-6
for single shear and Figure 3-7 for double shear. The specimen to be tested
is sheared between the hardened edges of the supporting block and the
block to which the load is applied. The shearing strength is the load at
separation divided by the cross-sectional area being sheared. Thus, if the
separation load is P and the specimen has a thickness t in. and is / in. long,
the direct shearing strength (in single shear) is S,, = P/tl. Similarly for
direct double shearing strength we have S,, = P/2tl.

0.005-in. Load
clearance

Figure 3-6 Fixture for direct single

shear. [Technical Editor Speaks. The
International Nickel Co., Inc., New York,
1943.]
Specimen

Table
 389
Section 3-17: Shear Yield Point

Load
0.005-in. clearance

Figure 3-7 Fixture for direct double

shear. [Technical Editor Speaks. The
International Nickel Co., Inc., New
York, 1943.]
Specimen

Table

Note that the phrase shear strength was prefixed by the word direct,
implying that this is the only stress placed upon the test specimen. However,
inspection of the test fixtures as illustrated by Figures 3-6 and 3-7 indicates
that bending stresses do exist and the stress cannot be considered as pure
shear. Therefore, the shearing stress as calculated above is an average stress.
This type of calculation is justified in analyzing bolts, rivets, welded parts,
of any other mechanical member where the bending moments are considered
negligible. Also, because strain measurements are difficult, if not impossible
to measure, few values of yield strength are determined by testing. It is of
interest to note that tests of bolts and rivets have shown that the strength in
double shear can at times be as much as 20% below that for single shear.
In those cases where the ultimate shear stress for a material is not easily
available, the following rule of thumb relations may be employed.
Wrought steel: S,, = 0.82S,

Malleable iron: S,, = 0.90S,

Cast iron? 9S. = 30s. (3-4)

Copper and copper alloys: S,, = 0.90S,

Aluminum and aluminum alloys: S,, = 0.65S,

SECTION 3-17
Shear Yield Point
As stated above, values for the yield point in shear (S,,,, pounds per square
inch) are generally not available. However, those values that are listed are
usually obtained by torsional testing of round test specimens. The equation
used for plotting the stress-strain curve for shear (that is, torsion) is
_ Gré
S (3-5)
ee
where r is the radius of the test specimen in inches, L the length of the test
specimen in inches, and @ the angular deformation (that is, twist) of the
shaft in radians. The yield point can then be determined from the plotted
 90
Chapter 3: Mechanical Properties of Engineering Materials

curve as discussed earlier. Where information concerning shear yield values

is not readily available, one may use the following approximations :
Aluminum and alluminum alloys: S,,, = 0.55S,,
=e (3-6)
Wrought steel: S syp = 0.585,

SECTION 3-18
Ductility
Ductility is defined as that property that permits a material to be deformed
without fracture. Ductility is usually expressed as a per cent elongation in
a 2-in. gage length. A ductile material has a value greater than 5% anda
brittle material less than 5°. Since the per cent elongation is not a measure
of the actual strain of the specimen, another method of measuring ductility
is also used. This method is based on a per cent reduction of cross-sectional
area between a specimen fractured in tension and the original area of the
test specimen.
Ductility,an often overlooked material property, can play a very important
role in deciding which material one should select for a particular design.
For example, it is possible to have two materials of approximately equal
tensile strengths and hardnesses, yet the one with greater ductility would be
able to withstand a greater overload than the material with lesser ductility.
Indeed, this conclusion would be obvious should one of these materials be

100,000
_- Structural silicon steel

90,000

80,000

70,000

60,000

50,000

psi 40,000
Tensile
stress, }p
High
30,000 purity
aluminum
annealed
20,000

10,000

IL a altl J
0 0.04 0.08 0.12 0.16 0.20 0.24 0.28 0.32 0.36
Strain, in./in.
(Determined on 8-in. gauge length)

Figure 3-8 Ductility curve (in tension) for some metals and
alloys. [Technical Editor Speaks. The International Nickel Co..
Inc., New York, 1943.]
 9]
Section 3-19: Malleability
12,000
Energy to Rupture
(total stretch X average load)
in. Ib.
10,000 Monel 1,446

Steel 1,116
Naval brass 931
Bronze (1.15% tin) 668

8,000 ;—

Ib
load,
Applied

Stretch in 3.95 in., in.

Figure 3-9 Ductility curves (in tension) for 3-in. diameter

bolt made of various materials. [Technical Editor Speaks. The
International Nickel Co., Inc., New York, 1943.]

brittle and the other ductile. Material used for making bolts would be an
application for a highly ductile material because bolt material is often
subjected to prestressing in addition to its normal loads. Still another
circumstance where good ductility is a highly desirable property is when
material must be cold worked, such as in deep drawing, spinning, cold
heading, bending, and so on. Unfortunately, materials that are both hard
and of high tensile strength usually are not as ductile as those with lower
hardness values and tensile strengths.
Figure 3-8 shows ductility curves for some metals and alloys based on
an 8-in. gage length rather than a 2-in. gage length. Figure 3-9 shows the
results of some strength-ductility tests on 3-in. diameter bolts performed at
Columbia University.

SECTION 3-19
Malleability
Malleability can be defined as the same as ductility, except that it is
applied to compression. Thus, one can say that malleable materials permit
high plastic deformation in compression without fracture. A malleable
 92
Chapter 3: Mechanical Properties of Engineering Materials

material is one that can easily be flattened or rolled without preheating.

Gold, aluminum, copper, and lead are representative of materials that are
highly malleable.

SECTION 3-20
Modulus of Resilience
The modulus of resilience (U,,, inch-pounds per cubic inch) is defined as
the ability of a material to absorb energy within its proportional limit.
Thus, the area of the triangle formed by the elastic part of the stress-strain
curve and the line dropped from the proportional limit perpendicular to
the strain axis determines the numerical value of the modulus of resilience.
Any area under the elastic curve less than that which determines the modulus
is called the strain energy of the material. A detailed discussion of strain
energy and its applications will be found in Chapter 6. For the present, we
will merely note that, for a uniform member subjected to a tensile stress
oc, the strain energy stored in the member is

U=to8
O.€ (327)
Since Hooke’s law applies (that is,.o, = Ee), equation (3-7) becomes
o2
U=—OE 3-8
(3-8)

If o,is the proportional strength S,, then equation (3-8) becomes

S2

p — i (3-9)

where U, is the modulus of resilience in inch-pounds per cubic inch of

volume. In a similar manner, we can express the moduli of resilience for
other types of loading, such as torsion, bending, and direct shear.

SECTION 3-21
Modulus of Toughness
Toughness (U,, inch-pounds per cubic inch) is the ability of a material
to absorb energy and plastically deform before it fractures. Thus, as in the
case for resilience, toughness can be calculated by evaluating the area under
the stress-strain curve. The maximum toughness ofa material is then
ef
U; = { o de
0)

To integrate this equation, one would have to know the explicit relation of
g as a function of ¢. Because this is not generally known (at least in the plastic
> The English alphabet will be used for strength. and the Greek alphabet for stress.
 oy
Section 3-21: Modulus of Toughness

region) other “‘schemes’’-are used to express a modulus of toughness. One

of these is simply
Ur = Sues (3-10)

where é, is the strain at fracture. Still another method is to multiply the

strain at fracture by the arithmetic average of S,, and S,,. Thus
Sees
U; = (et |., (3-11)

Toughness is usually associated with the ability of a material to with-

stand an impact or shock load. Two popular tests have been devised to
determine the impact strength of materials. One such test is called the
Izod impact test and the other is called the Charpy impact test.
The modulus of toughness is obtained by recording the difference between
the potential energy of the pendulum weight before free fall and after the
specimen is struck.* The values are indicated on a dial shown in Figure 3-10.

Figure 3-10 Izod impact testing

machine. [Courtesy of Ametek, Inc.]

This difference in potential energy of the pendulum weight is the impact

energy absorbed by the test specimen. A brittle material will generally
break, whereas a tough material will only bend. To assure fracture, the
specimens are notched to introduce a stress concentration.
+ Variations of the Izod and Charpy tests are also used for torsional impact and tensile impact.
 94
Chapter 3: Mechanical Properties of Engineering Materials

Although the data obtained by testing do not agree with calculated

values based on the area under the stress-strain curve, it does permit the
comparison of materials on a relative basis. Such information can help a
designer select the tougher of two or more materials.

SECTION 3-22
Hardness
Selecting a material that has good resistance to wear and erosion is very
much dependent upon the hardness and surface condition. For example, a
material that must withstand both dynamic loading and wear should have
both toughness and hardness. Such a material might be used for gears, cams,
mechanical slides, and so on. The most generally accepted definition of
hardness is the ability of a material to resist plastic indentation. Other
measures of hardness are determined by the scratch test, the file test, and
the Durometer test.
Several types of hardness testers are available for accurately measuring
the hardness of all shapes of wrought materials. These are the Brinell,
Rockwell, Vickers, and Shore Scleroscope hardness testers (see Figures 3-11,
3-12, 3-13, and 3-14, respectively).

Figure 3-11 Air-O-Brinell hardness

testing machine. [Courtesy Tinius Olsen
Testing Machine Co.]
 Figure 3-12 Rockwell hardness
testing machine. [Courtesy Wilson
Instrument Division of ACCO,
Bridgeport, Conn.]

Figure 3-13 Vickers pyramid

hardness testing machine.

95
 96
Chapter 3: Mechanical Properties of Engineering Materials

Figure 3-14 Scleroscope hardness

testing machine. [Courtesy Shore
Instrument and Manufacturing Co., Inc.]

The Brinell test is mainly used for materials whose thickness is }in. or
greater. The testing technique is to impress a 10-mm diameter hardened
steel ball into the surface of the material under 3000-kg load for 15-60 sec.
A smaller load is used when testing softer metals and alloys. The diameter,
in millimeters, of the dent made by the ball is measured by a microscope,
and the area is calculated. The Brinell hardness number is then found by
dividing the load by the calculated area of the indentation. Representative
values might be 250 BHN (3000 kg) or 100 BHN (500 kg).
One of the most popular types of testers is the Rockwell hardness tester.
Since no measurements are made (as compared with the microscope measure-
ments for the Brinell tester) and hardness can conveniently be read from an
indicating dial, the device lends itself to production or shop use. Normally,
materials of no less than ;; in. in thickness are tested. The procedure is to
indent the material with an initial load of 10 kg thus making a small penetra-
tion. The dial indicator is then set back to zero and a final heavier load is
applied (see Figure 3-12). After the dial indicator has come to rest, showing
that penetration has ended, the load is removed and the hardness is read
directly from a scale on the dial indicator. Depending upon the material
being tested, the penetrators can be a ;4-in. or 3-in. diameter hardened steel
ball or a diamond cone called a Brale. The following table shows a list of
dial scales, indenters, indenting loads, and the associated materials to be
tested.
 7
Section 3-22: Hardness

aa aR

Rockwell Scale Indentor Load, kg Material

Ra Brale 60 Very hard metals (e.g., tungsten

carbide)

Rp ale 100 Soft metals (e.g., soft steel,

Moyer copper alloys)
dia. ball
Rc Brale 150 Hard metals (e.g., hardened steel,
heat treated alloy steel)

Rp Brale 100 Very hard metals

RE alt BA 100 Soft metals (e.g., bearing materials,

Sam magnesium, aluminum)
dia. ball
Rp ein. dia. ball), Ry, Rx, RL, Rm; Rp; Rs, (all with different size ball indentors) are used for
soft materials (e.g., plastics, very soft metals, etc.)
RG uses a qe-in. ball indentor and is used for metals such as phosphor bronze

The Vickers hardness test is similar in principle to the Brinell test, as

hardness is determined by the ratio of the impressed load to the indented
area. However, rather than being a hardened steel ball as is the case in
Brinell testing, the indenter is shaped like a square pyramidal point and is
made of diamond. The diagonal lines formed by the diamond on the material
surface are measured by a micrometer and are converted to an area. This
area, divided into the applied load, is a measure of the material hardness.
Of all the methods used to measure hardness, the Vickers is considered to
be the most reliable. Hardness values are designated as follows: 250 VHN,o.,
or 150 VHN, 50.,- The first designation says that a 10 kg load using a diamond
pyramid reported a hardness of 250. The latter designation says that a 120 kg
load using a 2-mm diameter ball reported a hardness of 150.
The last of the more common hardness measuring devices to be described
is the Shore Scleroscope. This device measures the rebounding height of a
freely falling diamond hammer that is dropped from a fixed height. The
rebound is measured on a graduated dial scale that indicates the hardness.
For example, a typical hardness is reported as 18 Shore. An arbitrary scale
of 100 represents the highest hardness value. This number is based on the
hardest tool steel. The material tested should have a relatively smooth
surface for good results. The method is popular because it is easy and rapid
to use. However, the results obtained are the least reliable of all machine
methods.
For certain materials such as acrylics, phenolics, acetates, polyvinyl
chlorides, fluorocarbons, and other “high-impact”? molded or extruded
plastics, the Rockwell hardness tester can be used. The hardness of softer
plastics and wool felts is measured by the Shore Scleroscope.
 98
Chapter 3: Mechanical Properties of Engineering Materials

Figure 3-15 Durometer hardness

tester. [Courtesy The Shore Instrument
and Manufacturing Co., Inc.]

Synthetic rubber, natural rubber, foam plastics, synthetic fiber felts,

and similar materials are types that employ a Durometer hardness tester
(see Figure 3-15). This instrument measures the hardness of a material by
noting its resistance to elastic deformation and so no permanent indentation
remains. The hardness readings are given in Durometer hardness numbers
based on an arbitrary scale of 100.
Commonly used hardness numbers are not simply convertible each into
the other. Figure 3-16 shows a conversion plot between Brinell, Rockwell,
and Vickers hardness numbers. Also, Figure 3-17 shows the conversion

(o>) eet
Se a
=
Ho}
Poig ESI) te
8383
Soe gm
20 &
ao] 5 = Rockwe WA ee
280 Me,
x
=O oe oO
Bees s 260 &=
3 Ed= a62 40 .ge
S'S ~@ (Ss
Vickers (120 kg, 2-mm. ball) cae
7 ol OSS Se
OROees
ROR SES
LS () 200 & °§
nace ta 20k iso 2
Bio Sea Fig 160 55
tImo0 0 +140 3s 2
10 120 7 9
Bol Ss 8) a eee 0 Se ‘ieee =
100 140 780 220 260 300 340 380
Brinell hardness, 10-mm. ball, 3000 kg load

Figure 3-16 Conversion chart showing the approximate relation between

Brinell, Rockwell, and Vickers hardness numbers. [Technical Editor Speaks. The
International Nickel Co., Inc., New York, 1943.]
 Shore hardness
30 40 50 60 70 80 90
22 24 2628 | 3234 38/4246 | 55 | 65 | 75|85})95
100 100
120 60
150 140 70
160 80
180
200 90
200
100
220 110
250 240
120
260
130
300 280
140
300
150
320
350 160
340
170
360 180
400 380 190 ksi
strength,
Tensile
400 200
210
450 420
220
440 230
460 240
500 250
480
260
500 270
550 520 280
+ 540 Brinell
hardness 290

600
(Vickers)
hardness
pyramid
Diamond 560 300
580
600
650
620
640
700
660

680
750

800

850

900

950
(0) (10) 20 30 40 50 60 70
Rockwell c Hardness

EEA tpl xf ofe| || |c | |

72 80 90 100 (110)
Rockwell b Hardness

Figure 3-17 Hardness conversion and relationship to the ultimate tensile stress
of steel. [Courtesy of The International Nickel Co., Inc.]

99
 100
Chapter 3: Mechanical Properties of Engineering Materials

plot between Brinell, Rockwell C, Vickers and Shore hardness numbers,

and the ultimate tensile strength of steel. Note that these curves are nonlinear
and that the related values are only approximate. However, tests have shown
a relationship between the Brinell hardness number and the ultimate tensile
strength of carbon and alloy steels. This relationship (equation 3-12) is
approximate and should only be employed when more definitive data are
lacking: S, = 500(BHN) psi (3-12)
Note that equation (3-12) should only be applied for steels with a Brinell
hardness between 200 and 350 BHN.

SECTION 3-23

Endurance Strength (S,,, psi)

The endurance strength (also called fatigue strength) is the maximum,
completely reversed stress under which a material will fail after it has

(a) (b)
Figure 3-18 (a) Failure of high speed pinion by fatigue
cracking of gear teeth. High tooth pressures produced
pitting and “‘shelling-out,” but the final fatigue failures,
cracking, and fracture at the base of the teeth, were caused
by repeated bending stresses on the weakened tooth
sections. Torsional vibration of the shaft was probably a
contributory cause. The teeth were initially too soft to
resist wear. Thus this is a complex cause of failure. (b) At
the right is shown an enlarged view of part of the pinion.
[Batelle Memorial Institute: Prevention of the Failure of
Metals Under Repeated Stress. John Wiley & Sons, Inc.,
New York, March 1949.]
 101
Section 3-23: Endurance Strength (S_, psi)

experienced the stress for a specified number of cycles. Thus, when a value
for the endurance strength of a material is stated, it must be accompanied
by the number of stress cycles. Fatigue failures are quite damaging because
such failures occur with no warning and at a stress much lower than the
ultimate stress.
A fatigue failure usually has its origin in some surface imperfections
(sometimes an imperfection below the surface can be the cause). These
imperfections are manifold and can be due to manufacturing methods, heat
treatment, environment, handling, raw material production, size effects,
residual stresses, and surface coatings. Failure begins with the development
of a small surface crack, transverse to the direction of the tensile stress (for
example, rotating beam test). Because the load is alternating, the crack is
continually “opened and closed”’; this motion causes the adjacent faces to
rub against one another, making these surfaces both smooth and polished in
appearance. The remaining material finally is reduced to such a small area
that it can no longer withstand the load and so breaks suddenly. The final
area appears as a gray, granular structure, conveying the false argument
that a brittle fracture has taken place. This misconception continues to
persist. The fact that there are two distinct regions in a fatigue fracture is
often useful in helping to find just how a structural or machine member
failed because it is known that the material first started to separate at the
point where the parted surfaces are shiny.

Nace

Figure 3-19 (x 3) Design error. The drawing failed

to call for an undercut just beyond the threads to reduce
the cross section below that at the base of the threads,
and to decrease the stress concentration arising at the
base ofthe last thread. This was in hard steel, 285
Brinell. The fracture was by fatigue down to the base of
the open V, without deformation: from there down, by a
sudden deformation failure. The broken parts were
photographed together. This error caused failure in
flight and resulted in a crash, killing the pilot. [Batelle
Memorial Institute: Prevention ofthe Failure of Metals
Under Repeated Stress. John Wiley & Sons, Inc., New
York, March 1949.]
 102
Chapter 3: Mechanical Properties of Engineering Materials

Figure 3-20 (x 3) Propeller shaft

that failed in flight after 150 hours.
[Batelle Memorial Institute:
Prevention of the Failure of Metals
Under Repeated Stress. John Wiley &
Sons, Inc., New York, March 1949.]

Figures 3-18, 3-19, 3-20, 3-21, and 3-22 are photographs of actual fatigue
failures.

SECTION 3-24
Endurance Limit
The endurance limit (S/, in pounds per square inch and also called fatigue
limit) is the maximum, completely reversed stress for which it is assumed a
material will “‘never”’ fail regardless of the number of stress cycles. Thus, the
endurance limit is stated without an associated number of cycles to failure.
It is generally accepted as being standard that ferrous materials that survive
at least 10,000,000 cycles of stress reversals will have an infinite life. Non-
rrousiy material

Figure 3-21 (+) Fatigue failure of aircraft engine

crankshaft starting from keyway. [Batelle Memorial
Institute: Prevention ofthe Failure of Metals Under
Repeated Stress. John Wiley & Sons, Inc., New York,
March 1949.]
 103
Section 3-24: Endurance Limit

Figure 3-22 Fatigue failure of a shaft.

Failure started at the keyway where the
transverse cracks grew to be so large that
the remaining material could no longer
sustain the load and the shaft finally
broke. Note the shiny surface from the
keyway to the break, which is granular in
appearance. [Courtesy of Joseph T.
Ryerson and Sons, Chicago, IIl.]

mag not have an enduranc it; therefore only an endurance

streng be established for these materials.
Three types of endurance tests are performed. These are tension, torsion,
bending, and combinations of these. The most common test is that of reversed
bending and is performed by an R. R. Moore rotating beam test. Figure 3-23
shows such a machine and a standard test specimen. To perform a test, the
specimen is loaded with a selected weight. Turning on the motor rotates
the specimen, but not the weight. The longitudinal fibers of the specimen
alternately feel a tensile and compressive stress for each rotation. Rotation
is continued until rupture takes place, at which time the number of cycles
(revolutions) to rupture is noted. ASTM Special Technical Publication
No. 91-A entitled “Tentative guide for fatigue testing and the statistical
analysis of fatigue data” clearly specifies the procedure that is to be followed
and how the data are to be evaluated. The results of the tests are plotted on
log-log or semilog paper where the rupture stress is the ordinate and the
number of cycles to failure the abscissa.
Figure 3-24 is a typical plot of what the scatter of test points would look
like. More in accord with the ASTM fatigue test, Figure 3-25 shows not
only the scatter but the upper and lower survival levels. This figure demon-
strates that an 84% survival rate represents about 1.4 times the standard
deviation from the mean with a confidence level of 50%. This implies that
the mean of future groups, from the same test batch, would have a survival
rate greater than 84%, 84% of the time. A greater spread from the mean
(for example, 2.8 times the standard deviation) would result in a survival
rate of 99.5%.
 Drill and tap 1/4 in.
20 U.S.F., depth 7/16 in.

0.480” + 0.001”

5/8" per ft taper

0.300’ + 0.003”

(b)

Counter
Bearing housing Flexible coupling
Specimen

Fulcrum High-speed motor

W/2 W/2 Bearing housing
(c)

Figure 3-23 (a) Rotating beam test specimen. (b) The R. R. Moore rotating
beam fatigue testing machine. (c) Schematic of the R. R. Moore testing machine.
[Courtesy of Satec Systems, Inc.]

104
 105
Section 3-24: Endurance Limit

45,000

40,000

35,000

psi 30,000
Stress,

25,000

ol104Ee aah10° =
106 107 108
Number of cycles for failure

Figure 3-24 Endurance diagram for 0.37 % carbon steel annealed.

[From Elements of Strength of Materials, 2nd ed., by S. Timoshenko
and Gleason H. MacCullough © 1940 by Litton Educational
Publishing, Inc. Reprinted by permission of Van Nostrand Reinhold
Company. ]

120 4340, OQT 860°F, polished

16% survival
100 ow —a
90
80

a 50% survival
60 4340, OQT 860°, polished, with 0.025-in. hole
Stress,
ksi
16% survival
50 SEE

40;-- *asurviva OF ~~ ---— -

30
10° 2X108 5xX105 1068 2x106 5x106 107
N-number of cycles to failure

Figure 3-25 Endurance strength bands for rotating beam test.

Notice for ductile steel that the upper and lower bands tend to
converge at 10° cycles for ductile steel. [From V. M. Faires: Design
of Machine Elements, 4th ed., The Macmillan Co., New York, 1965,
p. 102.]

ASTM Special Technical Publication No. 9]-A discusses in greater detail

the important ramifications to be considered in various statistical aspects
of fatigue testing. Most often, fatigue curves as well as tabulated values of
endurance strengths and endurance limits are based upon a 50 % probability
curve. As a result, designers do not resort to using scatter-band curves like
Figure 3-25 unless they are concerned with a design that involves a statistical
approach. When the designer requires information of the highest order of
reliability, he should always contact the manufacturer and/or run tests.
 106
Chapter 3: Mechanical Properties of Engineering Materials

100,000 atx € | |

ore Denotes
80,000 specimen
unbroken
Figure 3-26 S-N diagrams from
endurance tests. © 1.20°% carbon steel,
60,000 quenched and drawn 860°F. @ SAE 3420,
quenched and drawn 1200°F. @ Alloy
structural steel. @ SAE 1050, quenched
and drawn 1200°F. © SAE 4130,
40,000
normalized and annealed. © Ordinary
psi
S,
Maximum
stress, structural steel. ® Duralumin. ® Copper,
alternately annealed. © Cast iron.
[From Oscar J. Horger (ed.): ASME
Handbook, Metals Engineering Design,
2nd ed. McGraw-Hill Book Co., New
York, 1965.]

10,000
100,000
1,000,000
10,000,000
100,000,000
Cycles to failure, VV, log scale

Figure 3-26 is a S-N plot showing the results of tests for several different
materials. Notice that the curves for steel display a decided break (or “*knee”’),
and many ofthese knees occur before or near 10 x 10° cycles. This establishes
the reason for using this value as the basis of the endurance limit for steel,
although many steels have values between: 2 x 10° and 10 x 10° cycles.
Using 10 x 10° cycles as the criteria for infinite life design, then, is rather
conservative.
It is of interest to note that nonferrous materials do not show a break in
their S-N curves and, as a result, a distinct endurance limit cannot truly be
specified. Instead a stress value for a particular number of cycles will be
given for various nonferrous materials. However, it has become customary
to use the endurance strength at 100 x 10° or 500 x 10° cycles as the
endurance limit for nonferrous materials.
Experiments have shown little correlation between the endurance limit
and such mechanical properties as yield strength, ductility, and so on.
However, some relationship between the endurance limit and tensile strength
for unnotched and polished specimens tested by the rotating beam method is
indicated experimentally. For steels not exceeding a tensile strength of
approximately 200,000 psi, an endurance limit ratio (also called fatigue
ratio) of 0.50 can be used when no specific endurance information is available.
Thus we have (for 50°, survival)

ee OS: for wrought steel where S,, < 200,000 psi

and BHN < 400 (3-13)

S}, = 100,000 psi for wrought steel where S,, > 200,000 psi
 107
Section 3-24: Endurance Limit

yy Wee for cast steel and cast iron (3-14)

S, = 0.388, for magnesium casting alloys and magnesium

wrought alloys (based on 10° cycle life) (3-15)
S, = 0.45S, for nickel-based alloys and for copper based
alloys (3-16)
Ss S85) for wrought aluminum alloys up to a tensile
strength of 40,000 psi° (based on 5 x 10% cycle
life) (3-17)
Sp US, for cast aluminum alloys up to a tensile
strength of 50,000 psi (based on 5 x 10% cycle
life) (3-18)
where all of the relations are based on a 50% survival rate.
The endurance limit for reversed axial loading of a polished, unnotched
specimen seems to have a 15% lower value than that for reversed bending.
As a guide to design (since the results are based on widely scattered data)
we can use relation (3-19) for reversed axial loading of steel specimens.

S), = 0.85 x endurance limit in reversed bending = 0.43S, (steel only)

(3-19)
On first thought, one would expect the axial fatigue values to agree “‘exactly”’
with the bending fatigue values because in bending the longitudinal fibers
are alternately stressed in tension and compression. The discrepancy is
attributed to the great difficulty in applying a purely axial load without
introducing a bending component. Also, in axially stressing the test specimen,
all of the longitudinal fibers are subjected to the same average stress. In
bending, however, the stress varies linearly from the neutral axis to the
outer fibers.
For reversed torsional testing of polished, unnotched steel, specimens,
the endurance limit is approximately 58% of that in bending. Therefore,
we have relation (3-20) for the torsional endurance limit.
S., = 0.58 x endurance limit in reversed bending
(3-20)
— 02295; (steel only)
For cast iron and copper respectively, relations (3-21) and (3-22) can be
used.
S,, = 0.8 x endurance limit in reversed bending Ae

=) 32S) (for cast iron)

(3-21)
Si, = 0.48 x endurance limit in reversed bending 3.22)
(3-
\|S 0225, (for copper)

> For higher values of tensile strength, the endurance limit peaks out at about 16,000 psi (at
55,000 psi tensile strength) and then dips somewhat at still higher values of tensile strength.
 108
Chapter 3: Mechanical Properties of Engineering Materials

All of the previous discussion relating to the endurance (that is, fatigue)
strength is based on testing a polished laboratory specimen of a fixed size
and geometric shape. As a result, the designer must modify these data by
those factors that adversely affect results determined under laboratory
conditions. The factors are a direct result of manufacturing, heat treating,
geometry, environment, and so on. In fact, F. B. Stulen, H. N.Cummings,
and W. C. Schulte [8] suggest that “‘the fatigue strength of a part based on
the actual alternating stress seldom exceeds 70 per cent of the handbook
value.” J. Marin [9] suggests that the list of factors given in Table 3-1 be
considered when determining an endurance strength or limit value to be
used for design purposes.

Table 3-1 Factors Influencing Fatigue Strength

Material Effects Environmental Effects

1. Chemical composition 1. Corrosion

2. Failure condition 2. Rest periods

3. Material variation 3. Superimposed static stress

4. Size and shape 4. Temperature

5. Speed 5. Varying amplitudes

. Understressing & overstressing . Exposure to nuclear radiation

Manufacturing Effects Miscellaneous Effects

. Fretting fatigue and fretting . Surface fatigue

corrosion

its). Heat treatment tw. Combined stresses

3. Method of manufacture

4. Stress concentration

nN . Surface treatments

Table 3-1 contains many factors that influence fatigue. These factors must
be used in design applications with great care because much of the available
information is based on specific specimens and tests. Literature pertaining
to the effect of these factors on fatigue properties is quite scattered. Thus,
the reader should make use of a few source references which, in turn, will
lead him to specific documents having much greater detail. Of course, it
goes without saying that one must always be ‘on top of” current work in
order to take advantage of the latest information. Among the many excellent
references available, (1) Metal Fatigue by G. Sines and J. L. Waisman,
McGraw-Hill Book Co., New York 1959; (2) Residual Stresses and Fatigue
in Metals by J.O. Almen and P. H. Black, McGraw-Hill Book Co., New York
1963; and (3) references [4], [10] and [11] cited at the end of this chapter,
have been found most useful.
 109
Section 3-25: Effects of Material Variation—Reliability Factor

A full discussion of all the factors listed in Table 3-1 is beyond the scope
of this text. Here we shall consider only those factors for which quantitative
data are available. This information will be used to adjust the endurance
limit S), to a value S,—a working endurance limit. However, for completeness,
some brief comments will be made in Section 3-30 on some of the more
significant nonquantitative factors in Table 3-1 that can affect the value of
S;0riS,;

SECTION 3-25

Effects of Material Variation—

Reliability Factor
Material variation is one very important factor often overlooked by a
designer or analyst when evaluating a mechanical component. Figures 3-24
and 3-25 show typical scatter values and scatter bands obtained in fatigue
testing of metals. As mentioned earlier, most plots and tabulated data of
endurance values are mean values and thereby imply a 50% survival rate.
Since fatigue analysis is, at best, an estimate, it becomes desirable to design
for some level of reliability that will acount for material variations in reported
S-N data.
Lacking specific reliability information, it is suggested [8] that “8 per
cent of the long-life strength may be assumed as a standard deviation
provided the material is of good quality.” Fatigue tests on various alloys
have generally displayed a statistically normal curve distribution, so that
the fatigue limit for fixed percentages of survivals can be obtained by sub-
tracting a specified (that is, desired) number of standard deviations from the
mean fatigue strength. Viewed from the standpoint of a fatigue strength
reduction factor, we can write equation (3-23).
Cr = 1 — 0.08 (D.M.F.) (3-23)
where C, = fatigue strength reduction factor based on reliability and
D.M.F. = deviation multiplication factor from Table 3-2.

Table 3-2 Reliability Levels

Survival Rate, % Deviation multiplication factor (D.M.F.)

90.00 1.28

95.00 1.64

98.00 2.05

99.00 2333

oo 0 3.08

99299 3.62
a
RRRR ATRSE EES
 110
Chapter 3: Mechanical Properties of Engineering Materials

SECTION 3-26

Influence of Size—Size Factor

The influence of size upon fatigue strength values can be a significant
factor. It would seem to be obvious that there must be some effect due to
size as a rotating beam test specimen is 0.3 in. in diameter at the smallest
section. Although little quantitative data are available, due to the unlimited
testing that would be required, endurance limits of the same material and
hardness have been found to tend to decrease with increasing size for
specimens subjected to bending and torsional loading (see Figure 3-27).
The rationale for this decrease of endurance strength with increasing size
is that a larger specimen is more apt to have internal defects (for example,
inclusions, nonuniform cooling, and so on).

120 12

100 1.0

80 0.8
Bending Figure 3-27 Size effect—bending,
g enditorsion axial, and torsional loads. [From C.
60 3 06 % Lipson and R. C. Juvinall: Handbook of
£ ~ Stress and Strength. The Macmillan Co.,
3 New York, 1963.]
r} T a 0.4
a

2
3
of 20
0.3-in.
diameter
limit
Endurance
% c 0.2

——
ee)

0
Sic 1.0 113) 2.0 2.5) 3.0

Size, in.

For bendi
and torsion
ng ofspecimens greater than 0.3-in. diameter, the
endurance strength is reduce dup to specimens of 0.5-in. diameter.
by 15°%
For speci
larger than 0.5-in. diameter, the reduction remains relatively
mens
to
approxim
constant up ately 2-in. diameter after which it begins to increase.
One cannot generalize about the reduction required for larger specimens,
but it can be as high as 25~30°,. Prudent design would suggest using a
Sere er re en For critical designs, testing of
component specimens must be performed for more accurate values. However,
lacking specific information, relation (3-24) may be used as a fatigue strength
reduction factor® due to size.
C. = 0:80 (for bending and torsion of specimens from 5
in. dia
to 2 1n. dia) (3-24)

° For varying loads of 1000 cycles or less, it is assumed that the loading is “‘static” and so Eo = 10)
 iT]
Section 3-27: Effects of Methods of Manufacture—Surface Finish Factor

For specimens in axial loading, tests have indicated that no size correction
is necessary. The hypothesis for this is that, regardless of specimen size, there
is a zero Stress gradient.

SECTION 3-27

‘Effects of Methods of Manufacture—

Surface Finish Factor
The method of manufacture has a most important effect on the endurance
properties of metals. Different manufacturing techniques, such as casting,
hot forging, cold forming, turning, grinding, polishing, welding, riveting,
bolting, and so on, all contribute to altering the endurance strength of the
material. One of the more obvious effects is that of the surface finish or
condition of a machine or structural component.

Machining
A great deal of work has been done to elucidate the effect of surface
finish on fatigue properties and the results are so diffuse that they cannot
be adequately covered. However, Chapter 7.5 of reference [11] cites many
references that contain more detailed information for the interested reader.
We will merely review those aspects that are of immediate use and concern
to the designer.
Noll and Lipson [12] collected the data concerning surface finish from
various sources and plotted average curves of endurance limit versus tensile
strength for each type of surface encountered. Figure 3-28 is the result of
this task. The data used for Figure 3-28 were further employed to plot a
series of curves that relate the tensile strength of each type of surface finish
as a function of per cent of the endurance limit. These curves can be used
as surface correction factors for the endurance limit of a steel. The curves
and factors are shown in Figure B-3 of Appendix B.
The surface finish reduction factor C, is used as a multiplying coefficient
in modifying the endurance limit of a material in the same way that Cp and
Cs are used.

The “hot-rolled curve” in Figure included data from

specimens having slight surface irregularities, oxide, and scale defects as well
as some decarburization.’ The “‘as-forged”’ curve included specimens with
larger surface irregularities as well as contained oxides and scale defects in
addition to having had a total surface decarburization. This combination of
surface defects and decarburization contributes greatly to reducing the
endurance limit.

7 Decarburization (see Section 2-9) not only reduces the static strength properties but also
lowers the endurance strength and leaves a residual surface tensile stress.
 1BRY
Chapter 3: Mechanical Properties of Engineering Materials

Brinell hardness

120 160 200 240 280 320 360 400

100 i

Endurance
ksi
limit,

Tensile strength, ksi

Figure 3-28 Endurance limit versus tensile strength for

various surface finishes of ferrous metals and alloys. The
results are based on unnotched test specimens subjected to
reversed bending. [From C. Lipson and R. C. Juvinall:
Handbook of Stress and Strain. The Macmillan Co., New
York, 1963.]

All the curves plotted are conservative ; thus, any improvement in machin-
ing, scale removal, reduction of surface irregularities, or elimination of
decarburization will increase the value of C,. To obtain a more refined
value of C,, one can interpolate between the polished, ground, and machine
curves using values from Figure 4-57 for different machining operations.
With non-ferrous metals and alloys such as aluminum, magnesium, copper,
brass, and so on the surface finish reduction factor need not be considered
when making fatigue calculation because the suppliers of these materials
have accounted for the surface finish in stating endurance strengths.

Welding
Welding, whjch has become a popular method of manufacture particularly
since the advent of automatic welding machines, can cause greatly reduced
endurance properties of materials. The basic reason for this is that a weld
produces a geometric change of shape at the joint and thus acts as a geo-
metrical stress concentration factor. Figure 3-29 shows both a single and
double butt weld. Load F causes stress lines to flow as shown, thereby
causing stress increases at points 1, 2, 3, and 4. In addition, poor welding
techniques, which prevent full penetration at such points as 1-2 and 5 and
permit porosity and slag inclusions, contribute to stress intensity increases,
 Ts
Section 3-28: Effect of Stress Concentration

Single butt weld Double butt weld

' Figure 3-29 Points of stress concentration on typical butt welds.

causing a reduction of the weld’s endurance strength. Grinding off the

reinforcing weld material (that is, that part of the weld above the material
surface) improves fatigue strengths by as muchas 10-20 % [21]. This improve-
ment was obtained for both longitudinal and transverse butt welds.
Reinforcing straps placed over a butt weld reduced its fatigue strength.
Generally welding is confined to hot-rolled, annealed, or normalized materials
because their strength is based on their composition rather than their heat
treatment or cold working. Only certain alloy steels, austenitic stainless or
heat treated aluminum are welded, and then mainly by spot or plug welding.
Welded joints that are air cooled rapidly as compared to the base metal
should be annealed after welding to reduce residual tensile stresses and
possible cracking. Some weld materials, however, become brittle with
annealing. Heavy sections should be preheated to a proper temperature
before welding to avoid severe tensile stresses in the weld on cooling.
More detail and methods of analysis will be found in Chapter 17. Also,
the Welding Handbook of the American Welding Society has a brief but
excellent discussion concerning fatigue in welding. Until more extensive
and reliable information is available, we may use the strength reduction
factors as listed in Table 3-3.

Table 3-3 Endurance strength reduction

factor C, for welds

Type and Location of Weld Cy

Reinforced butt weld 0.833

Toe of transverse fillet weld 0.667

End of parallel fillet weld 0.370

T butt joint with sharp corners 0.500

SOURCE: C.H. Jennings, Welding Design. Trans. ASME, Vol. 58, pp.
497-509, 1936.

SECTION 3-28
Effect of Stress Concentration
The causes of stress concentrations (also called stress raisers) are varied
and numerous. Some causes, namely, surface finish, nonmetallic inclusions,
and so on have already been discussed.
 114
Chapter 3: Mechanical Properties of Engineering Materials

A stress concentration (or stress raiser) is any material condition that

causes the local stress to be greater than the nominal stress. For some
conditions, it is possible to introduce a fatigue strength reduction factor
(that is, Cy for surface finish) and for others (for example, fretting) it is
possible only to rely on qualitative data and direct experience. Geometry or
shape of the specimen, one of the important stress concentration contributors,
does have (at least for simple cases) a rational basis. In those instances
which do not yield to analytical techniques, experimental methods (that is,
photoelasticity, brittle coating, membrane analysis, strain gage) or testing of
the actual component to be used have provided large quantities of useful
data for design. In fact, much work is still proceeding to determine these
factors for combined stresses.
The definition of the stress concentration factor is
maximum stress at the section of interest pax |
nominal stress at the section of interest Oo
(3-25)
K, =
iC max
rs ‘
Ss cr

where K, is considered as the theoretical (that is, geometric) stress concentra-

tion factor as it is a function solely ofthe specimen geometry. K,,isthe
theoretical factor
for shear.
Consider the flat plate in Figure 3-30a. Observe that the force flow lines
are uniformly spaced at each end of the plate. At the notch, however, the
outside force lines must change direction and are no longer uniformly
spaced. Thus, within the vicinity of the notch, more force lines are flowing
through a unit area. The local stress is thereby increased. The force flow line
that is caused to bend the most represents the highest stress. As the lines move
away from the notch, becoming straight again, the stress is reduced to its
nominal value.
Figure 3-30b shows a stepped shaft that has been divided into four hollow
tubes and a solid center shaft. The concentric tubes of the larger and smaller
diameter sections are joined by smooth lines as indicated. The thickness of
the outer tube at the point where the sections join is relatively thin. This

= os

(a) (b)

Figure 3-30 (a) Tensile stress flow lines. (b) Equitorque stress tubes for a
circular stepped shaft.
 13)
Section 3-28: Effect of Stress Concentration

means that the outer tube experiences a high stress. Note that the wall
thickness increases with decreasing radius because the thickness is chosen
so that the angle of twist per unit length is the same for each tube. This
requirement is necessary because the concentric tube model must not
experience any angular slip if it is to be equivalent to a solid shaft. As with
a flat plate, the force lines are ‘“‘crowded together” as they flow past a point
of changing (that is, restrictive) geometry.
An additional example illustrating the influence of geometric change is
that of a semi-infinite plate in tension with a small hole in the middle as
shown in Figure 3-31. Theoretical analysis indicates a stress distribution
such that the stress line tangent to the hole o,,,, is 3 times as large as the
nominal stress o9. Thus, the theoretical stress concentration factor K, is 3.
The many subjected to different loadings are
charted or ee h.appendix is by no means complete,
and the reader is advised to investigate the references cited and other sources
for those geometric combinations not included in this text.

Figure 3-31 Stress distribution for a semi-infinite plate

with a hole. Plate is in tension.

Look once again at Figure 3-30a. If the force lines are considered as
fibers, they tend to straighten out at those points where they pass a dis-
continuity. In this case, it is the V notch. As a result, the distance between
the notches is prone to increase slightly. This is called the biaxial effect due
to geometric stress raisers. It results in a slight reduction of the theoretical
stress concentration factor K,. However, the reduction is small (of the
order of 10°) and is generally ignored as a modifying factor in most design
applications.
Theoretical stress concentration factors K, are dependent solely upon
the geometry of the specimen concerned. If all materials were uniformly
homogeneous and free of surface marks or scratches, one would be justified
in using K, ‘‘as is” for fatigue stress calculation. However, materials are not
 116
Chapter 3: Mechanical Properties of Engineering Materials

homogeneous or free of surface defects. They can crack just as readily at a

tool mark or scratch as they can at a fillet, keyway, or edge ofa hole. Fatigue
tests have shown that the theoretical stress concentration factor is rarely
obtained (except for some high strength steels). Instead, a value smaller
than K, is indicated. It therefore becomes necessary to define a fatigue
stress concentration factor K,.
endurance limit of a specimen free of notches
K,= (3-26)
endurance limit of a notched specimen
Since unlimited testing would be required to produce values of K,, it
is desirable to relate the theoretical K, to the fatigue K, for various notch
sizes, materials and heat treatments. R. E. Peterson [14] suggests the ratio
in (3-27) called the notch sensitivity factor q

|
q ss
Kaan 3-27
(3-27)
or for torsion (or shear)

eee (3-28)
Se Ke aaa :

Solving equations (3-27) and (3-28) for K , gives SS

W
hea 1 a ge ah (3-29)
Kee td ee) (3-30)
Equations (3-29) and (3-30) can now be used to find a value of K,, provided
one knows what to assign for gq.
Although little information exists concerning q for a large number of
materials, Figure B-2 in Appendix B) provides data for ferrous material of
different hardness and an aluminum alloy. In those instances where one
lacks specific data concerning q, it is not unreasonable to choose q to be
0 <q < 0.2 for materials that are prone to be insensitive to notching (for
example, low strength cast irons). For a cast iron ASTM No. 50, T. E. Eagan
[15] states that one may use K, = 1.25. On the other hand, if one is uncertain
ofa value for q, #
The reader should note that K, is not only the theoretical (or geometric)
stress concentration factor but is based on static loading conditions.
Consequently, brittle materials such as cast iron and concrete, which do not
display a yield point, will suddenly fail when stressed to their ultimate
value. Therefore, in all cases of static loading involving brittle materials, K,
must be used to account for any stress raiser. One may write for the static
loading of brittle materials

N=—" Sa
(fortension)
og
N= ae Dae
(for compression) °
(3-31)
19; 1% ¢

where S,, and S,, are the ultimate strengths for tension and compression,
 117
Section 3-28: Effect of Stress Concentration

respectively, N is the factor of safety, and o, and a, are the nominal calculated
stresses for tension and compression.
Brittle materials are rarely considered for fatigue stress applications
because most such materials have low fatigue strength properties, and their
use would require a large factor of safety. However, certain cast irons (for
example, nodular) do have ‘‘acceptable’’ values for endurance limits and
can be analyzed rationally by using the Goodman line of failure (see
Chapter 6).
Considerable effort has been expended on explaining the theoretical stress
concentration factor; therefore, the reader may be surprised to learn that its
use is normally restricted to ductile materials subjected to varying (that is,
fatigue) loads (note its inclusion in this section on fatigue). The explanation
for the restriction is quite simple. It results from the fact that a ductile
material will not fracture immediately past the yield point but will “give”
instead, becoming plastic. This sues the stress concentration at ihe dis-
continuity. Consequently, tk tr n factc ;
S. However when the toads1SSTORE
varying, fatigue cracks can develop so that plastic yielding at a point of
stress concentration is of no value in delaying possible failure. Thus, 1/K ;is |
used asamultiplying coefficient forcorrecting the endurance limit! This is
discussed in Section 3.29.
The problems encountered in “real life’? situations are, as one might
expect, much more complex than those ordinarily encountered in an ‘“‘ideal”’
classroom situation. For example, it is most likely that a machine component,
such as a drive shaft, will have portions of different diameters—a keyway,
oil holes, press-fitted surfaces, and possibly other areas that act as stress
raisers. The novice very often will consider the point of the shaft that has
the highest stress concentration factor K,. This may be a grave mistake
because the mechanics of the problem and the location of the stress raiser
are the deciding factors for finding the “‘worst”’ spot for analysis. Admittedly,
such a location may sometimes be chosen by observation, but it is strongly
suggested that all locations be considered before one is decided upon as the
“worst of the lot”. Chapter 7, dealing with shaft design, clarifies this point.
Quite often must be considered for the
same location. For example, an oil hole at the fillet of a stepped shaft is
representative of such a case. The meager information available indicates
that the cumulative results of two such factors are greater than the individual
factors but less than the ied of son oe HOWENeE, roils and
Juvinall suggest heoretical st: concentra
(SeeemeiantheenetanirshnA OECD, G-29)) ¢or re
( 30)t
towept K,¢or me
A good designer must do his best to reduce or eliminate areas with severe
stress concentrations. Inspection of Figure 3-30 suggests that the intensity
of the stress concentration factor would be much reduced if the paths of
the force flow lines did not cause sharp changes or crowding of lines. Figure
3-32 illustrates various ways in which geometric changes can contribute to
C/7D

right view
 119
Section 3-29: A Working Equation for the Endurance Limit

Figure 3-32 Opposite: Different ways to mitigate stress concentrations. The

recommendations are obvious except perhaps for case (i). In this case, by cutting
away part of the nut bearing surface, the stress has been transferred from the bolt
to the nut preventing it (that is, the bearing stress) from being concentrated at the
bolt threads.

reducing stress concentrations. Notice that the modifications shown in

Figure 3-32 allow the force flow lines to pass through the specimen more
smoothly than if there were no modifications. In other words, it is desirable
to reduce or eliminate abrupt changes in the direction of these lines. Figures
3-33b, d, and f are also ways of reducing stress concentrations due to force
fits.

Slots equally
Damage surface spaced on the
due to fretting Damaged areas
circum ference
caused by fretting
Varying Varying
torque torque Varying bending
moment

SaaS Sas
5
4

d) (e)

Figure 3-33 (a) Press-fitted shaft in torsion. (b) Relief of fretting problem due
to torsion by permitting greater movement of the housing and thereby reducing
slip. (c) Press-fitted shaft in bending. (d), (e), and (f) Alternate ways to relieve
fretting caused by bending. Further improvement can be obtained by prestressing
(that is, surface compression) that part of the shaft that has not benefited from the
increased flexibility of the housing.

SECTION 3-29

A Working Equation
for the Endurance Limit
The various endurance limit reduction factors are used to establish a
modified (that is, working) endurance limit S,.
Sy CR CLONC WIGS, (3-32a)
Ses a CrCsCpCw(1/K 55)Si5 (3-32b)
 120
Chapter 3: Mechanical Properties of Engineering Materials

where
Cr = reliability factor (from equation 3-23)
Cs = size correction factor (from equation 3-24)
C, = surface finish correction factor (from Figure B-3, Appendix B)
Cw = weld correction factor (from Table 3-3)
S;, = endurance limit of the material
K , = fatigue stress concentration factor (from Appendix B)

Kye = fatigue stress concentration factor for shear

The coefficients in equations (3-32a and b) are used as required. Equation
(3-32) is extremely important in fatigue problems. The designer may further
decide to reduce the value of S, calculated by equation (3-32a) by factors due
to tensile surface stress, nonmetallic inclusions, effect of heat treatment, grain
flow and size, and so on. Unfortunately, these factors have no quantitative
values. The values assigned (that is, smaller than 1 if the effects are deleterious
and greater than | if the effects are beneficial) will depend on the designer’s
experience, judgement, and job observation. These and other nonquantit-
ative factors affecting S’, and S’, are discussed in Section 3-30.

SECTION 3-30

Effect of Some Important Nonquantitative

Factors
Fretting fatigue
Fretting fatigue takes place in those instances when two mating parts,
which are in close contact, rub against each other because of vibrations or
repeating loads. The rubbing is usually confined to local areas and contributes
to the deterioration of the contact surfaces. Deterioration may be so severe
that the damaged surfaces actually begin to crack. This must ultimately
lead to a reduced fatigue strength, and possibly to eventual fracture. Applica-
tions in which fretting fatigue and fretting corrosion problems must be
considered are i

fc ge : cal springs

Note that the examples given represent cases such as

interference fits, bolted and riveted connections, and ball and roller bearings,
which have little if any pressure but develop high contact pressures due to
repeated loading that leads to fretting failures.

Fretting corrosion
Fretting corrosion is often classified as a part of fretting fatigue but is
actually quite distinct and has a nature of its own. In fretting corrosion, the
 121
Section 3-30: Effect of Some Important Nonquantitative Factors

repeated loads are too weak to cause fracture by fatigue and any damage
to the material (cracking or corrosion) takes place at the edge of the con-
tacting parts or within the local contact area. In fact, we may think of
fretting corrosion as a type of damage that is associated with wear, surface
damage, and accumulated debris. On the other hand, in fretting fatigue,
we may think of the damage in terms of direct fracture.
Factors such as atmosphere, materials, temperature, humidity, load, load
frequency, number of cycles, relative motion between parts (that is, slip),
lubrication, surface finish, and closeness of fit all play a part in governing
the rate of fretting corrosion. Few of these factors can be controlled; but
the designer still wishes to know how to eliminate or reduce fretting corro-
sion and fatigue. He may consider (1) changing the combination of mating
materials [see reference 16], (2) using a lubricant such as MoS, (molybdenum
disulfide) in combination with corn syrup, grease, or petroleum jelly, (3)
insertion of plastic (for example, teflon, nylon) or rubber shims between
the contacting surfaces, (4) electroplating with cadmium, nickel, or zinc,
(5) introducing surface compressive stresses by mechanical means (for
example, shot peening, rolling) or by a heat treating method (method 5 is
particularly effective with interference fits) or (6) increasing the contact
load so that relative motion is eliminated. (Of course, if the relative motion
is not completely eliminated by increasing the contact load, the fretting
damage will worsen, and the designer must make design adjustments to
relieve the harmful effects of fretting.) Figure 3-33 shows some ways of
reducing fretting in press-fitted shafts subject to torsion or bending.

Heat treatment
Proper heat treatment of metals can be extremely beneficial in improving
the fatigue properties of a metal. In particular, it is one of several ways
residual stresses may be introduced which, if properly used in design, have
beneficial effects in reducing fatigue damage. Residual stresses are classified
into two categories. These are macrostresses and microstresses. Macro-
stresses are stresses that can be determined quantitatively and are based on
elastic-plastic analysis. Microstresses are those stresses concerned with the
granular structure, its loading and deformation or slip. These stresses are,
therefore, described by qualitative means. Before proceeding further, it would
be helpful to describe how a residual stress might be introduced into a metal.
For purposes of illustration, assume that a stress-free rectangular mild
steel bar is subjected to a pure bending moment M, large enough to cause
the bar to become completely plastic. Looking at Figure 3-34a we see that
the path OAB on the stress-strain curve has been traversed. The stress
distribution on the bar appears as shown in Figure 3-34b. The bending
moment is now slowly but completely removed. Removing M allows the
material to spring back and the stress-strain curve follows Hooke’s law.
It proceeds along BC, parallel to OA as shown in Figure 3-34a. Note that
there is now a permanent strain OC. The unloading stress distribution is
 {22
Chapter 3: Mechanical Properties of Engineering Materials

4
|
Tensile stress

/
S
- m
SL
yp
M M
( \
- h

( +)
Compressive
elo Tensile mee . "
strain strain us p (b)

| a3 ~_

Plastic - Elastic | 1.58,

— _ — —
strain recovery |r

—————otdsthai

Compressive stress
(a)

(c)

Figure 3-34 Ideal elastic-plastic model for obtaining a residual stress in the
outer fibers of a bar (or beam). The magnitude of the stress in part (a) is obtained
by noting that, because of equilibrium, the moment of the triangular stress
distribution taken about the neutral axis is equal to the moment of the rectangular
stress distribution taken about the same axis. The magnitude of the residual stress
is obtained by algebraically adding the stress values of parts (b) and (c) to obtain
part (d).

shown in Figure 3-34c. Superimposing the elastic unloading curve in

Figure 3-34c on to the perfectly plastic loading curve in Figure 3-34b results
in the final stress distribution as shown in Figure 3-34d. Part d shows the
residual stress distribution throughout the bar caused by plastic flow. Close
observation of this part d indicates that the top fibers, which were originally
loaded in compression, now have a residual tensile stress, and the bottom
fibers, which were originally loaded in tension, now have a residual stress
of compression.
The quenching process used in the course of heat treating a steel can also
introduce residual surface stresses that are either tensile or compressive.
Which of the two stresses finally appears depends on whether the quench
was thermal or was both thermal and metallurgical. For example, consider
 123
Section 3-30: Effect of Some Important Nonquantitative Factors

Residual stress

(a) (b)

Figure 3-35 Resulting residual stress pattern caused by quenching.

a round bar of steel, as shown in Figure 3-35, that has been heated above its
upper critical temperature (for example, 1500°F) so that the steel is basically
all austenitic. The specimen is then quenched (for example, in oil, water, or
brine) quickly and forms a metallurgical structure (that is, pearlite, bainite,
or martensite) depending on the quenching temperature and the time.
Analyzing this rather rapid process sequentially, we find that, due to
cooling, the surface of the bar (shown by the gray area in Figure 3-35) tends
to contract. Since it is, however, part of the hotter core, it is prevented from
doing so. The surface layer, then, is momentarily in tension while the core
is in compression.® Next, as the outside layers harden and form crystals
(for example, martensite), the tensile stress is somewhat reduced because the
transformation to martensite causes a slight expansion in volume. By now,
the surface is rather cool; the core, however, still being hot, continues to
cool and therefore contracts. This contraction of the core causes the outside
to contract again and again be placed in compression or reduced tension
while the core is in tension. Finally, as the core reaches the temperature to
form martensite, it expands. This expansion now acts on the outer layer,
which is quite hard and will not readily move with the core. As a result, the
stress in the outside layer of the bar is changed to tension or reduced compres-
sion. The resulting stress distribution would have a profile as shown in
Figure 3-35b.
On the other hand, if the bar had been heated to 1200°F, a temperature too
low to form austenite, it would not have experienced a phase transformation
with its associated volume increase on cooling. The outer layer of the bar
shown in Figure 3-35 would cool and tend to contract before the core.
However, it is prevented from doing so readily by the core. The outer layer
is subjected to tension whereas the core is in compression. As the core cools
further, it tends to contract and place the outer layer in compression, because
this layer has already cooled to about room temperature. Consequently,
the core is in tension. The final stress distribution is that shown in Figure
3-35c.
Note that the procedure for heat treatment involved both thermal con-
traction and phase-change volume expansion, whereas the latter process

8 Surface cracks may develop in brittle material due to high temperature gradients and subse-
quent heating or cooling. This cracking is called spalling.
 124
Chapter 3: Mechanical Properties of Engineering Materials

involved only thermal contraction. Quenching in general, regardless of the

initial temperature, must be performed carefully, otherwise warping or
surface cracks may develop that would be injurious to the material when
subjected to fatigue loads.
Since fatigue failures are always due to tension, a residual compressive
surface stress would help to counteract the effect of fatigue stressing.
In completing a heat treating process, temperating is often specified to
relieve unwanted residual stresses. Care must be exercised, however, in
order that a desired surface compression is not removed and perhaps even
changed to tension.
The method described above (that is, quenching and subsequent tempering)
will not only produce a compressive stress on the surface of a steel part
but will also result in hardening the section throughout. The gradations of
hardness as we move from the surface towards the middle portions will
vary according to rate of quench, quenching medium, and component size
and shape (see Section 2-8).
It is sometimes desirable to heat treat only the surface of a part in order
that its surface be hard while its middle remains tough. Two popular methods
used to attain this objective are flame hardening and induction hardening
(see Section 2-9). Both methods heat the surface of the steel, after which it is
rapidly quenched. This produces a residual surface stress as well as a
hardened surface. If the steel is of high enough carbon or alloying content,
the heated surface will undergo a volume increase due to the formation of
martensite. The surface, then, will be left with a residual compressive stress.
For induction hardening, the depth of hardened case is about 0.100—0.200 in.
depending on the carbon content, alloying constituents, and quenching
medium. For flame hardening, the case varies between % and { in. thickness.
It is also of interest to note that flame or induction hardened parts that
are press-fitted (therefore subject to fretting erosion) show greater resistance
to fatigue failure due to the residual surface compressive stress.
Any form of heat treatment in addition to the ones mentioned will produce
a residual stress in the component part. The kind of stress and whether it is
beneficial or injurious depends on many factors, and the size and shape of
the part are of paramount importance. Some of the other factors are carbon
content, heating temperature, type and temperature of cooling medium, and
cooling rate. If little thought is given to a heat treating déperation and no
account is taken of the physical shape of the part, not only can unfavorable
residual stresses occur but actual failure due to what is called ‘‘quench
cracking”’ [1] takes place. Temperature of the heated surface, final surface
hardness, and the depth of hardness all in combination determine the affinity
of a part towards cracking. In turn, the size of the component and carbon
content of the material have to be compatible with the hardness and depth
to avoid cracking.

[4]. Apparently, significant

the only effect of grain
 IPs)
Section 3-30: Effect of Some Important Nonquantitative Factors

ere

Grinding
There is much evidence [17, 18] to show that grinding produces unfavorable
tensile residual stresses in the surface of steels that decreases the endurance
strength. However, other investigators [13, 19] havé found that careful
grinding in the longitudinal direction of round and flat components did not
adversely affect the endurance properties of the material (for example,
AISI 521000 steel with 20% vanadium, which was quenched and tempered
to Rockwell C 45 and 59). For cases where the grinding was not carefully
performed, the endurance limit was reduced about 25%. Shot peening and
tumbling did, nevertheless, help to bring the material back to its original
endurance strength.

Polishing
Hand polishing in a direction parallel to loading (with fine grade emery
cloth) produces a greater fatigue strength than that produced by electro-
polishing [16]. The reason for this phenomenon is suggested by Cina [19]
as being due to a slight cold-working of the surface caused by mechanical
polishing. Electrolytic polishing, however, has little affect or can even remove
the benefit of cold working. The modern technique of electrolytic grinding
causes no residual stresses while producing surface finishes of S—10 yw.

Mechanical surface treatment

Surface treatment of mechanical or machine elements can have a very
marked effect on both the endurance and static strength properties of the
material used. These effects can be either negative or positive, depending on
the particular process involved. As indicated earlier, most grinding operations
are not beneficial because they result in an unwanted residual tensile stress
in the material surface. On the other hand, careful grinding and subsequent
polishing can produce a desirable residual compressive surface stress. Flame
or induction hardening followed by quenching can produce beneficial surface
compression or adverse tensile stresses—even cracking, depending on the
rate of quench, quenching medium, and so on.
There are several cold working processes that are used to introduce residual
surface compressive stresses. These methods are shot peening, surface rolling,
air hammering, stretching and drawing. Of these methods, shot peening and
surface rolling are the more frequently used operations. All of the methods
cause the metal surface fibers to stretch past the yield point and into the
plastic region. The material layers underneath the surface, having also been
stretched by the outer fibers, remain elastic and return to their original
length when the cold working operation is completed. This gives the outer
surface fibers a compressive stress—the layers of material below the surface
(that is, the core) remain in tension.
 126
Chapter 3: Mechanical Properties of Engineering Materials

Shot peening is performed by one of two types of machines. One machine

ejects pellets at a high velocity from a centrifugal wheel and is capable of
raining a large volume of shot onto the surface being treated. For volume
production the centrifugal type is preferred. The other machine uses
compressed air to “shoot” the pellets out at a high velocity. This method is
often employed in shot peening difficult spots, such as holes, recesses, and
so on. Shot peening results in a metal surface finish having a roughness of
about 65—200 microinches (in.)
The positive effect of shot peening can be quite significant and, at times,
even remarkable. Figures 3-36, 3-37, 3-38, and 3-39 as well as Table 3-4
attest to this improvement. Figures 3-36, 3-37, and 3-39 are obvious and
need no comment. Figure 3-38 clearly demonstrates the depth of compressive
stress caused by shot of different diameters. It is desirable that the depth
of the residual stress be deep enough to offset any tendency of surface failure
due to repeated loads. Table 3-4 indicates that markedly improved fatigue
strengths are to be obtained by shot peening. The amount of improvement,
however, depends on the surface condition of the material. It appears that
the manufacturing process having the “‘roughest”’ surface finish realizes the
greatest improvement in fatigue strength.

Shot peening

Stress
psi

“20,000 100,000 1,000,000 10,000,000

Life cycles

Figure 3-36 Effect of shot peening on the endurance strength of spring

steel [Courtesy Wheelabrator-Frye, Inc.]

Blea
onpeened

Computed
psi
stress,

100,000 |
l 000 10,000,000

Cycles to failure

Figure 3-37 Effect of shot peening on carburized spur and helical

automotive gears. [Courtesy Wheelabrator-Frye, Inc.]
 As heat treated via

Residual
ksi
stress,

ae | ieee am l (<i
8 16 24

Depth below surface, in. X 103

Figure 3-38 Profile of residual stress due to shot peening.

The material is SAE 5147 alloy steel with a hardness of
Rockwell C scale of 48. [From R. L. Mattson: Proc. Int.
Conf. Fatigue Metal. Institute of Mechanical Engineering,
London, 1956.]

Improvement
fatigue
in
%
strength,

Polished : As rolled As forged

Machined

Figure 3-39 Effect of shot peening of steel parts having different

surface finishes. [From C. Lipson: More realistic measure of shot-
peening effectiveness. Steel: (Aug. 6, 1951).]

127
 128
Chapter 3: Mechanical Properties of Engineering Materials

Table 3-4 Effect of Shot Peening on Endurance Limit with Different

Metals having Prior Surface Treatment

Surface
Treatment Prior Prior to Increase in
Mateiral Specimen to Shot Peening Shot Peening Endurance Limit, %

SAE 1020 steel Plate As rolled Polished 9

SAE 1045 steel Plate Normalized Polished 1]
SAE 1050 steel Plate As rolled Polished 22
SAE S4340 steel Standard Quenched and Polished 18
drawn
Ni-Cr-Mo steel Standard Carburized Polished 4
Alloy steel Bar Hardened Polished 2
Alloy steel Bar Hardened Machined 23
SAE 1020 steel _— Plate As rolled Hot rolled 34
SAE 1045 steel Plate Induction Hot rolled 50
hardened :

Rail steel Rail As rolled Hot rolled 32

0.65 steel Wire As drawn Hot rolled 42
SAE 1095 steel Wire As drawn Hot rolled 50
Standard Carburized Hourolted 3
Ni-Cr-Mo steel
NE 9470 steel Standard Carburized Hot rolled 50
NE 9240 steel Standard Carburized Hot rolled 53
NE 8650 steel Axle Quenched and As forged 100
NE 8650 steel tempered
Rea Normalized and As forged 54
NE 8650 steel tempered
Biaabar Quenched and Severely 90
4340 steel tempered ground
Shaft Quenched and Chrome plating* 90
tempered
Phosphor bronze Coil spring 40
Beryllium copper Coil spring 80
S-816 (Co-Cr-Ni Coil spring 80
base)
18-8 stainless Coil spring 10
13-2 stainless Coil spring 50
SOURCE: C. Lipson, G. C. Noll, L. S. Clock: Stress and Strength of Manufactured Parts. McGraw-Hill Book Co., Inc.,
New York, 1950.

*Shot peening performed before chrome plating.

Surface rolling? improves the fatigue properties, as does shot peening,

by causing the metal surface to sustain a residual surface compressive stress.
However, with surface rolling, it is easier to obtain greater depths of com-
pressive stress. These depths can be calculated (24, 25, 26], giving this method
a more rational basis than shot peening.
The method uses sets of hardened rollers that press against the surface
to be worked with a high pressure. The procedure is not as universal in
application as shot peening and is reserved mainly for cold working round
surfaces at significant locations of stress concentration such as fillets,

° Do not confuse this term with cold rolled steel, which is a method of manufacturing a standard
steel warehousing item. The item is continuously reduced in size by means of a rolling mill.
 129
Section 3-30: Effect of Some Important Nonquantitative Factors

Table 3-5 Surface Rolling Effects on the Endurance Limits of Specific

Materials

: Surface 3
Material Specimen Se Prior to Cold ; pe ee
to Cold Working ; Endurance Limit, %
Working

Cold rolling

SAE 1045 Bar Normalized Polished 6

SAE 1045 Bored Quenched and Polished $2
tempered
SAE 1045 Bored Quenched and Polished 33
tempered
SAE 1045 Bar Normalized Machined 27
SAE 1045 Notched Machined 120
SAE 1045 Notched Machined 52
SAE 1050 Press fit Normalized Machined 150
3.1 Ni Press fit Normalized Machined 33
Of35.€ Thread Quenched and Machined 33
tempered
0.20 C Bars Hot rolled Hot rolled 67
Alloy steel Shaft with Normalized and Polished 68
fillet tempered
Alloy steel Shaft with Normalized and Polished $6
fillet tempered
Alloy steel Shaft with Quenched and Polished 30
fillet tempered
a SS
SOURCE: C, Lipson, G, C, Noll, L. S, Clock: Stress and Strength of Manufactured Parts. McGraw-Hill Book Co., Inc., New York, 1950,

grooves, dianzeters subject to interference fits, and so on. Surface rolling is

also used in thread rolling, on splined surfaces, and on welded joints called
planishing. Table 3-5 indicates that much benefit is to be gained in improving
endurance limits by surface rolling.

Carburizing
Carburizing (see Section 2.9) is a surface hardening process primarily
intended to resist wear or abrasion and is produced on a low carbon steel.
Tests, however, have shown that fatigue properties also benefit greatly from
carburizing. For example, in tests [27] by the rotating beam method, a
;5-in. AISI 2317 normalized steel has an endurance limit of 48,000 psi. This
same steel was then carburized (0.05-in. case), water quenched, and tempered.
The rotating beam test for the treated specimen showed an endurance limit
of 120,000 psi—a clear increase of 150%. The same test for a AISI 2513
normalized, resulted in an endurance strength of 54,000 psi before carburizing
and an endurance strength of 123,000 psi after carburizing—an increase of
12377.
Other tests [28] on a 0.2% carbon steel, 0.3-in. diameter bar resulted in
an endurance limit of 33,000 psi. After carburizing with a 0.03-in. thick
case, the endurance limit rose to 45,000 psi. This test produced an increase
 130
Chapter 3: Mechanical Properties of Engineering Materials

of only 36.4°%, but nevertheless an increase. The same steel with a radial
hole had an endurance limit of 48,000 psi when tested before carburizing
and 62,300 psi after carburizing. This shows how carburization can over-
come the drawbacks of stress raisers.!° The same material when tested [27]
in reversed torsional fatigue showed an improvement of 159% (17,000 psi
to 41,000 psi) in its endurance limit.
Carburized and hardened gears are more desirable than through-hardened
gears when fatigue is a serious factor. Tests [29] have shown that carburized
and hardened gears improved fatigue strengths from 20 to 90% over that
of oil- or air-hardened gears.
A summary compiled by E. R. Gadd [30] indicates that carburizing and
hardening produce improvements of 32-105 % for unnotched specimens and
82-230 % for notched specimens. In particular, the advantage of carburizing
is greatly enhanced when the depth of case (that is, the thickness of the
hardened surface) is deeper than some surface defect or stress raiser. This is
due to the fact that the case is in compression.
Carburizing produces an outer case on the material with a hardness
range of 58-63 Rockwell C obtained by tempering at 300-400°F after
hardening. The thickness of the case and its hardness are of primary concern
in design. Both of these parameters are a function of time and temperature
[11]. Case thicknesses vary from a few thousands of an inch to as much as
4 in. or more. Commercial practice specifies a “‘light case’ as being from
0.02 up to 0.04-in. thick, a ““medium heavy case”’ as being from 0.04 up to
0.06 in. thick, and a “heavy case’’ as being over 0.06-in. thick. Figure 3-40
shows a representative hardness versus depth profile for a carburized
specimen.
The problem facing the designer in specifying the case depth and hardness
is whether the case hardened layer or the core will be most likely to fail
when subjected to a specified loading. Lipson and Juvinall [10] present
an analytical model that can help the designer resolve this dilemma. Although,
as shown in Figure 3-40, the hardness varies inversely with the depth of
case, the model assumed by Lipson and Juvinall is one in which there are’
two distinct hardness levels. The first is the outer layer, which is uniformly
hard throughout its thickness. The second is the core, which is also uniformly
through-hardened but is not as hard as the outer case. For a detailed
explanation of this concept, the reader should refer to reference [10].

Nitriding
Nitriding (see Section 2.9) produces somewhat similar results to those of
carburizing but does not produce the distortion caused by severe quenching
of carburization because no heat treatment is required after the steel is
heated to just below the critical temperature. However, nitriding does
induce higher residual stresses along with a sharp increase of strength and

'° Any machine operation (e.g., drilling, grooving, etc.) must be performed before carburizing.
 Jieogl|

3
Section 3-30: Effect of Some Important Nonquantitative Factors .

70

60
Effective case depth

50 -—— —

C
Rockwell
Hardness,

30
cae case depth ee ee
a
Surface Depth below surface

Figure 3-40 A representative hardness profile of a carburized

part. Total case depth means the total depth of carbon
penetration. Effective case depth is that depth below the surface
of the specimen at which a hardness of 50 Rockwell C is
exceeded. The case depth and effective depth are design
parameters and should be carefully considered and specified.
[From Oscar J. Horger (ed.): ASME Handbook, Metals
Engineering Design, 2nd ed. McGraw-Hill Book Co., New York,
1965.]

is more expensive than carburizing. Not all steels can be nitrided, and the
process is reserved for certain alloy steels.
The needed depth of case depends on the wear that the part will experience,
the distortion of the part due to machining, the size and shape of the part,
and the endurance limits required for both the core and the case. Case
depth varies with the time of nitriding and is carried out between 925 and
1000°F, depending on the material. Case thicknesses can be obtained from
as little as a few thousandths of an inch to as much as 0.030—0.040 in.
Attaining the latter depth takes about 60—-100hr, although nitriding is
primarily a light case process.
Certain nitriding steels have been developed that produce excellent case
and core properties. These steels are Nitralloy'’ 135 (modified), Nitralloy 135,
Nitralloy N, and Nitralloy Ez. Other steels that can be nitrided are certain
grades of stainless steel, alloy steels such as AISI 4140 and 4350, and various
high speed tool steels.

'l A product of the Nitralloy Corporation.

 iez,
Chapter 3: Mechanical Properties of Engineering Materials

Among the many advantages to be obtained by nitriding, the one of

primary interest to us is its affect on the fatigue strength of steels. Typical
results [1] showing improved fatigue limits are listed as follows:
en

Endurance Limit Endurance Limit

Specimen Condition Before Nitriding, psi After Nitriding, psi

Without a notch 45,000 40,000

With a semicircular notch 25,000 87,000

With a V notch 24,000 80,000

Other surface hardening methods are cyaniding, carbonitriding, induction

and flame hardening (the latter method discussed earlier under heat treat-
ment).

Corrosion
Corrosion of a metal surface has a very damaging effect on the static
strength properties of metals because, ultimately, it causes a reduced cross-
section and can lead to eventual failure. The effect of both corrosion and
stress on strength characteristics is called stress corrosion. When the load
is variable, the combination of corrosion and varying stress is called corrosion
fatigue. It is obvious, in the light of what we have already said about fatigue,
that the probability for fatigue failure is very much increased by the rough
surface and pit marks caused by corrosion.
Corrosion can be controlled in several ways: (1) selecting the right
material for the associated environment, such as stainless steel, copper alloys,
titanium, and so on; (2) using nonmetallic coating, such as organic film,
plastic film paints, and lacquers; (3) ceramic coating, such as porcelain;
(4) electrochemical surface treatment, such as anodic coating for aluminum
or magnesium; (5) sacrificial metallic coatings, such as zinc; (6) hot dipping
of the base metal in zinc, lead, or tin; (7) cathodic protection : and (8) electro-
deposition!* of chromium, nickel, cadmium, copper, zinc, tin, or lead on the
base metal.
Plating causes a reduction in fatigue strength. It is generally postulated
that the primary cause is residual tensile stresses in the deposited material.
Baking at specified temperatures (for example, 600—-800°F) has improved
fatigue strengths of some plates. Shot peening and then plating seems to
produce a much greater improvement. However, shot peening, plating, and
then baking can bring the fatigue limit to a point that is even higher than
that of the base metal [11].

'? Electrodeposited metals are also used to enhance appearance, to reduce friction, to provide
a wear resistant surface, and to rebuild parts that have been worn.
 133
Section 3-30: Effect of Some Important Nonquantitative Factors

Temperature
Large temperature deviations have a marked affect on the fatigue strength
of metals and must be of concern when designing power plants (fossil and
nuclear), oil refineries, chemical processing plants, supersonic aircraft, and
so on. As can be observed from Figure 3-41, both carbon and alloy steels
exhibit a higher fatigue limit at lower temperatures. However, the carbon
steels first show an improved fatigue limit with increasing temperature then
“drop off’ as the temperature continues to increase. The alloy steels show
no comparable improvement ;they immediately display lower fatigue values
as the temperature is increased.

Tensile strength
70,000 psi at RT
| aes Je

Fatigue
ksi
limit, Tensile strength
120,000 psi at RT
es eee eee sls
100

Tensile strength
160,000 psi at RT

—200 +200 600 1000

Testing temperature, F

Figure 3-41 Effect of temperature on the fatigue limit of

1035, 1060 carbon steel, and 4340 alloy steel. [From Metals
Handbook, 8th ed., Vol. I. The American Society for Metals,
Metals Park, Ohio.]
 134
Chapter 3: Mechanical Properties of Engineering Materials

If a mechanical component is subjected to both a steady and a varying

load at a relatively high temperature, the likely mode of failure becomes
time dependent (see Section 3-31). Thus, at high temperature and for long
periods of time, it is the creep-rupture strength of the material that will
determine failure, and variable loading will have little effect. One might
then conclude that failure (at elevated temperatures) due to fatigue will
most likely occur at an early time in the life of the material.

SECTION 3-31
Creep
Materials under a steady load at high temperatures for a long time period
begin to deform plastically. This deformation is called creep and is time
dependent because deformation increases with time until rupture takes place.
As was stated in the previous section, fatigue failure would most likely take
place early in the life of a material. However, because steam or gas turbines,
power plant piping, chemical process and refinery piping, food processing
equipment, and so on, are designed to operate at elevated temperatures for
long periods of time, there is great interest in the phenomenon of creep.
This interest is not confined to possible failure by rupture but includes
failure by large deformations that can make equipment inoperative.
Unfortunately, there is no relation between creep and the mechanical
properties of a material at normal temperatures but there are many inter-
related factors concerning creep, including metallurgical changes. Conse-
quently, this complex phenomenon has not been well-examined experi-
mentally or thoroughly explained by theory. Thus, if a designer is faced
with decisions concerning creep, his most reliable source of information is
a test program under simulated or actual conditions. But, the expected
operating life of most equipment designed to withstand creep is usually
10-20 years. It is apparent that actual long time testing is not likely to be
undertaken, although some sparse test data are available from suppliers of
alloy steels and nonferrous metals.
So-called long-time tests are undertaken for 1000 hr. This is the
recommended time specified in the ASTM Standards, Part I, Metals, 1958.
The tests are performed under carefully controlled stress (load), temperature,
time, and creep (elongation) conditions. To conserve time, tests for different
constant loads are performed simultaneously on different specimens of the
same material. The usual procedure is to plot the creep versus time curve,
but other combinations are possible. A theoretically shaped creep curve is
shown in Figure 3-42. The initial strain takes place almost immediately
and consists ofthe elastic strain plus plastic strain if the deformation extends
beyond the yield point. The first stage shows a decreasing rate of elongation
due to strain hardening. The second stage begins at a minimum strain rate
and remains constant because of the balancing effects of strain hardening
and annealing. The third stage shows a rapid increase of the creep rate
accompanied by severe necking and ultimate rupture.
 Se)
Section 3-32: Other Temperature Sensitive Properties of Materials

First | | Third |
stage Second stage stage
|<. Fracture
torinary (secondary creep) (tertiary
creep) | creep)

Total
in./in.
%creep
or
-

Initial strain

Time-hr.
Figure 3-42 Ideal long time creep versus time curve.

When there are large deformations, the cross-sectional area change is

large enough to alter the concept of fixed stress at a fixed load. When
warranted, some adjustments can be made to maintain a constant stress.
However, most interest usually concerns the second stage because no
designer is going to permit the material to enter the third stage. Thus the
creep versus time data of 1000-hr test is plotted, and the second stage is
extrapolated to the number of hours of equipment life. This is done for each
of the creep curves plotted. In making this extrapolation, it is assumed that
the 1000 hr test allowed the material to enter the second stage, and the
material behaved in a manner similar to Figure 3-42.
The problem may also be treated algebraically if the second stage is
extended to cut the ordinate at &9 (see Figure 3-42). Thus ¢, can be taken as
the initial creep and the equation of a straight line would then represent the
first and second stages of creep. The slope of the line, de/dt, is obtained from
testing at a particular stress. Other creep curves for the same material at
different stresses are then known by merely calculating a new slope from
an empirical relationship. Those who are interested in pursuing this subject
further are directed to references [1] and [9] which, in turn, contain additional
references.

SECTION 3-32

Other Temperature Sensitive Properties

of Materials
Modern civilization is widening the scope and range of materials being
used in extreme temperature environments. At the low temperature end of
the spectrum, there is the military with demands for equipment that operates
at --60°F and lower. The petroleum industry requires —150°F for its
dewaxing process. Food packaging and freezing require some temperatures
as low as — 100°F. Piping or shipping oil from Alaska will require materials to
withstand the same low temperatures as the military. The field of cryogenic
studies has grown vastly in recent years.
 136
Chapter 3: Mechanical Properties of Engineering Materials

At the high end of the spectrum, there is the ever rising demand of the
power generating field for materials to operate in a temperature environment
of 1000°F and above. Similarly, industries such as chemical processing,
petroleum refining, high speed aircraft, and so on are demanding materials
to withstand the severity of high temperature operation.

Low temperature effects

tensi and
for metals such as aluminum and aluminum alloys, nickel alloys, austenitic
steels, lead, and copper. Carbon and low alloy steels tend to become embrittled
at much higher temperatures than the aforementioned metals. Embrittlement
is measured by loss of toughness over a small temperature range (for
example, see Section 3.21) when tested by the Charpy or Izod machines.
The transition temperature is taken to be that for which the impact energy is
reduced by 50% of its ductile value. Figure 3-43 shows some average value
curves of toughness (energy in foot-pounds) versus temperature for a variety

To 120 ft-lb
100

9% Ni, V

AISI 304 L,K 40,

80 470,
4340001)

60 0.2% C Steel, K
Hastelloy B
9% Ni, K

Energy,
ft-lb AIS! 347, K
40
y 13% Ni

7, Nodular iron

20

Gray iron

300 200 100 0 100 200 300

Temperature, °F

Figure 3-43 Toughness versus temperature for several metals.

Note the sharp drop in toughness that takes place within a narrow
temperature range. [From V. M. Faires: Design of Machine Elements,
4th ed. The Macmillan Company, New York, 1965.]
 137
Section 3-32: Other Temperature Sensitive Properties of Materials

of metals and alloys. It is apparent from this figure that the range of tempera-
ture that causes embrittlement is relatively narrow.

High temperature effects

For short time static testing of metals temperatures,
at elevated itisgenerally

ing temperature whereas

with
temperature. Figures 3-44 and 3-45 show these variations graphically.
100

Tensile strength

s jas

ksi
yield
and
Ultimate
stress,
(ol Ce ~
alles
Za
%
in
reduction
and
Elongation
area,
iS
| i P
Elongation iz Yield point

0 200 400 600 800 1000

Temperature, F

Figure 3-44 Effect of temperature on the strength and

ductility of a mild carbon steel, ASTM A 212B.

45 Ni, 15 Cr,
4 Mo, 4W,
30 3Ti,1Al

Figure 3-45 Change of elastic modulus

25 with temperature. These steels are for high
temperature service. [From V. M. Faires:
17-7 Mo
Design of Machine Elements, 4th ed. The
PH stainless
Macmillan Company, New York, 1965.]
(15 Cr, 7 Ni, 2.25 Mo,

ee
1.15 Al)
20

10~§
psi
XEof
Elasticity,
Modulus

High-alloy steel (5 Cr-Mo-V)

0 400 800 1200

Temperature, °F
 138
Chapter 3: Mechanical Properties of Engineering Materials

The variations in directions of the curves in Figure 3-45 are caused by

metallurgical changes that take place as the temperature increases. This is
particularly true for steel.
The problem of designing for extreme temperatures is a special one in
which information concerning material properties is not overly abundant.
It would be wise for a designer to work closely with the supplier and test
department before making a final materials decision in this area.

SECTION 3-33
Wear
Wear can be defined as the surface damage caused by the environment
and the effect of one or more surfaces moving past each other while in
contact.
Lipson [31] classifies wear into the following catagories:

Galling (also called scuffing, scoring, and seizing)

This type of wear is considered to be caused by the adhesion’? between
two contacting asperities that are welded together and then sheared by the
relative motion between the sliding parts. A “strong” weld wil! transfer
material from one surface to the other. A ‘“‘weak”’ weld will produce a loose
particle. Severe galling will ultimately cause surface failure and seizure of
the moving parts. Careful selection of material with different hardnesses and
proper lubrication helps to prevent galling.

Abrasion
The loose particles (as indicated by the previously mentioned adhesion
theory), asperities of the harder surface, and any hard foreign matter finding
their way in between the rubbing surfaces can cause abrasive wear. These
particles embed themselves in the softer surface and tend to scratch and
score the harder surface. Abrasion, although not eliminated, can be controlled
by heat treatment, plating, anodizing, spraying, hard facing, or surface
hardening.

Pitting (also known as pitting corrosion or spalling)

This kind of wear is caused by repetitive contact stress and is a form of
surface fatigue.

Frettage (also known as fretting corrosion or false

brinelling)
Frettage is a surface damage caused by small movements between mating
surfaces, as in press fits. The constant oscillation removes any oxide film
that may form, which finally acts as an abrasive and causes more serious
'* This theory is also the basis for fretting corrosion as explained in Section 3.30.
 159
Section 3-34: Radiation Effects

damage. This form of surface failure and its control was discussed earlier
in Section 3.30.

Cavitation erosion
Cavitation erosion is caused by a high relative motion between a fluid
and a mechanical surface. If the fluid velocity is high enough, the local
static pressure at some point along the surface will drop below the liquid’s
vapor pressure. Local boiling begins, and small vapor bubbles are formed.
As equilibrium is restored, the static pressure rises and the bubbles collapse,
producing liquid particles with a high velocity. These high energy particles
impinge on the metal surface, causing cavitation pits. Repeated action of
this phenomenon will ultimately cause fatigue failure of the metal.
Propeller, turbine, and pump blades are usually subject to this kind of
damage and show pit marks at the blade edges. The most immediate and
straightforward way to fight cavitation is to use hardened materials or a
plate of chromium or chrome-nickel compounds. Other cures are to reduce
the vapor pressure with additives, reducing turbulence, changing the liquid
temperature, or adding air to the system to act as a cushion for the collapsing
bubbles.

Galvanic corrosion
Galvanic corrosion is a form of electrochemical attack where two dissimilar
metals, connected electrically, are within an electroyte. This model is called
a galvanic cell, and damage is caused by the dissolution of the anode (that
is, the metal that is higher in the electromotive series). The rate of attack
depends on the metals involved, the area of each metal, the type ofelectrolyte,
the current density, polarization of the cathode, temperature, and so on.
Another form of galvanic corrosion takes place when two like metals (or
two different points on the same metal part) are surrounded by an electrolyte
with different concentration pockets or different concentrations of dissolved
oxygen. These conditions form what is commonly known as a concentration
cell. Specifically, the cell formed by different electrolyte concentrations is
called a metal-ion concentration cell and that formed by the dissolved oxygen
is called an oxygen concentration cell.

SECTION 3-34
Radiation Effects
With the advent of nuclear reactors and nuclear handling equipment, the
designer should be aware of the problems and/or changes generated by
exposure of various materials to strong radiation fields. Fortunately, the
metals generally used for fabricating components are little affected by
exposure. In fact, in some cases, exposure may bring about some beneficial
effects. However, organic compounds can be severely damaged and even
destroyed by a little exposure to radiation.
 140
Chapter 3: Mechanical Properties of Engineering Materials

Properties such as the elastic moduli, electrical resistivity, thermal

conductance, and density are insensitive to radiation exposure. However
some alloys have experienced a threefold increase in yield strength. This is
accompanied by a smaller increase in tensile strength and a sharply decreased
ductility. It has also been observed that the rate at which work hardening
takes place is less for exposed materials.
The sharpest change takes place in the notch-impact transition tempera-
ture. This means that exposure to strong radiation (that is, neutron bombard-
ment) will make most steels and steel alloys subject to ductile fracture due
to their reduced ability to absorb energy. Figure 3-46 clearly shows this

70

A O unirradiated

60 B A 1X 1018 neutrons/em?
C O 5X 1018 neutrons/em2
D @ 7X 1018 neutrons/em2
50 E & 1.2X 1019 neutrons/em2

40

ee Oo D
30 e © ®
®

tens
A .A— «a
20
ft-lb
fracture
Charpy-V
energy,

IN YE

0 | allZi l
Sup
—100 —50 0 50 100 1 50 200 250 300

Test temperature, °C

Figure 3-46 Charpy V-notch impact energy curves for ASTM

A 212 grade B steel irradiated to various fast neutron (>1 mev)
doses at 175°F. [From Oscar J. Horger (ed.): ASME Handbook,
Metals Engineering Design, 2nd ed. McGraw-Hill Book Company,
New York, 1965.]

reduction in impact energy. However, the austenitic stainless steels do not

seem to be at all affected by radiation with respect to their notch sensitivity.
This is fortunate as most of the equipment (for example, pumps, vessels,
condensers, and so on) and piping used in nuclear energy and testing are
made of stainless steel.

PROBLEMS
1. In designing a part, the engineer will necessarily be referring to material property
tables similar to those of Appendix A. What should he keep in mind in using the
data from such tables?
 14]
Problems

. When is the stress of a structural member no longer linearly proportional to its

strain?

. What is meant by true stress? Engineering (or nominal) stress?

4. A camshaft on a knitting machine is made from AISI 1040 carburized steel and
is found to have too great a deflection when in operation. The maintenance
engineer has suggested replacing the shaft with one made of heat treated alloy
steel (for example, AISI 4340) in order to reduce its deflection. Will his suggestion
solve the problem? State the reason for your answer.
. A particular type of alloy steel is to be used for a shaft. Its modulus of elasticity
in tension is given as 29 x 10° psi, and the Poisson ratio is known to be 0.27.
Since the shaft will be subjected to torsion, you will also need to know its shear
modulus. Can you obtain this value with the given information? If so, what is it?
. A bar 0.75 in. in diameter by 10 ft long is subjected to a tensile load of 26,500 lb.
Measurements under load indicate that the bar increased 0.24 in. in length and
decreased in diameter by 0.0005 in. Find (a) modulus of elasticity in tension;
(b) Poisson’s ratio; (ec) shear modulus; (d) the general class of materials in which
this bar can be found.
. In tensile testing a 0.875-in. diameter plain carbon steel bar, the following data
were obtained: at 20,000 lb, the test specimen elongated 0.0087 in. in 8 in.; at
36,000 Ib, it elongated 0.022 in. in 8 in.; yielding was first observed at 33,000 lb,
and fracture occurred at 60,000 Ib. Find (a) the yield point strength ; (b) the ultimate
strength; (c) the elastic modulus in tension; (d) the AISI number identifying this
steel.
. A Zin. thick plate is being fed to a 150-ton mechanical press. How many holes of
lin. diameter can be punched if the material is AISI 1020 steel, cold drawn?
How many holes if the material is AISI 1020, annealed? How many holes if the
material is AISI 3140, annealed?
. A round bar of AISI 1020, annealed steel is 30 in. long. It is subjected to a tensile
load of 25,000 Ib. If the factor of safety is to be no greater than 3, find the diameter
of the bar for the limiting yield stress.
10. Find the maximum strain energy at impact of the 50 lb weight for (a) a bar made
of steel, (b) brass, (c) aluminum. Assume that 80% of the energy ofthe falling weight
is elastically absorbed.

Problem Figure 10

11. For each of the following materials, calculate the modulus of resilience: (a) AISI
1020 annealed steel; (b) AISI 4340 normalized steel; (c) AISI Type 302 stainless
steel, cold worked; (d) 2024-T4 aluminum; (e) manganese bronze A, half-hard;
(f) ASTM Type ZK60A-TS magnesium alloy ; (g) inconel X-750.
 142
Chapter 3: Mechanical Properties of Engineering Materials

12. Find the strain energy relation for the two systems shown. The bars are round.

po (2/5)L Ls

Problem Figure 12 é

13. After the weight W strikes the ‘‘stop” on the bottom of the bar, the tensile stress
should not be greater than 20,000 psi. Assuming that all of the energy of the falling
weight is elastically absorbed, determine (a) the weight W and (b) the weight W
if all of the bar were | in. in diameter. The bar is made of steel.

il :
15 in. dia.

Problem Figure 13

1 in. dia.

14. Discuss the relationship between brittleness and ductility.

15. What circumstances would make toughness an important property?
16. In selecting a material for the base of a machine which is periodically transmitting
dynamic loads, what important property would a designer wish to have? Name
such a material.
17. Using Table A-2 of Appendix A, calculate the modulus of toughness of any six
different materials. For each of the six materials chosen, calculate the modulus for
three conditions of the same material. (You will have 18 separate calculations.)
18. (a) Calculate the ultimate strength of the following AISI steels by means of
equation (3-12).
1. 1040 HR 4. 4140 CDA 7. 6150N 10. 9255 HRA
2. 1040 CD 5. 4140N 8. 8640 CD 11. 9440N
3. 1095N 6. 4640 CDA 9. 8740 N
(b) Calculate the per cent difference between the results of part (a) and the values
found in Table A-2 of Appendix A.
(c) Make the same comparison as in part (b), but this time compare your calculated
results with the values obtained from Figure 3-17.
19. Describe the mechanism of fatigue failure.
 143
Problems

20. (a) For the steels listed in Problem 18a, calculate the endurance limits S’.
(b) Using equation (3-14), compare (that is, calculate per cent difference) the
endurance limit of class 20, 30, and 50 cast irons with values found in Table A-1
of Appendix A.
(c) Repeat part (b) for class 65000, 105000 and 200000 cast steels found in Table A-1
Appendix A.
21. List six conditions that tend to make a metal more prone to fatigue failure.
22. A flat plate 4 in. thick and 3 in. wide contains two 4-in. semicircular notches that
are cut out opposite each other. What is the theoretical stress concentration factor
K, if the plate is loaded in (a) tension, and (b) bending in the plane of the plate?
23. A drive shaft at some point along its length has a change of diameter from 2 in.
to 14 in. with an 4-in. fillet radius at the discontinuity. Determine the theoretical
stress concentration factor K, if the shaft is subjected to (a) bending, (b) tension,
(c) torsion.
24. In Problem 23, what would one use as the theoretical stress concentration factor
if a 3-in. diameter oil hole were drilled radially to the 14-in. diameter part of the
shaft and located very close to the }-in. fillet? Determine proper values of K, for
(a) bending, (b) tension, (c) torsion.

25. The figure shows a class 25 cast iron bar subjected to a steady tensile load. What
is its factor of safety? What is the factor of safety if the steady load is compressive
and of the same magnitude?

1 in. dia. hole

1 in. thick

5000 |b 5000 |b Problem Figure 25

26. In Problem 25, what would the factor of safety be if the center ofthe 1-in. diameter
hole were located ?in. from the longer edge of the plate? (Consider the tensile
case only.)
27. Repeat Problem 25 for the case where the hole is (a) 3-in. diameter, (b) 3-in.
diameter.

28. The socket wrench shown in the figure is made of class 50 cast iron. What is the
maximum torque that can be applied if the factor of safety is not to exceed 2?
(Consider only the change of section at the fillet radius.)

Square head
for handle

Problem Figure 28
 144
Chapter 3: Mechanical Properties of Engineering Materials

29.*This figure shows a cantilever beam machined from AISI 1040, annealed steel.
The load F is unsteady and it is necessary therefore to base the design of this beam
on its endurance strength (for a finite life) or its endurance limit (for an infinite life).
For a 95% survival rate, establish the value to be used for S,. In solving this
problem use equation (3-13) and compare your final answer with the one that
would have been obtained by using (a) equations (3-12) and (3-13), and (b) by only
using Figure 3-28.

a dia. hole

dia.

Problem Figure 29

30.*Work Problem 29 for the case where the drilled hole is located at the 3-in radius
fillet.
31.*Work Problem 29 for the case of an unsteady torsional load.
32.*The shaft is made of AISI 4340 HRA steel, and is machined to the dimensions
shown. It is then quenched in oil and tempered at 1475°F. If an unsteady bending
moment is applied, determine a working endurance limit S, for each stress con-
centration.

Profiled kayway 1"

vO ele ee

plas
Gj SAREE %
Vr,

Problem Figure 32

* These problems are presented here only as a means of familiarizing the student with the
procedure followed in obtaining a working endurance limit S,. The application of this term
to a complete problem is carried out in Chapter 7 only.
 145
References

33.*Repeat Problem 32 for an unsteady torsional load.

34.*Repeat Problem 32 for an unsteady axial load. Assume there is no keyway for
this problem.

35.*(a) In Problem 32, what would be the expected improvement in the endurance
limit at both the groove and fillet if each were formed by surface rolling?
What is the reason for this improvement?
(b) Would the same improvement be achieved if the shaft were shot peened?
(c) What other ways can you think of that would improve the endurance limit of
the shaft?

REFERENCES
[1] Metals Handbook, Vol. 1, 8th ed. The American Society for Metals, Metals Park,
Ohio, 1961.
[2] Technical Editor Speaks. The International Nickel Co., Inc., New York, 1943.
[3] J. H. Frye: Selection of bar steel used as rolled or drawn. Metal Progr., 47: 497-504
(March 1945).
[4] Batelle Memorial Institute: Prevention of the Failure of Metals Under Repeated
Stress. John Wiley & Sons, Inc., New York, March 1949.
[5] S. Timoshenko and G. H. MacCullough: Elements of Strength of Materials, 2nd ed.
D. Van Nostrand Co., Inc., New York, 1940, p. 318.
[6] V. M. Faires: Design of Machine Elements, 4th ed. The Macmillan Co., New York,
1965, p. 102.
[7] Lionel S. Marks (ed.): Mechanical Engineers’ Handbook, Sth ed. McGraw-Hill
Book Co., Inc., New York, 1951.
[8] F. B. Stulen, H. N. Cummings, and W. C. Schulte: Preventing fatigue failures,
Part 5, Mach. Des., 33, (June 22, 1961).
(9] Joseph Marin: Mechanical Behavior of Engineering Material. Prentice-Hall, Inc.,
Englewood Cliffs, N.J., 1962.
[10] C. Lipson and R. C. Juvinall: Handbook of Stress and Strength. The Macmillan
Co., New York, 1963.
[11] Oscar J. Horger (ed.): ASME Handbook, Metals Engineering Design, 2nd ed.
McGraw-Hill Book Co., New York, 1965.
[12] G.C. Nolland C. Lipson: Allowable working stresses. Proc. Soc. Exp. Stress Anal.,
3(2): 89-101 (1946).
[13] C. H. Jennings: Welding design. Trans. ASME, 58: 497-509 (1936).
[14] R. E. Peterson: Relation between stress analysis and fatigue of metals. Proc. SESA,
11(2): 199-206 (1954).
[15] T. E. Eagan: Cast ferrous metals, Part 3, Stee/, 135(22), (1954); 135(29), 80-82
(1954).
[16] R. B. Heywood: Designing Against the Fatigue of Metals. Reinhold Publishing
Corp., New York, 1962, p. 158.
[17] H. E. Boyer: Effects of grinding on physical properties of hardened steel parts.
Trans. ASME, 40: 491-503.
[18] J. Frisch and E. G. Thomsen: Residual grinding stresses in mild steel. Trans.
ASME, 73: 337-342 (April 1951).
 146
Chapter 3: Mechanical Properties of Engineering Materials

[19] B. Cina: Effect of surface finish on fatigue. Metallurgia, 55: 11-19 (Jan. 1957).
[20] Shot Peening. Wheelabrator-Frye Inc., Mishawaks, Ind., 1962.
(21) R. L. Mattson: Proceedings, International Conference on Fatigue of Metals. The
Institute of Mechanical Engineering, 1956, London.
[22] C. Lipson: More realistic measure of shot-peening effectiveness. Stee/, (Aug. 6,
1951).
C. Lipson, C. G. Noll, and L. S. Clock: Stress and Strength of Manufactured Parts.
McGraw-Hill Book Co., Inc., New York, 1950.
H.R. Thomas and V. A. Hoersch: Stresses due to the pressure of one elastic solid
upon another. University of Illinois Engineering Experimental Station Bull. No. 212.
S. Way: J. Appl. Mech., 57: A69-A71 (June 1935). Discussion with reference to
articles by O. J. Horger entitled ‘Fatigue strength of members as influenced by
surface conditions” which appeared in Prod. Eng., Nov., Dec., 1940, and Jan.,
1941]. :
[26] R. E. Peterson and A. M. Wahl: J. Appl. Mech., 58: A74-A75 (June 1936).
Discussion of a report by H. F. Moore entitled ““A study of fatigue cracks in car
axles,” University of Illinois Engineering Experimental Station Bull. No. 165, June
14, 1927.
J.M. Lessells: Strength and Resistance of Metals. John Wiley and Sons, New York,
1954.
] H. F. Moore and J. B. Kommers: An investigation of the fatigue of metals.
University of Illinois Engineering Experimental Station Bull. No. 124, 1921.
29] R.A.C. Fosberry and H. D. Mansion: Bending fatigue of gear teeth. Motor Industry
Research Association Report, 1950-1957.
] E. R. Gadd: Fatigue from the metallurgists viewpoint. J. Roy. Aeronaut. Soc., 57:
565 (Sept. 1953).
] Charles Lipson: Wear Corssiderations in Design. Prentice-Hall, Inc., Englewood
GliftsN eel G7e
Manufacturing Processes
and Design
SYMBOLS

rms = root mean square surface o = standard deviation of the sum of

roughness, microinches (juin.) assembled parts, +in.
T = total tolerance accumulation, in. o,, = standard deviation of the nth
T, = tolerance of the nth assembled assembled part, +in.
part, in.

The development of a device or a machine that is functional on paper is

only part of the overall task of design. If the designer’s creation is to leave
the drawing board and become a physical piece of hardware, it must be
manufacturable. Manufacturable means that all of the component parts can
be fabricated and assembled at a competitive cost. To meet this objective,
the designer must have more than just a “passing acquaintance” with manu-
facturing processes. In particular, he must be familiar with those manufac-
turing and production facilities within his own company.
The designer is not expected to be a tool maker, a machinist, and so on.
Rather, the implication here is that he must know the capabilities of basic
manufacturing and shop process equipment. Only by knowing these can he
properly design component parts, select materials, specify tolerances,
consider assembly procedures, and so on.
Different production methods, materials, tolerances, and procedures are
likely to be employed when a part is to be fabricated in millions, or in
thousands, or only as a single unit. As stated in Chapter 1, designing involves
compromise, and method of manufacturing merely adds yet another facet
to the overall decision making process concerning design.
In the present chapter, we will briefly summarize the basic methods of
manufacturing and indicate the function and limitation of each method. We
shall also review dimensioning and finally close the chapter with comments
on good design practice.

147
 148
Chapter 4: Manufacturing Processes and Design

SECTION 4-1
Introduction
Manufacturing processes can be classified’ into two general groups known
as primary and secondary processes. The primary processes (see Figure 4-1),
namely, casting, powder metallurgy, forging, cold heading, cold extruding,
stamping, deep drawing, spinning, roll forming, hot extruding, brake forming,
and so on, are those that convert raw material into shapes. These forming
methods include both hot and cold working processes and, in general, will
still require further finishing operations in order to obtain an end product.
The secondary processes are those that bring the part to the dimensions
and surface finish specified. One may also include in the secondary processes
such operations as heat treating and surface treatments for corrosion
resistance, hardness, and appearance. Secondary processes are planing,
shaping, turning, milling, drilling, boring, reaming, broaching, grinding,
honing, lapping, polishing, and special methods of metal removal such as
electrodischarge machining and electrochemical machining. Operations like
screw cutting, tapping, thread milling, gear cutting, and so on, are secondary
processes that are merely adaptations of one of the aforementioned processes.

SECTION 4-2
Primary Processes
Figure 4-1 illustrates schematically some of the primary manufacturing
processes. A complete discussion of these procedures is beyond the scope
of this text. Therefore, we will confine ourselves to succinct descriptions of
the most prominent techniques.

Castings
Castings are identified by the type of mold or the force required to fill the
mold. Molds are either permanent or expendable. The pattern of sand, shell,
and plaster molds, however, can be used repeatedly for making new molds.
Castings are made by pouring molten metal into a mold or die. As the metal
cools, it takes the shape of the cavity. Tables A-1, A-8, A-10, A-11, and A-12
in Appendix A indicate the wide variety of metals that can be used in the
cast processes. The basic casting methods are described below.
Sand Castings. 1. The green sand process is one in which moist, bonded
sand is packed around a wood or metal pattern. The pattern is then removed
and molten metal is poured into the cavity. When the metal solidifies, the
mold is broken and the casting is removed. Almost any metal can be used,

' There is no clearly defined method of classifying manufacturing processes. Upon investigation,
one will find that some references base manufacturing classifications on hot and cold working
methods, others on casting, forming and machining, and still others on the type of stress created,
etc.
 149
Section 4-2: Primary Processes

-| Number

oy anal | Punch

A P ) Die
Rolling BS Coining
Gy Coined
part

Steps in spinning Original blank

Final shape ait it :
Stipa of sheet metal
Follower held
Spinning in tailstock

Extruding TAF =
Form attached
to headstock
spindle

Shear
spinning ix

Tube
spinning

Swaging or
kneading

Casting removed
from mold
CASTING
Deep drawing
Cavity Granules

Y os

;
Nbr
Wire & tube LD 0d 008 09 é
drawing
j 2
08 Op pone ©. Od
ROD GY
i) Oho
00 2p OStd ORY

Yys KS Piston moves to Molded part

Stretching W left and applies removed from
Mold pressure mold

MOLDING

ayihiaiew:
(a) Convex

Contoured
flanging
(b) Concave

Figure 4-1 Schematic drawings

of primary processes.,
 150
Chapter 4: Manufacturing Processes and Design

with virtually no limit as to the size or shape of the part. The method permits
casting complex components at a low tooling cost and is the most direct route
from pattern to casting.
Some machining is always necessary with the green sand process, and large
castings have a rough surface finish. Close tolerances are difficult to achieve,
and long, thin projections should not be cast. It is possible, however, to
design for bosses, undercuts, and inserts. The minimum core hole diameter
advisable is :—,in., and the minimum section thicknesses advisable are
aluminum, ;;in.; copper, 3 in.; iron, #,in.; magnesium, 35 in.; and steel,
+ in.
2. The dry sand process is similar to the green sand process except that
core boxes are used in place of patterns. This method is usually limited to
smaller parts than method |. The sand is bonded with a setting binder, and
the core is then oven baked at 300—400°F to:remove the moisture.
3. The carbon dioxide sand process is one 1n which the sand molds are
bonded with sodium silicate solutions and set by forcing carbon dioxide gas
through the sand. This type of mold is strong and permits the production of
better dimensionally controlled castings than either methods | or 2.
4. The core-sand mold process is one where the molds are put together
completely from oven baked cores set with organic binders such as oil or
dextrines.
Figure 4-2 shows the construction of a typical sand mold.

Weight Risers Sprue

Chill bar

Core print

is
Parting
line y : : ON Runner

Molding board Sand

Figure 4-2 Typical sand molds for ferrous and nonferrous

metals. [From 1970 Metals Reference Issue, Machine Design,
Penton Publishing Co., Cleveland, Ohio.]

Shell Mold Casting. This is a process where the molds are made by coating
the sand with a thermosetting plastic (see Figure 4-3). The mold is then
supported on the outside by sand or shot, and molten metal is poured. When
the metal has solidified, the mold is broken away from the finished casting.
Shell molding produces castings with smooth surfaces, uniform grain
structure, high dimensional accuracy, rapid production rate and minimal
amount of finishing operations. The minimum section thickness castable is
 15d
Section 4-2: Primary Processes

Shell-mold half

wes

Pattern plate

Ejector pins

Figure 4-3 Shell mold being stripped from pattern. [From

1970 Metals Reference Issue, Machine Design, Penton Publishing
Co., Cleveland, Ohio.]

je in., but section differences where the maximum-minimum ratio is

greater than 5 to 1 should be avoided. Bosses, undercuts and inserts are
easily cast, and the minimum cored hole diameter is 4-4 in. The method is
relatively costly and is limited to few metals.
Full-Mold Casting. A process in which a green sand or cold-setting resin
bonded sand is packed around a foamed plastic pattern (for example,
polystyrene). The plastic pattern is vaporized with the pouring of the molten
metal into the mold. An improved casting surface can be obtained by putting
a refractory type of coating on the pattern surface before sand packing. The
pattern can be one piece or several pieces, depending upon the complexity of
the part to be cast.
The plastic patterns are handled easily. They do not require any draft, and
produce no flash on the casting. For small quantity production this method
can be expensive, depending upon the size and complexity of the casting.
The minimum recommended section thickness is 0.1 in., and the maximum
section thickness is unlimited. Bosses, undercuts, and inserts present no
problem in this type of casting. The minimum cored hole diameter is j in.
Permanent Mold Castings. These are formed by a mold that can be used
repeatedly. Some applications have reported as many as 25,000 castings
being made from the same mold. Usually, the mold requires some redressing
after about 3000 uses. The molds are machined of metal (for example, gray
iron) for casting nonferrous metals and cast irons. Machined graphite molds
are used to cast steel.
This method produces castings that have a good surface finish as well as
a good grain structure, low porosity, and high dimensional accuracy. The
initial mold cost is relatively expensive, but castings weighing as little as
several ounces to castings weighing about 500 Ib can be produced. Permanent
mold castings are limited to relatively simple shapes and forms. The maximum
recommended section thickness is about 2.0in. The minimum thickness,
however, depends upon the material being cast as follows: 7¢in. for iron,
 iey4
Chapter 4: Manufacturing Processes and Design

3-t in. for aluminum, 3 in. for magnesium, and 3-74 in. for copper. The
minimum castable holes are ;—; in. in diameter.
Die Casting. A process used extensively in the quantity production of
intricately shaped zinc, aluminum, lead, and magnesium alloys. The method
is limited in use with tin and copper alloys.
Castings are formed by forcing molten metal under pressures of 1500—
25,000 psi into an accurately machined steel die. The steel die, which is water
cooled, is held together by a hydraulic press until the metal casting(s)
solidifies. To ease the ejection of the cast parts, a lubricant is sprayed on the
die forming surface and the ejection pins. For small diecast components,
multiple cavity dies are used.
The surface finish of the resulting castings is quite smooth and has
excellent dimensional accuracy. Although the cost of production is relatively
cheap, the initial die cost is high. Diecasting is limited to nonferrous metals
and by the size of the part that can be cast. The maximum wall section thick-
ness is usually restricted to ;x in. but certainly to no greater than 0.50 in.
Figures 4-4a and b show the two basic types of die casting machines. :
Plaster Mold Castings. These are made by pouring a nonferrous alloy (for
example, aluminum, copper, or zinc alloy) into a plaster mold, which is then
broken to remove the solidified casting. The castings produced by this
process are smooth, have high dimensional accuracy, low porosity, and can
be madein many intricate shapes. The method, however, has disadvantages in
that it is limited to nonferrous metals, small castings, and also requires a
relatively long time to make the molds. The minimum wall section thickness
is 0.040—0.060 in. for a cast area that is less than 2 sq. in. For larger cast areas,
the minimum wall thickness increases accordingly.
The Shaw process, developed within the last decade, is a most promising
method. It employs the use of plaster molds. This procedure produces
castings with fine detail and excellent dimensional accuracy and surface
finish.
Ceramic Mold Casting. This process uses a mold made ofa ceramic powder,
binder, and gelling agent. The mold can also be made of a ceramic facing
reinforced with a sand backing. The method is restricted to casting intricate
parts requiring fine detail, close tolerances, and smooth finishes. The
minimum wall thickness recommended is 0.025—0.050 in. There is, however,
no limit to the maximum wall thickness.
Investment Castings. These are made when parts are desired that are
intricate in shape, have excellent surface finish, and require a high degree of
dimensional accuracy (for example, 0.003—0.007 in. the first dimensional
inch, and 0.002 in. for each additional inch). In addition, this method of
casting permits the use ofa variety of metal alloys (see Table A-12 in Appendix
A) and does not have any metal flashing to be removed from the finished
casting.
The technique of investment casting requires careful workmanship and
expensive patterns and molds. The minimum castable wall thickness is
 193:
Section 4-2: Primary Processes

Air valve
RAINS
SCTIRNSAY
NSSieNS

Movable platen

ee
Tie bars

|ZAZTZ Vi
MWVKLLLMLLLLL
SLL LL Le
ee eee)

Die plate
Ladle

Plunger
piston rod

NK
< 2 “Y R

YU VL LLL. an

ce Ge “gs3 4
SN VT
Or LLL pei

Eee ave 7 Nitralloy liner

Heat-treated steel cooling

Portion Portion ‘
connections
Plunger comes through this half
o, die to eject sprue

(b)

Figure 4-4 (a) The hot-chamber die casting machine where the molten metal in
the metal pot is pushed through the gooseneck and the nozzle into the forming die
by the plunger. (b) The cold-chamber of the die casting machine where a
predetermined quantity of molten metal is poured into the machine by a ladle,
after which a plunger forces the material into the die. The rate of production is
higher for the hot-chamber machine except for high melting temperature materials
(for example, brass, bronze). [Courtesy of the New Jersey Zinc Company.]
 154
Chapter 4: Manufacturing Processes and Design

Figure 4-5 Gas turbine blades and vanes made by the method of investment
casting. The materials used were cobalt and nickel base alloys for their temperature
and strength properties. [Courtesy of Howmet Corporation, Pechiney Ugine
Kuhlmann Group.]

about 0.025—0.050 in., and the maximum thickness should not exceed 3 in.
Figure 4-5 illustrates the complexity of products that may be investment cast.
Centrifugal Casting. This is a method of casting a shape by pouring the
molten metal into a rotating flask containing the mold. The molds are made
of sand, metal, or graphite (depending on the metal cast) and are rotated
about their axial centers either in a horizontal or vertical position.
Relatively large diameter and bulky components are made by centrifugal
casting. Typical products produced by this method are pipe, gun barrels,
hollow shafts, machinery drive rolls, long sleeves, tubing, and so on. Where
the diameter to length ratio is rather large, the rotational axis can be vertical.
Parts cast in this manner are ring gears, gear blanks, engine cylinder liners,
bearings, rings, and so on. Figure 4-6 shows a horizontal casting machine.
When cores are used in vertically mounted molds, the method is called
semicentrifugal casting, and it is used to make parts such as wheels, turbo-
supercharger diaphragms, disks, flywheels, and so on. Figure 4-7 shows a
schematic section of a semicentrifugal casting of a cored flywheel.
 1S)
Section 4-2: Primary Processes

Flask
Sand lining
Orifice
Top ADEN Pouring basin

Bottom rollers

Figure 4-6 True horizontal type centrifugal casting machine. The sand is packed
progressively when the flask is in the vertical position before it is placed on the
rollers. [Courtesy of American Cast Iron Pipe Company.]

Pouring basin Center core

Strainer core
Feeder reservoir

Holding
fixtures

Revolving
table

“Semicentrifugal’’ casting of fly wheels and gear blanks.

Figure 4-7 Section through the flask of a semicentrifugal mold and

casting. [Courtesy of American Cast Iron Pipe Company. ]

Generally, the method ofcentrifugal casting is expensive and is also limited

in the shapes which can be cast. However, the castings made by this method
are very sound and have good dimensional accuracy. It should also be noted
that this method is the only way to obtain a large cylindrical part.
Continuous Casting. This is a method whereby a molten metal is con-
tinuously gravity fed from a ladle into an ingot mold ofa desired shape which
is open at both ends. As the metal “‘falls” through the mold, it takes its shape,
and is rapidly cooled by a water spray. It is then cut to specific lengths.
The continuous casting process has the advantage of being low in cost and
having a high rate of production. In addition, this method permits cast-
ing materials that cannot be extruded. However, continuous casting is
restricted to shapes of uniform cross section (that is, in the direction of
casting), and cast holes must be in the direction of the casting. The minimum
 156
Chapter 4: Manufacturing Processes and Design

wall thickness recommended is ¢ in., and the maximum thickness depends

ontheshape beingcast. The minimumsize of castingis5 in.,and the maximum
size is about 9 in.

Powder metallurgy
Powder metallurgy, P/M, is an old art. Historically, it was employed as
far back as 3000 B.c. in ancient Egypt. Porous metal bearings first appeared
in quantity just after World War I and came into their own during World
War II.
The material employed in making P/M parts consists of pure metal powders
or mixtures thereof with some additive metal powder (for example 95%
copper and 5% tin, a low tin bronze). The particle size used can be from
1-100 microns (1). The method of manufacture involves placing the powder
(or mix) into a forming die for compacting. After the part is formed by the
die, it is presintered (that is, heated at a low temperature) for the purposes of
increasing the component strength, removing lubricants, and so on.
Following this step, the part can be machined and subsequently sintered
(heated to a higher temperature in a controlled atmosphere) in order to
obtain the desired mechanical and physical properties. In the manufacture
of sintered bearings (for example, Oilite), the part is impregnated with a
lubricating oil that fills the pores of the component. This operation is
performed by dipping in oil or by spraying with oil during the sintering

Figure 4-8 A wide representation of parts made by powder

metallurgy. [Courtesy of the Amplex Corporation, Division of the
Chrysler Corporation.]
 ID7,
Section 4-2: Primary Processes

process. The final component can be coined, heat treated, machined, and so
eon:
Typical parts made by P/M techniques are gears, ratchets, pawls, levers,
cams, clutch friction facings, rollers, guides, splines, bushings, turbine blades,
spacers, permanent magnets, and so on (see Figure 4-8). This method of
manufacture has the advantage of controlling the mechanical and physical
properties of a finished part at a rapid rate of production with little scrap loss.
In addition, it is possible by means of compacting and sintering to alloy
metals that cannot be alloyed by conventional methods. Finally, powder
metal parts can be made with fairly close tolerances and smooth surfaces.
Often they eliminate the need for secondary operations.
On the debit side, P/M parts are somewhat limited as to the size and shape
which can readily be fabricated. This method is uneconomical for short runs
unless the part is large and has a high unit value.

Forging
Forging dates back to ancient times and, before the advent of modern
production methods, was associated with the “village blacksmith shop.”
Forging is a production method whereby heated metal stock is shaped to a
desired form by compressive forces or by sharp hammerlike blows. Virtually
all ductile materials may be forged by first preheating the work piece (that is,
a billet, a wrought bar, a cast or sintered ingot, or a powder metal form) toa
forging temperature. However, there are some materials, such as stainless
steel, that are forged at room temperature or slightly above.

Figure 4-9 Forging ofa six cylinder crankshaft of a diesel engine. The upper
photo is that of the forged crankshaft “‘as is.”” The lower photo is an etched
longitudinal section through the crankshaft. Note how the fiberlike flow lines
follow the contour of the forging, and how they are closely spaced (that is,
densely). [Courtesy of the Wyman-Gordon Company. ]
 158
Chapter 4: Manufacturing Processes and Design

Figure 4-10 Opposite: A 50,000-ton hydraulic press.

[Courtesy of Wean B-L-H, Inc.]

Forging has a marked beneficial effect on metals Their strength and

toughness are improved because the “‘kneading”’ of the dies results in a
beneficial orientation of their grain structure. The constant hot working
performed by the forming or shaping dies causes the material to become more
dense and the grain “‘flow lines”’ to follow the contour ofthe final component
shape. An illustration of this phenomenon is shown in Figure 4-9.
There are several forms of forging, but there is some disparity identifying
processes with names in different references. However, the following classifi-
fication and description of each may be considered as basic.
Open Die Forging. Open die forgings, also known as hand or smith
forgings, are made by using hammers or presses in conjunction with black-
smith tools or flat type dies. There is little, if any, lateral confinement of the
work piece, and the desired shape is obtained by manipulating the forging in
between blows.
This procedure, which employs low cost tooling, is relatively simple, but
it has less control in determining grain flow, mechanical properties and
dimensions than do other forging methods. The technique is applicable to a
wide range of forging sizes and is restricted to short run production. Also,
the final cost of production can be higher than other forging methods because
finished machining is often required. Open die forging has some other
disadvantages, namely, relatively poor material utilization, restriction to
simple shapes, difficulty of maintaining moderately close tolerances, and
absolute need for skilled labor.
Closed Die Forgings. These are produced by hammering the work piece
into a desired shape by means of closing dies. The hammering or pressing is
performed, respectively, by a mechanical or hydraulic press. Small and
medium size forgings are generally made in presses ranging in capacity from
500—10,000 tons. Figure 4-10 is a photograph ofa 50,000-ton press used to
make large integral forgings.
Closed die forging, as compared to open die forging, makes good utilization
of the work piece material and has excellent reproducibility with good
dimensional accuracy. The resulting mechanical properties are also better
than those developed in open die forging. Furthermore, this method is
suited for rapid production rate. But, because of the initially high cost of
tooling, it should not be considered for short-run production.
Upset Forging. This is a method of forging intended for large quantity and
rapid production of symmetrical parts such as gears, gear clusters, worm
gears, shafts, axles, and so on. In the past, rivets, nails, bolts, and the like
were also produced by upset forging, but the method has largely been replaced
by cold heading.
Upset forging is performed in a horizontal forming machine which is rated
by the diameter of the largest billet or form it will accept. Figure 4-11 shows
an automatic upset forging machine for producing rear axles with flanges.
 159
Section 4-2: Primary Processes

CO OO

K POR
PISO INSve

yO
oe

<2
 160
Chapter 4: Manufacturing Processes and Design

Figure 4-11 An automatic upset forging machine capable of accepting work

pieces 6 in. in diameter. This machine produced the axles shown in Figure 4-12.
[Courtesy of the Ajax Manufacturing Company.]

Figure 4-12 Axles made on an automated 6-in. forging machine. The sequential
steps required to form an axle by upset forging are shown from left to right.
(Courtesy of the Ajax Manufacturing Company.]

Usually where the upsetting of a large amount of metal is required, the

process is completed by a series of steps in several machines rather than by
one automatic machine (see Figure 4-12).
The final product manufactured by upset forming lends itselftoreasonably
intricate forms with good dimensional accuracy as well as rapid production
rates. However, the size of the product is limited, and it does not have a
finish equal to that of other forgings. Also, tooling can be quite expensive.
 161
Section 4-2: Primary Processes

Other processes
Other important primary processes for forming and shaping metal
products are (1) cold heading, (2) cold extruding (also known as impact
extruding), (3) stamping, (4) deep drawing, (5) spinning, (6) roll forming,
(7) hot extruding, (8) press-brake forming, (9) section contour forming,
(10) rotary swaging, (11) wire and metal ribbon forming, (12) thread and
roll forming, (13) hobbing, (14) rubber forming, (15) hydraulic forming,
(16) explosive forming, (17) electrohydraulic forming, (18) electromagnetic
forming, (19) pneumatic-mechanical forming and (20) electroforming.
The use of a particular method depends upon a variety of factors: the type
of item to be fabricated, the quality and accuracy required, the quantity to
be produced, the material to be formed, the limiting cost of the item, the size
of the item, and so on.
The processes are too numerous to describe in this text. However, the
reader can find many fine books (for example, reference[2]) that give detailed
descriptions of these methods. In particular, reference [3] is especially
recommended because it discusses production processes and their influence
on design.
Figures 4-13 to 4-21 inclusive show respectively the production processes
of stamping, drawing, spinning, form rolling, hot extruding, and press-brake
forming.

Figure 4-13 Sixty-ton high speed

reproduction type stamping press. [Courtesy
Niagra Machine & Tool Works.]
 162
Chapter 4: Manufacturing Processes and Design

Figure 4-14 Different stages ofa

typical stamping produced by a
multiple-die high speed stamping press.
The photo shows the forming of the
“pull-tab”’ for a soft drink (or similar)
can. [Courtesy Niagra Machine & Tool
Works.]

Drawing die

Metal blank

A
Drawplate

Race
or
Punch blankholder

LLL

Figure 4-15 Drawing sequence for cylindrical cup. Punch is shaped to produce
desired configuration on interior of cup. Drawplate, or blankholder, and drawing
die hold the metal blank. Drawing begins in step 2 as blankholder and drawing die
move downward and stretch the blank over the punch. Completed part in step 3
is ejected from the drawing die by the knockout ram after the tooling has returned
to the starting position. [From 1/970 Metals Reference Issue, Metals Design,
Penton Publishing Co., Cleveland, Ohio.]
 163
Section 4-2: Primary Processes

Figure 4-16 Left tube was drawn to

14-in. diameter, 94 in. long with a 0.035-in. wall.
This tube was formed from a hot-rolled steel
plate 43-in. in diameter by 0.140 in. thick. The
tube was formed by five draws. The right
tube is made of aluminum and was also
made in five draws. The original stock was
24 in. in diameter by 2 in. thick. [Courtesy
of McCauley Metal Products, Inc.]

(a) ; (b) ' (c)

Figure 4-17 Shapes that are readily

, formed by spinning. [From 1970 Metals
Jb Reference Issue, Machine Design. Penton
(d) (el Y f Publishing Co., Cleveland, Ohio.]
e

(g) (h) (i)

 164
Chapter 4: Manufacturing Processes and Design

Figure 4-18 Yoder M-2 roll forming machine with flying cutoff, used in the
manufacture of building siding. This unit is capable of making siding of prepainted
aluminum or stainless steel. [Courtesy Yoder Company. ]
 C=
°
al
n
~oO
%
es)
7 OO
O
=

y.]

4-19
Figur
Typica
shape
alumi
by
forme
extrus
hotthe
Co
Met
Rey
 166
Chapter 4: Manufacturing Processes and Design

Figure 4-20 A 900-ton press brake bending a machine base. [Courtesy of

Cincinnati Incorporated.]
 167
Section 4-3: Forming of Plastics

Casket

Corrugating Narrow flange

Figure 4-21 Representative shapes that can be formed by press brake die
sets. [Courtesy of Cincinnati Incorporated.]

SECTION 4-3
Forming of Plastics
There are about 15 different methods of forming products made of plastic
material (see Table A-15 in Appendix A for properties of some plastics).
However, the primary methods are hot compression molding, injection
molding, transfer molding, extrusion forming, laminating, cold molding, and
casting.
Hot Compression Molding. A press operation in which a molding powder
(or preform) is placed inside a heated mold and melted. The male portion of
the mold is then lowered into the open cavity forcing the melting plastic to
assume the shape ofthe mold. Figure 4-22 shows a typical set of molding dies
used for hot compression molding.
Injection Molding. (See Figure 4-23.) A forming process that is very similar
to die casting. In this process, granular thermoplastic material is fed from a
hopper into a heated chamber by means of a plunger. The granules melt, and
the semiliquid material is formed through a nozzle into a mold cavity by an
arrangement of sprues and runners. Because the dies are cooled, the melted
plastic solidifies rapidly while assuming the mold cavity shape.
Injection molding is a high production rate process that produces a product
having a good surface finish as well as dimensional control. Many parts can
be produced with little finishing work required.
 168
Chapter 4: Manufacturing Processes and Design

Movement
of plunger

Plunger

Top clamp
plate

Preforms
Runner
plate

yaw
lees

me
Movement
of mold

Ejector

Figure 4-22 Schematic of hot compression forming process. [Courtesy

Hooker Durez Division, Hooker Chemical Corp.]

Torpedo Granular
Ejector pin plastic

Molded part :
Plunger

Figure 4-23 Injection molding. [Courtesy Tennessee Eastman

Corporation. ]

Transfer Molding. This process is similar to injection molding and avoids

some of the disadvantages of hot compression molding (for example, uneven
flow). In transfer molding, a thermosetting powder or preform is placed into
a heated transfer chamber. When the material charge is melted, it is forced
into the cavity of a closed mold where polymerization takes place (Figure
4-24). In comparison with hot compression molding, the dies are initially
closed and the product is formed not so much by pressure as by the free flow
of the melted plastic.
 169
Section 4-3: Forming of Plastics

POT TRANSFER

Clamping plate

Pot plunger

Seal groove——-

Cull pickup

Figure 4-24 Transfer molding process. [Courtesy Hooker Durez

Division, Hooker Chemical Corp.]}

The molds are kept hot, and the formed part is ejected from the mold
without any cooling other than that which is due to the ambient temperature.
Transfer molding permits the fabrication of thin-sectioned components in
addition to being capable of “molding in” metal inserts.
Extrusion Forming. This process is illustrated in Figure 4-25. Thermo-
plastic resin in granular or powder form is placed in a feed hopper. The

Heaters

110 Air or water

gs A\ did, OD
WY 7 Romy <
D YE ENENEN): Ze Conveyer

Figure 4-25 Schematic of the extrusion process. [From E. Paul De

Garmo: Materials and Processes in Manufacturing, 3rd ed. The
Macmillan Company, New York, 1971.]
 170
Chapter 4: Manufacturing Processes and Design

material then enters a heated cylinder within which a feed screw forces the
plasticized material through a hot die having the desired extrusion shape. The
extruded form passes from the die onto a conveyer or set of rollers and is
solidified by air or water.
Laminating. A fabricating method whereby materials such as paper, cloth,
wood, asbestos, and so on are immersed in a liquid thermosetting resin and
cut to some appropriate length. A sufficient number of sheets to produce a
desired thickness are then stacked in a hydraulic press and compressed at
an elevated temperature to produce a laminated sheet. It is possible to
produce shapes other than flat sheets by using shaped molding dies between
the forming press platens.
If directional properties of the final product are of importance, it is possible
to produce a relatively homogeneous material by using a shredded or ground
filler (that is, paper, wood, fabric, and so on) in the resin, and then follow the
normal molding procedure. Figure 4-26 emphasizes the directional differences
of the mechanical properties of two gears.
Cold-molded parts are made by compressing a powdered resin in a non-
heated die by means of a plunger type press. The compacted parts are then
oven baked at an elevated temperature (that is, 225—S00°F, depending on the
resin) for 12-14 hr.

Figure 4-26 The left gear was cut

from laminated blanks. The right gear
was cut from a material blank containing
a ground fabric. [Courtesy of The
Formica Corporation. ]

The powdered materials are either the nonrefractory type having a

phenolic resin base or the refractory type containing an inorganic binder (for
example, asbestos). Thus, any components formed by this method are
effective in withstanding heat and also have good electrical resistivity.
Casting. An inexpensive way offorming parts for a “short run”’ production
or for use as a prototype. Molds can be made of wood, plaster, metal, or lead.
Molds made ofthese materials only permit the casting of simple shapes (that
 ial
Section 4-4: Welding

is, no undercuts, reentrant curves, and so on). More complicated shapes are
molded by means of split molds or molds made of rubber.
The materials commonly used in casting are acrylics, styrenes, polyesters,
epoxies, silicones, and nylons. Frequently, these materials are mixed with
reinforcing fillers such as glass cloth, glass mat, flock, wood, and Fuller’s
earth.
Plastic castings are used in making jewelry, ornaments, novelties, and so
on, but are also used in the fabrication of drill jigs, punches, dies, and a
variety of other tools.
Miscellaneous methods of molding or shaping are blow molding, vacuum
forming, slush molding, and filament winding.

SECTION 4-4
Welding?
The American Welding Society defines welding as “‘a localized coalescence
of metals wherein coalescence is produced by heating to suitable temperatures
with or without the application of pressure and with or without the use of
filler metal. The filler metal either has a melting point approximately the
same as the base metals or has a melting point below that of the base metals
but above 800°F.”’
There are 34 different welding processes. Figure 4-27 is a master chart of
these processes and shows their connection to a major welding method. The
proper choice of a particular method of weld must be carefully considered
by the designer. And, before making a final selection, the designer will have
to consider, evaluate, and weigh such factors as the metals to be joined, the
joint design, the thicknesses (or bulk) of metals, the type of load, the equip-
ment available, the production rate, and the environment to which the weld
will be subjected. It is apparent that there are no “hard and fast”’ rules that
one can use in making a decision with the exception, perhaps, ofa particular
case where one requirement or condition is of overall importance.
Because of space restrictions, it is not possible to discuss in detail each of
the processes in Figure 4-27. We shall confine ourselves to describing briefly
the most widely used welding methods.°

Gas welding
Gas welding is “a group of welding processes wherein coalescence is
produced by heating witha gas flame or flames with or without the application
of pressure and with or without the use of filler material.”

? For the analysis of welded joints, refer to Chapter 17.

3 The definitions quoted are those found in the American Welding Society’s booklet No. AWS
A3.0-69 entitled Terms and definitions. For an in-depth analysis of welding, the reader is advised
to study the American Society of Welding Handbook(s), Vols. 1 through 5.
 I72
Chapter 4: Manufacturing Processes and Design

furnace
torch brazing induction
infrared a (FB) prazing resistance
brazing (T8) (18) brazing
cold (IRB) (RB) dip
welding brazing
(CW) (08) Oxy
diffusion acetylene
welding welding
(DFW) OAW)
xy
explosion’ hydrogen
welding welding
(EXW) (OHW)
pressure
forge gas
welding welding
(FOW) Brazing (PGW)

resistance’
friction :
welding welding
(FRW) Solid Gas ey
state welding
welding
Tesistance
ultrasonic -seam
welding welding
(usw) (RSEW)

electron Projection
beam welding
welding (RPW)
EBW)
Other Resistance
processes welding flash
electroslag
welding
welding (FW)
(EW)

induction ane ee
welding welding
welding UW
(UW)
(IW)

laser beam percusson

welding welding
(LBW) (PEW)

thermit stud
welding welding
(TW) (SW)
carbon-arc. pesto
welding welding
ea) shielded submerged\ \_(PAW)
metal-arc flux gas arc
welding oe gas tungsten welding
c
SED welding retaliate welding Sey
F
(FCAW) Hassel
MAW (GTAW),

Figure 4-27 Master chart of welding processes. [Courtesy

of The American Welding Society.]

Of the three types of gas welding, oxyacetylene welding (OAW) is the one
most frequently employed. This method uses a mixture of oxygen and
acetylene to produce heating. Fluxes may be used to reduce oxidation and
to promote a better weld joint. This type of welding is suitable for both
ferrous (including cast iron) and nonferrous metals and is capable of welding
thick metal sections.
Oxyhydrogen welding (OHW) is used for low melting point metals such as
aluminum, magnesium, and lead.
Pressure gas welding (PGW) uses an oxyacetylene flame for a heat source
but does not require a filler rod. Instead, fusion is obtained by applying
pressure to the heated parts, either while being heated or after the parts are
heated. This form of welding can be used for joining both ferrous and non-
ferrous metals.
 [73
Section 4-4: Welding

Arc welding
Arc welding is “a welding process wherein coalescence is produced by
heating with an arc or arcs with or without the application of pressure and
with or without the use of filler metals.’ As indicated in Figure 4-27, there
are eight different arc welding processes. These are (1) carbon-arc welding,
(2) shielded metal-arc welding, (3) flux cored arc welding, (4) gas metal-arc
welding, (5) gas tungsten-arc welding, (6) submerged arc welding, (7) plasma-
arc welding, and (8) stud welding.
The most widely used of these methods is the shielded metal-arc welding
(SMAW) process. It is defined as “‘an.arc welding process wherein coalescence
is produced by heating with an arc between a covered metal electrode and the
work. Shielding is obtained from decomposition of the electrode covering.
Pressure is not used and filler metal is obtained from the electrode.”’ Figure
4-28 shows the nomenclature associated with arc welding.

Electrode
core wire

Extruded
coating

Metal
Base

real
Gaseous shield
Arc stream
Slag
Weld metal
Base metal Molten pool

Penetration or Throat of weld Root of weld

Lower leg of weld

Figure 4-28 Elements of the arc welding process.

The shielded metal-arc process is employed in both manual and automated

production setups. Electrodes are available that permit the welding of ferrous
metals (including cast iron), all grades of carbon steels, low-alloy high-
strength steels, stainless steels, copper bearing steels, copper alloys,
aluminum, nickel, nickel alloys, and bronze. This welding technique is used
in many fields, particularly in the manufacture of machinery, transportation
equipment, piping systems and in various structures (for example, buildings,
trusses, machine bases, and so on).
The next two most widely used arc welding methods are the submerged
arc welding (SAW) process and the plasma-arc welding (PAW) process.
Submerged arc welding (SAW) is “an arc welding process wherein
coalescence is produced by heating with an arc or arcs between a bare metal
electrode or electrodes at work. The arc is shielded by a blanket of granular,
fusible material on the work. Pressure is not used and filler metal is obtained
from the electrode and sometimes from a supplementary welding rod.”
 174
Chapter 4: Manufacturing Processes and Design

This method can be used in fully automated equipment where the feeds of
both the electrode and granular flux are controlled. The method is also
adaptable for semiautomatic equipment where the feed of the electrode and
granular flux are controlled manually. Since the granular flux must cover the
joint to be welded, this method is restricted to parts in a horizontal position
and is particularly suited for welding long straight joints. Also, fewer passes
are needed to weld thick metal sections than are usually required by shielded
metal-arc welding.
Submerged arc welding can be used to weld low carbon steels, high-
strength low-alloy steels, chromium steels and austenitic chromium-nickel
steels. With special methods, it is also possible to weld high-alloy air-
hardening steels. Figure 4-29 shows the arrangement and nomenclature of
the submerged arc welding process.

f
To automatic wire feed
Welding wire
Welding flux
feed tube ><

Fused welding flux

Granulated Solidified fused flux

Flux-retaining Idi
welding
plates Finished weld surface

Plate metal
Weld metal

Weld backing + AC or DC

Figure 4-29 Submerged arc welding. [Courtesy Union Carbide Corp., Linde
Division. ]

Plasma-arc welding (PAW) is “‘an arc welding process wherein coalescence

is produced by heating with a constricted arc between an electrode and the
workpiece (transferred arc) or the electrode and the constricting nozzle
(nontransferred arc). Shielding is obtained from the hot, ionized gas issuing
from the orifice, which may be supplemented by an auxiliary source of
shielding gas. Shielding gas may be an inert gas or a mixture of gases. Pressure
may or may not be used, and filler may or may not be supplied.”
Plasma-arc welding is used for quality welding and can easily weld 5-in.
thick aluminum sections or stainless steel sections up to 4 in. thick. Since
there are no products of combustion, the welded joints have no porosity and
display a strong resistance to high stresses and impact loading.
 LZ:
Section 4-4: Welding

The plasma torch is constructed with an electrode centrally within a metal

cup that guides an inert streaming gas past the electrode. In the plasma-arc
torch, the discharge end of the cup is smaller in diameter than the upper
diameter so that a discharge nozzle is created. In addition, the inner wall of
the nozzle is lined with a ceramic material.
Although a plasma stream can be created with any gas, a gas that is
nonoxidizing should be used. Another important requirement is the thermal
conductivity of the gas rather than the temperature it may attain. Thus, gases
of higher conductivity can transfer more heat, making it possible to weld
“bulky” sections more easily. Argon, helium, and hydrogen are the gases
most frequently used. Hydrogen has the higher thermal conductivity and
produces hotter arcs than those produced by argon or helium.
Other forms of arc welding, namely, carbon-arc (CAW), flux cored arc
(FCAW), gas metal-arc (GMAW), gas tungsten-arc (GFAW) and stud
welding (SW) are used forjoining particular metals or for mass production.
For example, carbon-arc welding is used to join galvanized sheet steel, brass,
bronze, and aluminum, whereas flux cored arc welding uses a flux cored
electrode, continuously fed from a spool for quantity production.
Stud welding is accomplished by means of a stud-welding gun, which welds
a stud to the surface of a workpiece. The method is extensively employed in
the automotive, shipbuilding, railroad, and building construction industries.

Resistance welding
Resistance welding is “a group of welding processes wherein coalescence
is produced by the heat obtained from resistance of the work to electric
current in a circuit of which the work is a part, and by the application of
pressure.” Figure 4-27 indicates that there are six types of resistance welding
processes. These are (1) resistance spot welding, (2) resistance seam welding,
(3) projection welding, (4) flash welding, (5) upset welding, and (6) percussion
welding. Resistance welding is widely used for quantity production. By means
of proper controls and tooling, it is readily adaptable to automation,
including any required preheating or heat treatment after welding. The most
widely used types of resistance welding are the spot, seam, and projection
forms of welding.
Resistance Spot Welding (RSW). “A resistance welding process wherein
coalescence at the faying surfaces* is produced in one spot by the heat obtained
from the resistance to electric current through the work parts held together
under pressure by electrodes. The size and shape of the individually formed
welds are limited primarily by the size and contour of the electrodes”’ (see
Figure 4-30).
Spot welding is primarily restricted to thin metals (for example, 0.001-in.
thick to +-in. thick for steel and magnesium, 3}-in. thick for aluminum),

4 Faying surface is ‘that surface of a member which is in contact or in close proximity with
another member to which it is to be joined.”
 176
Chapter 4: Manufacturing Processes and Design

Figure 4-30 Schematic of a spot weld. [Courtesy of the American

Welding Society.]

namely, steels, stainless steels, aluminum, magnesium, nickel, nickel alloys,

bronze, and brass. Some dissimilar metals can be spot welded, but with
difficulty.
Resistance Seam Welding (RSEW). *‘A resistance welding process wherein
coalescence at the faying surfaces is produced by the heat obtained from
resistance to electric current through the work parts held together under
pressure by the electrodes. The resulting weld is a series of overlapping
resistance-spot welds made progressively along a joint by rotating the
electrodes.” In principle, seam welding is similar to spot welding except that
the weld is continuous by virtue of the rollers rather than discontinuous as
in spot welding (see Figure 4-31).
Seam welding is primarily used for quantity production but is restricted to
joining metal gages that are thinner than those which can bejoined by spot
welding. The “‘normal”’ range of thicknesses compatible with seam welding
is 0.100—4 in.
Projection Welding (RPW). “A resistance welding process wherein
coalescence is produced by the heat obtained from resistance to electric
current through the work parts held together under pressure by electrodes.
The resulting welds are localized at predetermined points by projections,
embossments, and intersections.”
Projection welding is a process similar to spot welding except that the
projections tend to localize the heat, permitting thicker materials to be
 LT.
Section 4-4: Welding

Figure 4-31 Schematic of a seam weld. [Courtesy of the American

Welding Society.]

welded. Simultaneous welds are readily made by this method, and result in
a stronger welded structure than that obtained with spot welding.
Flash Welding (FW). In this process abutting surfaces to be welded are
clamped in fixtures and brought within close proximity (or light contact) of
each other so that an electric arc is produced between the surfaces causing
them to heat to a fusible temperature. At this point, the two surfaces are
forced together, completing the weld. Forcing the two surfaces together
causes the metal to be displaced (that is, bulge) outward from the welded
joint. This “upset’’ metal is usually removed after welding. Preheating (for
large bulk parts) and postheating (that is, heat treatment) can readily be made
part of the overall welding cycle.
Materials that are easily weldable by spot welding are also weldable by
flash welding, although the method is used mostly with ferrous metals.
Copper, copper alloys, and some aluminums cannot be relied upon to
produce satisfactorily welded joints. However, dissimilar metals can readily
be welded by this method, including even refractory metals such as tungsten,
molybdenum, and tantalum.
Upset Welding (UW). A process similar to flash welding except the parts
to be welded are held in close contact with each other before the electric
circuit is closed. Thus, there is no flashing in this method. Upset welding is
extensively used in the fabrication of tubular sections, pipe, and heavy steel
rings; it is also used for joining small ferrous and nonferrous strips.
Percussion Welding (PEW). “A resistance welding process wherein
coalescence is produced simultaneously over the entire abutting surfaces by
 178
Chapter 4: Manufacturing Processes and Design

heat obtained from an arc produced by a rapid discharge of electrical energy

with pressure percussively applied during or immediately following the
electrical discharge.”
Percussion welding is used for special joining situations (for example,
joining dissimilar metals that cannot be welded economically by flash
welding). This welding method is also used to weld pins, studs, bolts, and so
on, to other components as well as to join sections of pipe, rod, or tube to
each other or to flat sections.

Brazing
Brazing? is ‘“‘a welding process wherein coalescence is produced by heating
to suitable temperatures and by using a filler metal having a liquidus above
800°F (427°C) and below the solidus of the base metals. The filler metal is
distributed between the closely fitted surfaces of the joint by capillary
attraction.” As indicated by Figure 4-27, there are six brazing methods,
namely, (1) infrared brazing (IRB), (2) torch brazing (TB), (3) furnace
brazing (FB), (4) induction brazing (IB), (5) resistance brazing (RB), and
(6) dip brazing (DB). Among these methods, the primary difference is the
manner in which the metals to be joined are heated. Also, only four of the
six methods are of industrial importance, torch brazing, furnace brazing,
induction brazing, and dip brazing. These methods are defined and briefly
described in the following paragraphs.
Torch Brazing. A joining process that may employ acetylene, natural gas,
butane, or propane in combination with air or oxygen to supply the heat
required to melt the filler rod and diffuse it into the surface of the base metal.
This technique 1s not extensively used for continuous mass production.
Furnace Brazing. A high production fabrication method where the heat is
supplied by gas or electric heating coils. The furnaces are of the box type or
the continuous type, which employ a wire mesh belt to transport the parts to
be brazed. Furnace brazing requires that preformed shapes of filler metal be
placed on the parts to be joined prior to entering the furnace. This method of
brazing 1s well suited to high production and can avoid the use of fluxing by
maintaining an inert atmosphere in the furnace.
Induction Brazing. Like furnace brazing, induction brazing requires the
use of preformed shapes offiller metal. Heat is produced by placing the parts
to be brazed within the field of a high frequency induction coil. The work
pieces are heated by eddy currents because the parts to be joined offer
electric-magnetic resistance to the changing induction field. Heating is very
rapid, and by properly shaping the induction coils, the heat can be applied
in the local area of the joint to be brazed. Induction brazing can be used for
mechanized production when properly designed tooling and feeding devices
are incorporated into the production setup.

° Brazing includes the joining method called silver soldering. This term is erroneously used to
denote brazing by employing silver-base filler rods.
 179
Section 4-4: Welding

Dip Brazing. Parts can be dip brazed by one of two methods. In chemical
dip brazing, the parts to be joined are prepared with preformed filler metal,
after which they are placed into a molten bath of brazing flux. In the molten
metal bath process, the assembled parts are first prefluxed and then immersed
into a molten bath of filler metal. This latter method of brazing is restricted
to small parts, whereas the former method is more adaptable for joining
large parts.

Other welding processes

The master chart of welding processes (Figure 4-27) shows, in addition to

the aforementioned methods, a series of solid state welding techniques and
some other processes. These methods are defined by the American Welding
Society as follows:
Ultrasonic Welding (USW). “A solid state welding process wherein
coalescence is produced by the local application of high frequency vibratory
energy as the work parts are held together under pressure.”
Friction Welding (FRW). “A solid state welding process wherein
coalescence is produced by heat obtained from mechanically induced sliding
motion between rubbing surfaces. The parts are held together under
pressure.”
Forge Welding (FOW). “A solid state welding process wherein coalescence
is produced by heating and by applying pressure or blows sufficient to cause
permanent deformation at the interface.”
Explosion Welding (EXW). “A solid state welding process wherein
coalescence is effected by high velocity movements produced by a controlled
detonation.”
Diffusion Welding (DFW). “A solid state welding process wherein
coalescence of the faying surfaces is produced by the application of pressure
and elevated temperatures. The process does not involve macroscopic
deformation or relative motion of parts. A solid filler metal may or may not
be inserted.”
Cold Welding (CW). “‘A solid state welding process wherein coalescence is
produced by the external application of mechanical force alone.”
Thermit Welding (TW). ‘A group of welding processes wherein coalescence
is produced by heating with superheated liquid metal and slag resulting from
a chemical reaction between a metal oxide and aluminum with or without
the application of pressure. Filler metal, when used, is obtained from the
liquid metal.”
Laser Beam Welding (LBW). “‘A welding process wherein coalescence is
produced by the heat obtained from the application of a concentrated
coherent light beam impinging upon the surfaces to be joined.”
Induction Welding (IW). ““A welding process wherein coalescence is
produced by the heat obtained from resistance of the work to induced electric
current with or without the application of pressure.”
 180
Chapter 4: Manufacturing Processes and Design

(a) BUTT (b) CORNER

$ Sa
\
Ss

if
Fillet Single bevel

R=
Single bevel Double bevel

o aye

R
Single J
Double J

Ro ||—
Single U

Ger
bel

Single V

Double fillet
Ay

Single J Single fillet, single J Double fillet, double J

Figure 4-32 Fundamental types of welded joints. [From S. W. Bernard, E. O.

Waters, and C. W. Phelps: Principles of Machine Design. The Ronald Press
Company, New York, 1955.]
 18]
Section 4-5: Secondary Production Processes

Electroslag Welding (EW). ““A welding process wherein coalescence is

. produced by molten slag which melts the filler metal and the surfaces of the
work to be welded. The weld pool is shielded by this slag which moves along
the full cross section of the joint as welding progresses. The conductive slag
is maintained molten by its resistance to electric current passing between the
electrode and the work.”
Electron Beam Welding (EBW). “‘A welding process wherein coalescence
is produced by the heat obtained from a concentrated beam composed pri-
marily of high velocity electrons impinging upon the surfaces to be joined.”’
Figure 4-32 shows schematically the basic types of welded joints and the
plate edge preparation (that is, single U, double U, single J, and so on)
required to insure a full depth weld when joining thick plates. Also shown in
the figure, are the standard symbols denoting each type of weld. Figure C-1
in Appendix C summarizes the American Welding Society’s nomenclature
of standard welding symbols used for specifying welds.

SECTION 4-5
Secondary Production Processes
Secondary production processes, in contrast with primary production
processes, remove metal from a workpiece in order to achieve a specified
shape or configuration. The methods of metal removal may be classified into
three categories, namely, chip removal, chipless removal, and metal removal
by heat. The most widely used of these methods is chip removal. It employs a
hardened cutting tool to form (and remove) the chip. The most commonly
used forms® of chipless machining are electrochemical machining, chemical
milling, electrical discharge machining, electrolytic grinding, and photo-
forming. The third form of metal removal by heat employs a hot flame, which
is capable of melting the work piece material (for example, oxyacetylene,
plasma, and so on).
Figures 4-33, 4-34, and 4-35 show schematically the various ways in which
machines are used for chip removal. The designer is not expected to be a
machinist, but he is expected to be able to design various parts for a machine,
structure, system, and so on, and, therefore, be knowledgeable about the
capabilities, type of cutting, and dimensional limitations of machine shop
equipment. Because it is not possible, due to space limitations, to study or
review the various secondary production processes, the reader is advised to
refer to the many fine texts (for example, reference [2]) and handbooks
available on this subject matter.
Figures 4-33, 4-34, and 4-35 are basic representations because, in almost
every type of operation, special or high-production rate machines exist that

© New developments in chipless machining employ laser beam and ultrasonics. However,
these methods have, as yet, not come into wide use due to various limitations.
 Figure 4-33 Schematic of machine tool operations for machining flat surfaces.
[From E. Paul De Garmo: Materials and Processes in M anufacturing, 3rd ed. The
Macmillan Company, New York, 1971.]
=
Most Machines Machines
0 ; : Commonly Less Seldom
pated Diagram of Mal and Used Frequently Used
Tool Motion Machines Used

tool rf he
4; work
Sain Horizontal Vertical
sete ge shaper shaper

=| =
a TOO!
Planing LS Ge wes Planer

it
tool

Oa
x Slab milling

a
Lathe
ee
Millin Milling
illing :
[ work tool aes machine specia
as
SS ra Face milling attach-
aae
ment)
1
t

es iF T
Pra
tool 2

is ae. Lathe porns

~- mill

lis te
tool
k 4

machine

[ | [
5 2) pas Lathe

Grinding SDSL tool stie

grinder
(with
:
specia
attach-
ment)
IL
tool
Sawing H Cutoff Contour
| work saw saw

T “| 1
hen EE Vertical
shaper

Tool and work motion

——— Feed only

182
 Figure 4-34 Schematic of machine tool operations for machining internal
cylindrical surfaces. [From E. Paul De Garmo: Materials and Processes in
_ Manufacturing, 3rd ed. The Macmillan Company, New York, 1971.]

Most Machines | Machines

Commonly Less Seldom
Operation Diagram of Work and Used Frequently Used
Tool Motion Machines Used

Milling
machine

_ 1 work if - Drill oe Boring

Drilling fool =< press atne mill

Horizontal
boring
machine

Lathe
work Boring Milling
(( mill machine
Boring =
SS Horizontal Drill
ae coe press
| machine

i Lathe
Drill
press
Boring Milling
Reaming (t ahi machine

Horizontal
boring
machine
4 ——

Rare Lathe

tool Cylindrical eu
Grinding y Me ne special
ae Se attach-
oy ment)

Le =

tool
work Contour
Sawing ea iy
= i saw

| 1

tool | work
: Broaching
Broaching machine

183
 184
Chapter 4: Manufacturing Processes and Design

Figure 4-35 Schematic of machine tool operations for machining external

cylindrical surfaces. [From E. Paul De Garmo: Materials and Processes in
Manufacturing, 3rd ed. The Macmillan Company, New York, 1971.1

Most Machines Machines

; Commonly Less Seldom
Operation Diagram of Work and Used Frequently Used
Tool Motion Machines Used

Vertical
shaper
Turning
Milling
machine

Lathe

ae Cylindrical (with
Grinding : special
grinder oe

ment)

Contour
saw

Figure 4-36 Electrochemical machining of burs (electrochemical deburring) on

an automobile engine connecting rod. [Courtesy of the Ex-Cell-O Corporation. ]
 185
Section 4-5: Secondary Production Processes

Figure 4-37 Parts made by chemical milling. [Courtesy of the Chemcut

Corporation. ]

employ the same basic method of metal removal (for example, drill press and
radial drill, engine lathe and turret lathe, and so on).
Electrochemical Machining (EMC). In this process the metal work piece is
shaped by the removal of metal from its surface as a result of electrochemical
action. The tool (cathode) moves towards the work piece (anode), completing
the electrolytic circuit through the electrolyte. The feed rate of the tool,
electrolyte, and current density, depends upon the material being removed
and the final shape desired. Figure 4-36 shows a typical electrochemical
operation.
Chemical Milling (CHM). Also called chemical blanking and chemical
machining, chemical milling is a controlled etching process. The process does
not use electricity but depends on the etching action of an acid or alkali
(depending on the material) that uniformly attacks all exposed areas of the
work piece. Thus, in order to obtain a desired configuration or shape, it is
necessary to use a mask or a protective coating on those surfaces which are
 156
Chapter 4: Manufacturing Processes and Design

not to be etched. Figure 4-37 shows some typical parts produced by this
method.
Electrodischarge Machining (EDM). This process, aiso known as electro-
spark machining, is intended to machine “difficult”? materials (for example,
carbides, tool and die steels, heat resistant alloys, and so on). The work piece
is brought to the desired shape by electric sparks which remove the metal
being worked by vaporization. The tool is shaped to produce the desired
configuration. Sparking is initiated when the forming tool is brought into
close proximity to the work piece. Electrodischarge machining is a slow
process and is not intended for production purposes. Figure 4-38 is a
photograph of a typical part produced by the EDM process.

Figure 4-38 A splined hole machined by

the electrodischarge method. [Courtesy of
the Ex-Cell-O Corporation.]

Electrolytic Grinding (EGG). Also called electrochemical grinding, this is

a process similar to electrochemical machining. In this machining method,
the grinding wheel is the cathode, and the work piece is the anode. Although
this method was originally developed for sharpening carbide cutting tools
(which normally causes rapid wearing of diamond grinding wheels used in
normal sharpening), electrolytic grinding can also be used for grinding other
hard materials. The results produced by this method are of high quality (for
example, a surface finish of 5-10 yin. can be obtained, and dimensions can
be held within 0.005 in.), and are also obtained more rapidly than from
ordinary grinding methods. Figure 4-39 shows a schematic setup for the
electrolytic grinding of a cutting tool.
 187
Section 4-5: Secondary Production Processes

Electrode wheel

Current pickup brushes

Electrolyte GMMMIMMMML MALS ee

SS Grinder spindle
LAI fete

Insulating abrasive
particles

Work table
a

Figure 4-39 Schematic setup for electrolytic grinding. [From E. Paul

De Garmo: Materials and Processes in Manufacturing, 3rd ed. The
Macmillan Company, New York, 1971.]

Photoforming. A method akin to chemical milling in that metal is removed

by chemical etching. However, it is different than chemical milling in that
the areas not to be etched are masked by shapes that are photographically
produced to very close dimensions. In addition, photoforming removes all
the metal from an exposed area, whereas chemical milling merely removes
enough material to produce cavities in the work piece surface. Photoforming
is usually intended for relatively thin material and is used for making masks
for the manufacture of transistors, printed circuits, strain gages, and so on.
Figure 4-40 shows the sequential process of machining a part by photo-
forming.

® aN
Q =9 uw) @®
Transparency

(a) Metal sheet (b) (c) (d)

covered with
photoresist

Figure 4-40 Photoforming. (a) A transparency containing an image of the

desired shape is applied to a metal sheet containing a photoresistive coating.
(b) The sheet is exposed to light, and a subsequent bath removes all resist material
except that shaded by the image. (c) The remaining resist then serves as a mask in
a chemical etching operation, which (d) removes all unmasked metal to produce a
part having a profile identical to the initial image. [/970 Metals Reference Issue,
Machine Design. Penton Publishing Co., Cleveland, Ohio.]

Ultrasonic Machining. An important unconventional machining method

intended for work on hard and/or brittle materials. This method employs a
 Figure 4-41 Ultrasonic machining
(impact grinding). (a) The head unit of
the machine. (b), (c), and (d) Representa-
tive difficult types of products that can be
produced by this method of fabrication.
(b) Long-life heading and stamping dies
cut from tungsten, carbide steel. (c) A
ceramic spacer with multiple drilled
holes. (d) Holes and slots cut in a
ceramic spacer for an electronic tube.
[Courtesy Raytheon Manufacturing
Company.]

188
 189
Section 4-6: Engineering Drawings and Dimensions

tool of a softer material (cold rolled steel, brass, or copper) than the work
- piece which has the contour of the shape to be cut. This tool is mounted on
a high frequency, low amplitude (for example +0.002 in.) transducer, and
oscillates as shown in Figure 4-41 at approximately 25,000 cycles per second
(cps). While this motion is taking place, a slurry containing abrasive particles,
such as silicon carbide, boron carbide or aluminum oxide, is directed at the
tool and the work piece, cutting the shape of the tool into the work surface.
The tool is subject to some wear, but is inexpensive to replace.

SECTION 4-6
Engineering Drawings and Dimensions
The result of a designer’s efforts must be translated into a set of instructions
to the shop in order that the part or parts can be fabricated and assembled.
Thus, a set of engineering drawings are prepared showing the sizes (that is,
to scale), shapes, and dimensions to which parts are to be made. Un-
fortunately, many designers consider this phase of engineering design to be
trivial. Realistically speaking, however, it may be of greater importance than
the design solution itself.
Consider, for the moment, what an engineering drawing represents. It is a
detailed set of instructions (that is, orders) that tells the machinist, molder,
die caster, and so on, “make this part in accordance with the information
indicated and to the dimensions specified—any unauthorized deviations or
errors made in fabrication are your responsibility.’ This statement is, of
course, exaggerated. Nevertheless, it is meant to convey the importance of
the complete and proper dimensioning of engineering drawings—the
responsibility for which rests with the designer. Careless dimensioning can lead
to increased production costs and/or outright waste as a result of errors.
Due to the fact that no part can be manufactured to an “‘exact”’ dimension,
shop drawings are prepared in accordance with a system of tolerances and
allowances. Many companies, by reason of their shop facilities and experience,
rely on their own standards for dimensioning drawings. We will base our
discussion on the widely used ANSI’ (American National Standards
Institute) ‘Preferred limits and fits for cylindrical parts” (Standard B4.1-—
1967) published by the American Society of Mechanical Engineers (see
Tables C-1, C-2, C-3, C-4, and C-5 in Appendix C). These tables indicate
the limits imposed on different sized components that will result in a particular
class offit between two mating parts. The tables provide a common basis for
dimensioning components for interchangeable manufacture.
When closer fits than those indicated by the tables are required, the designer
may reduce the tolerance of the mating parts. However, taking such action
could result in increased fabrication costs. In order to avoid any increased

7 ANSI was formerly known as the United States of America Standards Institute (USASI), and
prior to that was known as the American Standards Association (ASA).
 190
Chapter 4: Manufacturing Processes and Design

cost, the designer would of necessity have to resort to selective assembly.

The idea behind selective assembly is to specify large tolerances for the
mating parts, and then grade them by gaging in smail, medium, and large
fits. Thus, a small shaft mating with a small hole, a medium shaft mating with
a medium hole, and a large shaft mating with a large hole will all possess the
same fit allowance. One should, nevertheless, keepin mind that the additional
cost of purchasing the gages as well as the labor required for gaging may
offset any saving achieved by selective assembly.

SECTION 4-7
Definitions of Dimensioning Terms
For a thorough understanding of fits and tolerances, the following terms®
must be clearly understood.
Allowance. The allowance is the tightest fit between mating parts. For
interference fits, the allowance is negative. ‘
Nominal Size. The nominal size is the designation used for the purpose of
general identification. For example, a 24-in. diameter nominal pipe is actually
2.875 in. in diameter.

Hole tolerance zone

| Shaft tolerance zone

a & #

eS°
Basic
size

Lei.29 gee 1.494 + .002

Sa 1.498 1.498
~.004 7.493 1.493

be 1.500 +,004 | 1.500 + .002 1,500 | 1,502

— 000 1502 1.500
Unilateral tolerance Bilateral tolerance Max. material Min. material
method method

Limit dimensioning

Figure 4-42 Examples of the three systems of dimensioning. The trend is to

adopt limit dimensioning as the preferred system.

® This list does not include all of the ANSI terms but only those that are pertinent to the under-
standing of specifying fits and tolerances.
 JOT
Section 4-7: Definitions of Dimensioning Terms

ax
Max. clearance eC
0.002
: n
allc wance
— ; = allowance
(min. clearance) ;
Zz La LD + hole tolerance
+ shaft tolerance

(a) (b)
Minimum clearance (i.e. tightest fit) Maximum clearance (i.e. loosest fit)
is equal to the smallest hole dia. is equal to the largest hole dia. minus
minus the largest shaft dia. The the smallest shaft dia. The max. clearance
minimum clearance is the allowance is determined as indicated in the above figure

Clearance Fit

Max. salctliall ; :
0.003 min. interference t ee

=ltRe 0.013

2.013 PONS
2.008
Shaft

Hole

(c) (d)
Interferce Fit 2.000
2.005

Max. clearance
Max. hole
Max. shaft it

Min. shaft

|Interference

Min. hole

(e)

Transition Fit

Figure 4-43 The representations shown are in accordance with the basic hole
system. (a) and (b) A clearance fit using the limit dimensions of Figure 4-42.
(c) and (d) Maximum and minimum interference fit based on the dimensions as
shown. For interchangeability, there must be clearance between parts when they
are at maximum material condition (MMC), as in (a). Interference between parts
is dimensioned on the basis of the least material condition (LMC), as in (c).
 192
, Chapter 4: Manufacturing Processes and Design

Tolerance. A tolerance is the total permissible variation in the size of a

part.
Basic Size. The basic size is that size from which limits of size are derived
by the application of allowances and tolerances.
Unilateral Tolerancing. Unilateral tolerancing is a system of dimensioning
where the tolerance (that is, variation) is shown in only one direction from
the nominal size. Unilateral tolerancing allows the changing of tolerance on
a hole or shaft without seriously affecting the fit (see Figure 4-42).
Bilateral Tolerancing. Bilateral tolerancing is a system of dimensioning
where the tolerance (that is, variation) is split and is shown on either side of
the nominal size (see Figure 4-42).
Limit Dimensioning. Limit dimensioning is a system of dimensioning where
only the maximum and minimum dimensions
are shown. Thus, the tolerance
is the difference between these two dimensions. Two methods of designating
limit dimensions are considered as standard. One method is the maximum
material method in which the large dimension is placed above the smaller
dimension for male parts, and the reverse is true for female parts. This
method is well suited for small lot quantities because it is likely that the
machinist himself may check the dimensions ofthe parts. In so doing, he will
be verifying initially the larger dimension of the male part and the smaller
dimension of the female part. The other method is the maximum number
method and is preferred by production and quality control departments.
In this method of designating a dimension, the larger number is always
placed above the smaller number, regardless of whether the part is male or
female.
Clearance Fit. A clearance fit is one having limits of size so prescribed that
a clearance always results when mating parts are assembled (see Figure
4-43a and b).
Interference Fit. An interference fit is one having limits of size so prescribed
that an interference always results when mating parts are assembled (see
Figure 4-43c and d).
Transition Fit. A transition fit is one having limits of size so prescribed that
either a clearance or an interference may result when mating parts are
assembled (see Figure 4-43e).
Basic Hole System. A basic hole system is a system of fits in which the
design size of the hole is the basic size from which the allowance is subtracted
to obtain the diameter of the shaft. The basic hole is the preferred system
because standard drills, reamers, broaches, plug gages, and so on can be used
and shafts can then easily be machined to fit.
Basic Shaft System. A basic shaft system is a system of fits in which the
design size of the shaft is the basic size to which the allowance is added to
obtain the diameter of the hole. (Normally most design offices adhere to the
basic hole system because of standard tooling and the fact that tolerance
tables are based on the basic hole system. However the use of the basic hole
system has the great advantage in that it is possible to use a standard size
shaft.)
 [93
Section 4-8: Classes of Fits

SECTION 4-8
Classes of Fits
As indicated in Section 4-6, the ANSI Standard B4.1—1967, “Preferred
limits and fits for cylindrical parts,” is widely used for establishing tolerances
for various classes of fits. The letter symbols appearing in this standard
(Tables C-1, C-2, C-3, C-4, and C-5 in Appendix C) represent the following
classes:? RC (running or sliding fit), LC (locational clearance fit), LT
(transition fit), LN (locational interference fit), and FN (force or shrink fit).

Running or sliding fits

Running or sliding fits (Table C-1, Appendix C) provide a similar running
performance with suitable lubrication allowance throughout the range of
sizes. The clearance for the first two classes, used chiefly as slide fits, increases
more slowly with diameter than the other classes, in order that accurate
location is maintained even at the expense of free relative motion. There are
nine types of RC fits which are defined as follows.

RCI. Close sliding fits are intended to locate accurately parts that must
assemble without perceptible play.
RC2. Sliding fits are intended for accurate location, but with greater
maximum clearance than class RC1. Parts made to this fit move and
turn easily but are not intended to run freely and, in the larger sizes,
may seize with small temperature changes.
RC3. Precision running fits are the closest fits that can be expected to run
freely and are intended for precision work at slow speeds and light
journal pressures. However, they are not suitable where appreciable
temperature changes are likely to be encountered.
RC4. Close running fits are intended chiefly for running fits on accurate
machinery with moderate surface speeds and journal pressures
where accurate location and minimum play is desired.
RC5 and RC6. Medium running fits are intended for higher running speeds
or heavy journal pressures or both.
RC7. Free running fits are intended for use where accuracy 1s not essential
or where large temperature variations are likely to be encountered,
or under both of these conditions.
RC8 and RC9. Loose running fits are intended for use where materials
such as cold-rolled shafting and tubing made to commercial
tolerances are involved.
A graphical representation of RC fits is shown in Figure 4-44.

° These letter symbols (i.e., RC, LC, etc.) are not used on manufacturing drawings but are
intended to guide the designer to select tolerances consistent with his design requirements.
Other symbols such as HS, g4 (H5g4), H6hS, etc., are in accordance with the American, British,
Canadian (ABC) system.
 194
Chapter 4: Manufacturing Processes and Design

6 RCQ
4 : RC8
RC7
2 = RCI RC2 seRCOa RCA RES RCO rat
0 peal two _\eerat___\weal__Nsl__ S|
eS aa
2
4
size
Basic
6 Holes —@ Shafts &

8 Scale: thousandths of an inch for —

10 a diameter of 1 in.

Figure 4-44 Graphic representation of running and

sliding fits.

Location clearance fits

Location clearance fits (Tables C-2, C-3, and C-4 in Appendix C) are
intended to determine only the location of the mating parts; they may
provide rigid or accurate location, as with interference fits, or some freedom
of location, as with clearance fits. Accordingly, they are divided into three
groups: clearance fits, transition fits, and interference fits. These fits are more
fully described as follows:
LC. Locational clearance fits (Table C-2, Appendix C) are intended for
parts that are normally stationary but can be freely assembled or
disassembled. They run from snug fits for parts requiring accuracy of
location, through the medium clearance fits for parts such as spigots,
to the looser fastener fits where freedom of assembly is of prime
importance.
LT. Transition fits (Table C-3, Appendix C) are a compromise between
clearance and interference fits for application where accuracy of
location is important, but either a small amount of clearance or
interference is permissible.
LN. Locational interference fits (Table C-4, Appendix C) are used where
accuracy of location is of prime importance and for parts requiring
an alignment with special requirements for bore pressure. Such fits
are not intended for parts designed to transmit frictional loads from
one part to another by virtue ofthe tightness offit, as these conditions
are covered by force fits.

A graphical representation of clearance locational fits, transition location

fits, and interference locational fits is shown in Figure 4-45a, b, and c
respectively.

Force fits

Force fits (Figure C-5, Appendix C) or shrink fits constitute a special type
of interference fit, normally characterized by maintenance of constant bore
pressures through the range of sizes. The interference, therefore, varies
almost directly with diameter, and the difference between its minimum and
maximum values is small to maintain the resulting pressures within reasonable
 195:
Section 4-8: Classes of Fits

size
Basic

Holes WB Shafts
Scale: thousandths of an inch for
a diameter of 1 in.

2 LT2 LT4 [SG aman (77

size
Basic
2 Holes & Shafts
Scale: thousandths of an inch for a diameter of 1 in.

(b)

LN3
2 LN2 =

Basic
size 0; Bf GaN
= Holes Be Shafts ——

Scale: thousandths of an inch

for a diameter of 1 in.

(c)

Figure 4-45 Graphic representation of (a) clearance

locational fits, (b) transitional locational fits, and
(c) interference locational fits.

limits. There are five types of force fits, which are described as follows:
FN1. Light drive fits are those requiring light assembly pressures and
produce more or less permanent assemblies. They are suitable for
thin sections or long fits or 1n cast iron external members.
FN2. Medium drive fits are suitable for ordinary steel parts or for shrink
fits on light sections. They are about the tightest that can be used
with high grade, cast iron external members.
FN3. Heavy drive fits are suitable for heavier steel parts or for shrink fits
in medium sections.
FN4 and FNS. Force fits are suitable for parts that can be highly stressed
or for shrink fits where the heavy pressing forces are impractical.
 196
Chapter 4: Manufacturing Processes and Design

A graphical representation of force fits or shrink fits is shown in Figure

4-46,

size
Basic
Holes WB Shafts
Scale: thousandths of an inch
for a diameter of 1 in.

Figure 4-46 Graphic representation of

force or shrink fits.

SECTION 4-9
Selecting Tolerances
Selecting tolerances is not a random process. It requires careful judgement
based upon design calculations, production facilities, and’cost. As a rule,
tolerances should beas large as possible because they generally will determine
the method of manufacture and, thus, markedly affect the cost of production.
This aspect is clearly indicated by Figure 4-47 where it is readily observed
that, as the tolerances decrease, the relative cost of production rises (quite
rapidly for small tolerance values).
It is essential that in selecting a tolerance the designer be aware of the
accuracies attainable from various machine shop operations and other

Figure 4-47 Relative cost of production as a

function of tolerances. The curve is based on a
variety of data, and it is only indicative of a
trend.

production
of
relative
Approximate
cost

0.004 0.012 0.020 0.028 0.036 0.044

Tolerances, in., (plus or minus)
 197
Section 4-9: Selecting Tolerances

manufacturing processes such as casting, molding, forging, and so on (see

Sections 4-2, 4-5, and Figure 4-48). To assist him in making a wise decision,
the designer may refer to Table 4-1 and Figure 4-48. A tolerance selected from
Table 4-1 determines the grade of machine work required. Knowing the

Figure 4-48 Machining processes expected to produce work within a given

tolerance. [Courtesy of the ASME, ANSI Standard B4.1—1967.]

Grades

Lapping and honing

Cylindrical grinding

Surface grinding

Diamond turning

Diamond boring

Broaching

Reaming

Turning

Boring

Milling

Planing and shaping

Drilling

Table 4-1 Standard Tolerances Arranged According to Size of Part and Grade of
Machine Work (value in thousandths of an inch)

Nominal
Grade Grade Grade Grade Grade Grade Grade Grade | Grade
Size Range
4 5 6 qf 8 9 10 11 12
Inches
Over To

0.04- 0,12 ; i : 0.4

0.12- 0.24
0.24- 0.40
0.40- 0.71
O71- 1.19
Nelo
1.97- 3.15
3.15=- 4.73
4.73- 7.09
7.09- 9.85
9.85- 12.41
12-41— 15.75 HNOmMNNwAANSO<
15.75- 19.69
19.69- 30.09 WwNN
=
MEW
N=

30.09- 41.49 i
41.49- 56.19
56.19- 76.39 no
76.39-100.9 Oe
as
am
rai
100.9- 131.9
131.9- 171.9
171.9- 200 Dnt

SOURCE: ASME: ANSI Standard B4,1-1967.

 198
Chapter 4: Manufacturing Processes and Design

grade, the designer can then determine the likely machining operation.
Table 4-1 has been prepared so that it also indicates the level of tolerance
achievable in any one machining operation for all of the nominal size ranges
listed.
As an illustration, we see from Table 4-1 that a grade 10 milling operation
(Figure 4-48) can be expected to produce a tolerance of 0.004 in. for a work
piece within the 1.19—1.97 in. range. On the other hand, the expected tolerance
for a larger work piece, for example the 4.73-7.09 in. range, is 0.006 in. Thus,
it becomes apparent that it is easier to achieve a smaller tolerance with smaller
work pieces for the same machining operation. Of course, one must also note
that any machining operation is capable of yielding a range of grades as
indicated by Figure 4-48. The smaller tolerances (that is, lower grades) are
most likely to be obtained from machines that are in excellent condition and
from careful workmanship. The larger tolerances can be expected from
machines that are in somewhat less than average condition or from poor
workmanship.
To further illustrate the use of the tolerance tables and Figure 4-48 consider
Example 4-1.

Example 4-1. Figure 4-49 is a drawing of a drive shaft that is used to drive a small
centrifugal fan by means of a V-belt drive. The power input is through a gear
keyed to the shaft. The shaft is free to rotate in self-lubricated flanged bearings,
which are held in the machine frame. Fit requirements are as follows: The
flanged bearings are to have a locational interference fit (LN2) with the frame,
the shaft is to have a free running fit (RC4) with the inner diameter (I.D.) of
the bearing, the V-belt pulley is to have a force fit (FN5) with the shaft, and the
gear is to have a free running fit (RC7) with the shaft.

V-belt pulley
Machine frame

1’
dia.
nominal

Flanged sleeve bearing

Figure 4-49 Example 4-1.

 199
Section 4-9: Selecting Tolerances

Determine the limit dimensions for the flanged bearings, the pulley mounting
hole, the gear mounting hole, and the shaft.
Solution: First we establish the locational interference fit LN2 between the
frame and the flanged bearings. Using the basic hole system, and referring to
Table C-4 in Appendix C, we establish the following:
Hole diameter in the frame: 1.5000/1.5010 in.
O.D. for the bearings: 1.5016/1.5010 in.
Thus, we see that the loosest fit between the bearings and the housing is
0.0000 in. (that is, “‘line-to-line”’ fit) and the tightest fit is a negative allowance
or interference fit of 0.0016 in.
Now we proceed to establish the limit dimensions for the pulley, I.D. of
the bearing, and the mounting hole for the gear. From Figure 4-49, the
nominal size to be used for the basic hole system is 1-in. diameter. Referring
respectively to Tables C-1 and C-5 in Appendix C we obtain the following
tabulated data.

Size Range in.

Standard Standard Standard
8 Limits 8 Limits ae Limits
z5 : pas
as Hole | Shaft EB Hole | Shaft 1 3] Hole |shaft
oO Oo oO
H8 ie H9 d8

0.8 |}+1.2 : 8) || ae2A0)

0.71-1.19
2.8 0) Dif 0

0.95-1.19

Thus we obtain the following for

The gear
The tolerance for the hole in the gear is 0.0020 in.
The tolerance for the shaft for the gear is 0.0012 in.
The allowance between the shaft and gear is 0.0025 in.
The flange bearing
The tolerance for the hole in the flange bearing is 0.0012 in.
The tolerance for the shaft for the flange bearing is 0.008 in.
The allowance between the shaft and flanged bearings is 0.0008 in.
The V-belt pulley
The tolerance for the hole in the V-belt pulley is 0.0012 in.
The tolerance for the shaft for the V-belt pulley is 0.008 in.
The allowance between the shaft and the V-belt pulley is —0.0033 in.
Figure 4-50 shows the limit dimensions for the shaft. The limit dimensions
for the mounting hole of the gear are 1.0000/1.0020 in. diameter. The limit
dimensions for the I.D. of the flange bearing are 1.0000/1.0012 in. diameter.
The limit dimensions for the mounting hole of the V-belt pulley are 1.0000/
1.0012 in. diameter.
 200
Chapter 4: Manufacturing Processes and Design

Length of Length of Length of

engagement engagement of engagement
of the pulley the bearings of the gear

1.0033 gi, 9.9992 gis, 0.9975 gi,

1.0025 0.9984 0.9963

Figure 4-50 Limit dimensions for the shaft of Example 4-1 (not to scale).

From Table 4-1 the grade of machine work for the shaft is as follows:
For the part which mounts the gear: grade 8
For the part which mounts the bearing: grade 7
For the part which mounts the pulley: grade 7
From Figure 4-48, it may be seen that the shaft can be turned (with care)
by a lathe operation. Similarly, we may note from Table 4-1 the grade of
machine work required to make the holes in the gear, bearings, and pulley.
Thus, the gear hole requires grade 9 machine work, the bearing hole requires
grade 8 machine work, and the hole in the pulley requires grade 8 machine
work. Referring to Figure 4-48, we find that it is possible to bore or ream the
holes in the gear, pulley, and bearings. @

Example 4-1 requires the fabrication of a shaft having three different

diameters. Such a shaft is more costly to make than one that has only one
diameter. The primary reasons for the increased cost are that the lathe must
be stopped, axial measurements made, the cutting tool readjusted, and
additional measurements are required for checking. Inaddition, the transition
corner between diameters of the V-belt pulley and the flange bearing requires
the corner radius to be such that the bearing flange is in contact with the
shaft shoulder. This requirement also adds to the cost. Thus, we see that the
fabrication cost of the shaft can be reduced by using a uniform diameter
along all of its length. (The reader should realize that in practice a uniform
diameter shaft is not always feasible; the material presented here is for
illustrative purposes.) To attain our objective, we will use the basic shaft
system.
From Example 4-1, we note that the smallest shaft tolerance is 0.0008 in.
for both the flange bearings and pulley. Because in the basic shaft system the
largest diameter shaft is the basic size, the limit dimensions for the shaft
become 1.0000/0.9992 in. diameter. To establish the hole diameter for the
gear, flange bearings, and the pulley, it is necessary to know the allowance
and maximum limit for each of these parts. The minimum diameter is
 201
Section 4-9: Selecting Tolerances

established by adding the allowance to the basic size (that is, 1.000 in. in this
example). The maximum diameter is established by adding the tolerance tothe
minimum diameter.

Gear Flange Bearings Pulley

Minimum 1.0000 + 0.0025 = 1.0025 1.0000 + 0.0008 = 1.0008 1.0000 + (—0.0033) = 0.9967

Maximum 1.0025 + 0.0020 = 1.0045 1.0008 + 0.0012 = 1.0020 0.9967 + 0.0013 = 0.9980

Referring to Table 4-1 and Figure 4-48, we find that all of the shaft must
now be machined by the same methods and to the same grade required by
the dimensions determined from the basic hole system. The reader should
verify the conclusions.
To minimize the necessity for maintaining an inventory of all sizes of bar
stock and to limit the number of tools and gages used, the designer should
try to select basic sizes from Table 4-2 and tolerances and allowances from
Table 4-3. For example, if the diameter of a bar is determined to be 32 in.
use | in. if possible. Along this same line of reasoning, select, if possible, a
tolerance or allowance from Table 4-3 that can be gaged with tools stocked
in the shop rather than selecting a dimension that would require new gages
or tooling.

Table 4-2 Preferred Basic Sizes (fractional and decimal) (continued next
page)

2 0.0100 ES C 2.2500 205

z 0.0125 = 22 2.3750 a
x1 0.015625 = Ps1 2.5000 25
fi 0.0200 A - 2.6250 2
ms 0.0250 2 2e 2.7500 275
*
pee 0,03125 J y
all
2.8750
2) =

a 0.0400 0.04 3 3.0000 3.0

= ae 0.0500 = 34 3.2500 3.25
= = 0.06 oe 3.5000 3.5
a4 0.0625i = 343 3.7500 3075
S 0.0800 -s 4 4.0000 4.0
43 0.09375 a 4i ;
4.2500 4.255
= 0.1000 0.10 u 4.5000 4.5
tl 0.1250 e 3
ni 4.7500 4.75
# 2 0.15 5 5.0000 5.0
 202
Chapter 4: Manufacturing Processes and Design

Table 4-2 (continued)

35 0.15625 7 it
r 5.2500 5.25
2 0.1875 a i 5.5000 535

= - 0.20 3 5.7500 575

A 0.2500 0.25 6 6.0000 6.0
© 2 0.30 ot 6.5000 6.5
216 0.3125 2 7 7.0006 7.0
= 0.35 ie 7.5000 7.5
; 0.3750 a 8 8.0000 8.0
- a 0.40 8s 8.5000 8.5
ne 0.4375 " 9 9.0000 9.0
< 0.5000 0.50 ot 9.5000 9.5
7c 0.5625 = 10 10.0000 10.0
ie Y 0.60 104 10.5000 10.5
2 0.6250 = Wl 11.0000 11.0
ac 0.6875 = ii 11.5000 11.5
E e 0.70 12 12.0000 12.0
3 0.7500 0.75 125 12.5000 12.5
= = 0.80 13 13.0000 13.0
“ 0.8750 e 134 13.5000 igs
E 2 0.90 14 14,0000 14.0
| 1.0000 1.0 14h 14.5000 14.5
_ is 1 15 15.0000 15.0
IF 1.1250 = = 15.5000 15.5
2 Ss = 16 16.0000 16.0
4 1.2500 1.25 @ 16.5000 16.5
Es 1.3750 = 7 17.0000 17.0

Z . 1.40 ve 17.5000 17.5

51 y
1.5000 1.50 8 18.0000 18.0
ie 1.6250 2 igh 18.5000 18.5
3
3 1.7500 1.75 19 19.0000 19.0
i7 ae
1.8750 = 195 19.5000 19.5
2 2.0000 2.0 20 20.0000 20.0
oF1 2.1250_ = 201 20.5000 2
21 21.0000 J
SSS

SOURCE: ASME: ANSI Standard B4.1—1967.

 203
Section 4-10: Vague and Superfluous Dimensions

Table 4-3 Preferred Series for

Tolerances and Allowances
of All Shafts and Holes
(dimensions in thousandths
of an inch)

0.1 I 10 100
= ka 125
ONS ta a =
= IG ie 160
= 18 18 =
0.2 2 20 200
= DD WD =
O25 O85 ow 250
= 2.8 28 =
OS 2 30 =
- 355) 35 -
QA A 40 =
= 4.5 45 <
OS 6 50 =
Oo 6 60 =
0.7 70 =
oO co (ea)S |
YI
co
10

SOURCE: ASME: ANSI Standard B4.1-1967.

SECTION 4-10
Vague and Superfluous Dimensions
As indicated earlier in this chapter, an engineering drawing 1s a document
that provides the shop with the information necessary for the fabrication of a
part or parts. A drawing, therefore, should not contain any vague or super-
fluous data. Vague or superfluous data prevent the shop from making the
component because information is either unclear or lacking, but too much
information can cause difficulty because it provides the shop with a choice. In
fact, lack of information is not as damaging as too much information. With
too little information, the part cannot be made. With too much information,
the part can be made and wasted because of the choice given the shop.
3.500
2.480

1.000 2.000 0.500

0.995 1.990 0.495

Figure 4-51 Superfluous dimensionin;

B EG
 204
Chapter 4: Manufacturing Processes and Design

Figure 4-51 provides an example of superfluous dimensioning. Assume that

the shop chooses to make the part so that length AD is 3.500 in., length AB
is 0.995 in., and length BC is 1.990 in. The result is that length CD is 0.515 in.
and exceeds the required limit. Similarly, if the choice was to make AD
3.500 in., AB 0.995 in., and CD 0.495 in., the length of BC would be 2.010 in.
and would exceed the allowable limit. The solution to this dilemma is to
leave out one of the dimensions. The selection of the dimension to be omitted
depends upon the functional importance of each of the dimensions. If for a
particular reason (for example, information or clarity), all the dimensions
are to remain as in Figure 4-51, then one dimension should be marked
‘‘Ref.”’!° and the tolerance removed.

SECTION 4-11

Accumulation and Nonaccumulation of

Tolerances
Again referring to Figure 4-51, we see surface C can be displaced from
surface A by a dimension that can vary between 2.985 and 3.000 in. If it is
important that C be held to a smaller variational distance from A, the method
of dimensioning must be changed. Figure 4-51 represents a system called
chain dimensioning where the tolerances accumulate (that is, cumulative
tolerances). Itis an acceptable method if, for example, length BC is important
because of assembly requirements. However, if there is no valid reason for
chain dimensioning, this system should be avoided. Instead, a nonaccumula-
tive method (that is, noncumulative tolerances) should be employed. This
point is illustrated in Figure 4-52.
Notice in Figure 4-52 that the location of surface C varies less from datum
surface A than it does in Figure 4-51 where chain dimensioning is used.
3.500
3.480

3.000 0.500
2.990 Ref.
1.000
0.995

Figure 4-52 Nonaccumulating

B . Cc tolerances.

A D

' Ref. means “‘reference only” and that the dimension so marked is not be be used for manu-
facturing purposes.
 205
Section 4-13: Statistical Determination of Tolerances

Datum surfaces such as A and E (that is, for vertical dimensions) are also
very convenient for shop work because many machines are set up for
cartesian coordinate locations. It is also convenient to use the center lines of
a hole as the origin of a cartesian coordinate dimensioning system. In this
instance, because the center of the hole does not exist physically, actual shop
measurements are made by putting a small tolerance, ground pin in the hole
and measuring from the pin diameter to the points ofinterest.

SECTION 4-12
Tolerance Stackup
It is possible for an accumulation (that is, stackup) of tolerances to cause
an inoperable or malfunctioning assembly. As a simple illustration of this
point, consider Figure 4-53. This figure shows a bolted flanged joint connec-
tion. This particular bolted joint uses a shoulder screw rather than the con-
ventional stud or bolt in order to limit the gasket compression. Figure 4-53a
shows the connection with the limit dimensions for each part. Figure 4-53b
shows the bolted joint with all the parts at a maximum and the length of the
shoulder bolt at a minimum. Figure 4-53c shows the bolted joint with all of
the parts at a minimum and the length of the shoulder bolt at a maximum.
Note, in Figure 4-53c, that the result of the “stacked-up”’ tolerances of the
assembly will allow leakage from the vessel past the gasket seal.

SECTION 4-13
Statistical Determination of Tolerances
In designing for quantity production, the cost of the final product will be
strongly influenced by the amount of inspection required. Obviously, the
greatest savings are realized where there is little or no inspection.
Considering the less than perfect condition of production equipment that
is frequently used, some component parts are bound to be larger and some
smaller than specified. Consequently, as described in the previous section,
some parts and/or assemblies cannot be functionally assembled. "?
The method used in the previous section illustrating tolerance stackup if
applied to a production run would be expressed as
(Pe Tg Tsih Soe eb: (4-1)
where 7is the total tolerance accumulation at assembly, and 7,, T>,... T,,
are the tolerances of each of the assembled parts.
Equation (4-1) is useful when one or two (or very few) assemblies are to be
made. However, for quantity production (without complete inspection)

'1 In some production shops, parts that are greater than and less than dimensionally specified
are used to complete some additional assemblies, thereby reducing waste.
 207
Section 4-14: Geometric Dimensioning and Tolerancing

Equation (4-1) is not useful because it implies that the largest or the smallest
assembly is just as likely to be attained as the intermediate size assembly.
Statistical analysis does not bear this out. Instead, results show the dimensions
of the components will follow a normal distribution curve. Consequently,
the frequency distribution curve of the dimensions of the final assembly will
also be a normal curve. Assembly dimensions, thus, can fall on either side of
the same mean value.
Since it is not likely that the extreme bounds ofthis curve will be satisfied,
designers have agreed to define a bilateral tolerance (natural spread) as being
equal to 3 times the standard deviation (3c) on either side of the mean.
It can be shown that the standard deviation o of the sum of parts assembled
externally is expressed as

o=jf/or+o3+o3+-..4+07 (4-2)

where o,, = the standard deviation of the nth part.

Also, from statistics, it can be shown that the expression for the total
tolerance accumulation at assembly is

Th Tee IG eee ee ern (4-3)

where 7 = total tolerance accumulated at assembly in inches and 7, =

tolerance of the nth part in inches.

SECTION 4-14
Geometric Dimensioning and Tolerancing
On paper, the designer or draftsman draws geometrically perfect shapes,
which are never achieved in actual production. Thus, a part can be made
that satisfies the drawing dimensions and yet is not functional due to
geometric variations. The characteristics of form or shape such as straight-
ness, flatness, parallelism, squareness, angular displacement, symmetry,
concentricity, roundness, and eccentricity can have an adverse effect on the
assembly and function of component parts. These characteristic features are
concerned with the tolerances of form and position, and are covered by the
ANSI Standard Y14.5—1966'* entitled ‘“‘Dimensioning and tolerances for
engineering drawing.”
Space limitations prevent further discussion of geometric dimensioning.
The reader is directed to refer to the aforementioned ANSI standard. Also,
commercial and governmental standards generally contain a complete
description of the rules and symbolisms used in geometric dimensioning.

12 Parts of this standard have also been approved or adopted by the International Standards
Organization (ISO R129 and R406) and by Military Standard 8C.
 208
Chapter 4: Manufacturing Processes and Design

SECTION 4-15
Surface Quality
Most manufactured parts do not require any special quality of surface
finish other than that obtained by the method of fabrication. However, there
are other component parts (such as bearings, pistons, cylinders, gears,
machine feed rolls, and so on) and certain fit requirements (for example,
running and sliding fits, interference fits, and so on) that make it necessary
to specify the quality of surface finish required. Furthermore, because there
is a close relationship between surface quality (that is, roughness) and the
endurance limit of metals (see Figure 3-28), it becomes imperative that a
designer carefully consider the surface finish required and the manner in
which it is to be achieved.
The definitions and symbols used for specifying surface quality are stated
and detailed in the ANSI Standard B46.1—1962 entitled “Surface texture.”
Figure 4-54 shows the meaning of the surface finish terms and the related

Roughness width Lay direction

Roughness-width cutoff
Waviness
Roughness height width

Waviness height Waviness width

ae Fe earevoit cutoff
.002-2
Roughness height Ta Lay
(arithmetical average)
—— 63 /, 020—<#——_ Roughness width

Figure 4-54 Surface texture definitions. [Courtesy of ASME,

ANSI Standard B46.1-1962.]}

symbol terminology used on engineering drawings. The terms used in this

figure are defined as follows:
Roughness. The finely spaced surface irregularities that are caused by
cutting edges and machine tool feeds. These surface irregularities are
indicated by the height, width and direction of the overall surface pattern.
The height of the irregularities is expressed in microinches (that is, 1 pin. =
0.000001 in.) sometimes called mu inch. The width of the irregularity is the
maximum permissible spacing between repetitive units of the surface pattern
and is stated in microinches. The roughness width is also used to establish the
roughness width cutoff value, and is always greater than the roughness width
unless otherwise specified.
 209
Section 4-15: Surface Quality

Roughness Width Cutoff. The maximum width of surface irregularities that

is included in the measurement of roughness height, measured in inches.
Waviness. Irregularities of the nominal surface that are of greater spacing
than roughness. The height is given in inches as the distance from peak to
valley between waves. The width is given in inches between adjacent waves.
Waviness may result from machine deflection, work deflection, vibration,
chatter, heat treatment, or warping. Roughness may be considered as super-
imposed on a wavy surface. The ANSI Standard B46.1—1962 lists commonly
used waviness height values.
Lay. The direction of the predominant surface pattern produced by tool
marks or grains of the surface determined, ordinarily, by the production
method used. Symbols used to indicate the direction of lay are shown in
Figure 4-55. The ANSI Standard B46.1—1962 specifies commonly used values
for roughness height and lay, depending upon the type of machined surface.
Roughness (in microinches) values should be based upon an arithmetic
average that is the average deviation of the surface from the mean. However,

Figure 4-55 Lay notations with surface texture symbols. [Courtesy of ASME,
ANSI Standard B46.1-1962.]

LAY SYMBOLS
Lay Symbol Designation Example

La arallel to the line represent- Direction of

Y le P tool marks
|| ing the surface to which the symbol
is applied.

Lay perpendicular to the line Direction of

tool marks
| representing the surface to
which symbol is applied.

Lay angular in both directions to Direction of

ye line representing the surface to tool marks
which symbol is applied.

M Lay multidirectional

Lay approximately circular relative

( to the center of the surface to which
the symbol is applied.

Lay approximately radial relative

R to the center of the surface to
which thesymbol is applied.
 210
Chapter 4: Manufacturing Processes and Design

because of earlier types of instrumentation,’ > roughness was specified by

the root mean square (rms) average height.
As indicated in the ANSI Standard B46.1—1962, the root mean square
average is theoretically 11 °% higher than the measurenients obtained for the
arithmetical average. Since readings from point to point on the same
machined surface would most likely vary more than 11%, industry has
adopted arithmetical average ratings without changing the ratings on older
and previous drawings. In fact, surface roughness versus machining charts
(and other literature) still specify roughness ratings in terms of rms. However,
the designer should adopt the latest standard, which uses the arithmetic
average.

Drill 31/64, depth 9/16

Drill Q (0.332), depth 1-1/8
Tap 3/8-24 NF, depth 1
Ream 0.500 + 0.001 ¥ , depth 1/2

Roughness i teheight
3.500 + 0.005
(height) lye
Roughness width
|0.062 ae (in.)
t4J 63 Bree
fa : ;
a —-— Lay direction
J
(circumferential)

Bore 2.00 I meres pont

(u in.)

Figure 4-56 Typical applications of the surface roughness. [From Kent's

Mechanical Engineer's Handbook, 12th ed., John Wiley & Sons, Inc.,
New York, 1950.]

The symbol used to specify roughness and the manner of its application is
shown in Figures 4-54 and 4-56.
The quality of surface finish of a material is directly related to its method
of manufacture and the tolerance demanded by the designer. It is apparent
that smaller tolerances require finer finishes (hence higher costs) and that
the design of a part should be such that its method of manufacture is
compatible with the surface finish required.
By means of Table 4-1, Figure 4-48, and Figure 4-57, the designer can
relate a production process with an expected tolerance and surface finish.
However, the characteristic values determined from Table 4-1 and Figures
4-48 and 4-57 should be used only as a guide. The designer should consider
each case on its own merit. For example, from Table 4-1, a 1.5-in. nominal
diameter shaft with a standard diameter tolerance of 0.004 in. indicates that

‘3 Measuring instruments that read directly the arithmetic average on a face dial are now readily
available and in use (e.g., Profilometer and Brush Analyzer).
 211
Section 4-15: Surface Quality

Figure 4-57 Surface roughness produced by common production methods.

[Courtesy of ASME, ANSI Standard B46.1—-1961.]

Roughness Height, Lt in.

Process 2000 sud es
500
= 12555 6355327 16 8 4 2
Flame cutting
Snagging =
Sane =
Planing, shaping

Drilling et I lea
Chemical milling re —
Elect, discharge mach. a
Milling |} le
- =a 4
Broaching
Reaming
Boring, turning heen | eae |
Barrel finishing |
Electrolytic grinding + |
Roller burnishing
Grinding
Honing

Polishing "| | ]
Lapping
Superfinishing

Sand casting
Hot rolling
Forging
Perm. mold casting

Investment casting
Extruding
Cold rolling, drawing
Die Casting

Key

GER Average application

(7) Less frequent Application

The ranges shown above are typical of the processes listed.

Higher or lower values may be obtained under special conditions.

a grade 10 machining operation is required. Reference to Figure 4-48

indicates that the shaft diameter can readily meet the tolerance requirements
by turning on a lathe. However, Figure 4-57 shows that for average applica-
tions, turning can produce a range of surface finishes as fine as 16 rms or as
rough as 250 rms. What finish shall the designer specify? A quick glance at
Figure 4-58 shows that a 16rms finish will cost about 2.65 times the pro-
duction cost of a 250 rms finish. Thus, the designer must consider the partic-
ular application before assigning a surface finish if he is not to increase costs
unnecessarily.
 212
Chapter 4: Manufacturing Processes and Design

Average machine

finishes

age
|

tas
Average Critical

ag ara
Noncritical

0.00003
0.00005
0.00008
0.0001
Ve
0.0002
0.0003
y
0.0005
0.0008
ie
0.001
9 0.002
0.003
: 0.005
:
0.008
i 0.010
0.013
0.016
iy
0.020
0.025
°
0.032
0.040
‘
0.050
of
Relative
units
in
required
time
cost, : 0.063
0.080
typical
Applicable
in.
tolerances,
unilateral
dimensional
0.100
2
0.125
0.160
0.200
0.250
ise) N o foe) vv N a jo)
o Oo =!
500 250 125
2000 1000
Surface finish, in., rms

Figure 4-58 Relative cost versus surface finish.

As a guide to what surface finish should be specified for certain applications,

the designer can refer to information such as that given in Figure 4-59,
industrial or governmental standards, or his own accumulated knowledge
of an industry and its requirements.

SECTION 4-16
Design Hints
Without any doubt, cost is virtually the single most important factor in
determining the ultimate success or failure of a product. No one (except,
perhaps, the obstinate designer) will argue the fact that “the best mousetrap
in the world” is unsaleable when it is too expensive. There are, however, a
few exceptions to this viewpoint (for example, times of national emergency,
the moon shots, and so on). Nevertheless, even in these instances, the cost
 213
Section 4-16: Design Hints

Figure 4-59 Characteristic maximum surface roughness finishes for some

typical machine components. [From Roger W. Bolz: Production Processes—Their
Influence on Design. The Industrial Press, New York, 1956.]

rote Surface or Part

Maximum Roughnessf, mu in., rms

°
|
50 63

HH
t+
Valve stems (automotive)
Cylinder bores (automotive) *
Piston pins
He
Seats for antifriction brg. races
Crankpins
Pistons
Cam lobes (automotive)
Rod bushings (automotive)
Cylinder bores (O-rings or leather packings) *
Surfaces for copper gaskets
Surfaces for soft gaskets
Surfaces of fluid seals—sliding or rubbing’
Ground screw threads and worms
Pressure - lubricated bearings
Piston rods (O-rings or leather packings) *
Housing fits (no gaskets or seal)

Hi
-
Slide ways and gibs
Teeth of ratchets and pawls
Valve seats
aa
Gear teeth (heavy loads)
Gear teeth (ordinary service over 10 D.P.)
Gear teeth (ordinary service under 10 D.P.)
Journal bearings (general)
Journal bearings (precision)
Rolled threads
Die or top-cut threads
Clearance surfaces (machined) |
a

a
|_ Datum surfaces (for tolerances under 0.001-in.) IL a

re
Datum surfaces (for tolerances over 0.001-in.)

rH
Push fits
Milled threads
Sliding surfaces of mating mechanisms or parts—general
Sliding surfaces of mating mechanisms or parts—precision
|
Mating surfaces—brackets, pads, faces, bases, etc. | :|
Pressed fits—general, keys and keyways | | | oe ah IL
Valve seats J : | ale al
Worm gears—general
Worm gears—heavy loadings || " ] al
Rolling surfaces—general, cams and followers, etc. | | | | |
Rolling surfaces—precision heavy-duty | Bl
Rotating surfaces—general, pivot pins and holes, etc. | 4
Rotating surfaces—precision
Friction surfaces—brake drums, clutch plates, etc. aan
Chased threads i
HH —- EXE —— a
* With regular cross-hatch pattern
t Smoothest possible finish
+ Waviness not considered
 214
Chapter 4: Manufacturing Processes and Design

factor is not wholly ignored. Thus, without intending to be facetious, a good

designer is one who practices the art of “penny pinching” as part of his
overall design thinking.
Penny pinching must be tempered, naturally, in accordance with the
objectives of a particular application. It does establish the basis upon which
a designer may satisfy functional requirements while being cognizant of
tolerances, finishes, manufacturing methods, quantity required, material
choice, availability, and so on, and the effect of these factors upon costs.
In this one section, it is not possible to discuss and detail an extensive
number of design hints and their effect on function and cost. However, the
following suggestions are offered for the reader’s consideration. It is the
authors’ hope that they will be of assistance in establishing the proper
‘design attitude.” These recommendations and suggestions are not presented
in order of importance, but are classified as general and specific. The general
recommendations primarily deal with factors that overtly affect cost. The
specific recommendations are mainly concerned with improving strength
and/or rigidity.

General recommendations

1. “If you can buy it, don’t make it!” It seems obvious that, if one could
purchase standard components which could be assembled in “‘erector set”
fashion and satisfied design objectives, a device or machine of minimum cost
would be achieved. The reader will certainly agree that it would not be
sensible to design his own ball bearing or chain and sprocket when they are
readily available in a wide variety of sizes and capacities. Yet many designers
will design a gear, clutch, brake, and so on, because they claim no supplier
has exactly what they need. This is an expensive attitude; it would be a great
deal more practical for the designer to consider design changes which would
allow him to use commercially available components. A thorough investi-
gation on the part of the designer into the vast number of standard com-
ponents, parts, and so on, that are available would often eliminate the
necessity of his becoming his own supplier. In those situations where a
standard component of the required size or capacity is not available and
design change is not possible, a design modification or rework of the com-
mercial component or part should be considered before embarking on an
individual design.
2. Always consider the ease with which the final product can be assembled,
disassembled, and maintained. For example, items such as wrench clearance,
bolt location, lubrication points, and so on, can be the source of endless
difficulty and criticism once the product “hits the market place.’’ As a further
illustration of this point, consider the embarrassment of having designed and
built a sophisticated machine, which is ready for delivery to the customer,
only to find that there are no lifting lugs or openings in the base for rigging
equipment. Thus, the machine must remain just where it is—right on the
shop assembly floor.
 ZID
Section 4-16: Design Hints

3. One should design parts that are simple in shape. Whether a component
is to be made by a primary or secondary manufacturing process, the cost can
rise rapidly with design complexity. Often, it is possible to design a much less
expensive and simpler part(s) (or redesign an existing part) by making the
more complicated part into an assembly of two or more parts. This idea is
aptly illustrated in Figure 4-60.

he aeoa 1A 18 NEF-3

WITT
LY
16

rome
ile625 Oo
| Original design

Sa

Sy

OLN
1agg"= LE) 1 4g *000
“"— 001
New design

Figure 4-60 Note that in the original design, the shaft is made of one
piece of metal having three diameters (exclusive of the threads). The
flange is welded directly to the shaft. This not only can lead to the
shaft warping but it takes 0.77 man-hr to make. The new design requires
only 0.44 man-hr to make. However, the shaft is subject to the effect of
welding heat and can be centerless ground since there is only one
diameter. [From Roger W. Bolz: Production Processes—Their Influence
on Design. The Industrial Press, New York, 1956.]

4. The designer should carefully consider the method of manufacture to

be used for the most economical production. This requires careful considera-
tion of shop facilities, the quantity to be produced, and the material to be
used. Of course, with limited shop facilities and small production, the
designer may not have too much choice. On the other hand, for quantity
production but limited shop facilities, it is worthwhile for the designer to
consider having the part(s) made in outside facilities. A serious consideration
affecting choice of manufacturing method is how rapidly the tooling costs
can be amortized over the quantity produced.
Proper selection of a production method will usually result in less scrap
and fewer rejects. It is also wise for the designer to consider a fabricating
operation in which several operations can be performed by one machine or
setup, thereby reducing setup and production time (for example, turret
 216
Chapter 4: Manufacturing Processes and Design

Figure 4-61 An aluminum valve stem body machined from

extruded rod (A) used 460 lb of material to make 1000 bodies.
By changing to an aluminum forging (B), the amount of material
used to produce 1000 bodies (C) was reduced to 220 Ib.
(Courtesy of Scovill Manufacturing Company. ]

Figure 4-62 A valve stem fora

railroad car tank was originally machined
from a Monel sand casting at a cost of
$18.75 each for quantities of 2000. The
valve stem was then redesigned as an
investment casting, reducing the cost to
$9.00 each per 2000 produced. [Courtesy
of International Nickel Company. ]

lathes, screw machines, and so on). Figures 4-61 and 4-62 show typical
examples of the savings that can be obtained by using proper production
methods.
5. The observation that the cost of different materials varies is a trivial one
but, from the viewpoint of machinability or formability, material selection
becomes important. For example, Figure 4-63 shows the relative machining
time for several different metals. Although the lighter alloys are more
expensive than the ferrous metals, the obvious savings in machining time
(that is, labor savings) can be of such significance as to offset the higher
material cost.
When one has little choice in material selection (for example, with high
temperature problems), cost reductions must be sought in the area of
manufacturing methods.
 217,
Section 4-16: Design Hints

Magnesium alloys

Aluminum alloys

Brass Figure 4-63 Chart comparing the

relative machining time of several
Cast iron different metals.

Steel
(0) 1 2 3 4 5 6
Relative machining time

6. Try to design simple parts. Obviously, simple parts will require fewer
manufacturing operations which, in turn, will result in lower costs. In
addition, savings will be realized as a result of simpler and fewer production
tools being needed. The importance of minimizing manufacturing operations
by virtue of simple design is emphasized by reference [3], which states, “‘the
greatest saving which can be obtained in designing for production usually
results from reducing the number of separate processing operations required
to complete a part.”’* An example of this principle is exemplified in Figure
4-64, in which the number of operations to produce the different parts is
reduced from 45 to 21. Also, the number of parts was reduced from 14 to 10
while the cylinder was made 40% lighter. The final cost of production was
reduced by approximately 600%.

OT a ae cee TH
RI
y, | | n° 55):
kIT pith
c paca
eal 1 hess
xX?L0 ee
iA 7

en
ESOC ra AX la 2EKX Original design
TX XOK Pe ccs oe pagel
WKKRY

i Bo
| eaeYZSZSZSZV NPS pend | New des ign

Figure 4-64 Original and new design of hydraulic actuating cylinder for the
extension or retraction of an airplane tail wheel. Note the greater simplicity of
the new design. [Courtesy of Electrol Inc.]

7. Do not overlook the manner in which a part to be processed is to be

held in order that it may be “worked on”’ by some machine or cutting tool.
The designer should consider providing shoulders, centers for turning,
bosses, flats, holes, lugs, and so on, which may be needed for properly holding
a work piece in a jig, fixture, clamp, or machine. Magnetic chucks and holding
devices should be considered for magnetic materials that are awkwardly
shaped or otherwise difficult to hold in place.

14 “Next in importance is reduction in the number of parts used and following that comes
savings effected through the use of stock parts or interchangeable parts.”
 japaa4 Japaay

218
sO_ 4aSii so _ 4asis

ULLYS
Q peed
paya

39944/09U] 39341109) 1991409uU| 41991109

Y,
fe
SUO!1IaSPjNOYUs
aq 4O WAOJIUN “SS9UHIIY}
B4IBYM je007 Jesnjonsys j2907 uLIYs oof a61e) jaj|i} Sasned
Bulpualq
4o 1461)
0} AABay suO!}9as JOuUeD
aq ‘paplone ssauyeam panosduy
uBisap
ssauxeam yeam jeJaW a4NONsIs
ay} suoljo—as pjnoys
aq pazed0|
Je ay} do} 10} 49}18q
400g uBisap 40 abeyulsysJa}ap
jeJaw pas} pue “UOIINGI4}SIP
(P)

atioe|
[an Pesnot/) asneoag 194914) SUuO!}DaS
aney
,,.ods walle samoyjs 6uljooo ‘sawi} poob
| aes }| es Buijseo uBisap Sasinbas
EY} JJEM
ees a suo!jo—as
aq AapeW
OU 424914}
ueyy yj6uasjs sjuawasinbas
pue
yey ay} saquinu
yo Bululofpe

e
(e)
suo!oas
aqe “WNUWIUIW

1991109
ay} Ue} $saj aq Ajqesajaud pynous sped pue sassog $o ssauxdIY|
Say

ya1J09u|
4991109U|
0} yBnoua 4914} Ing ‘uIofpe Aay} UOI}DaS Hul]sed ay} JO ssaUxDIY)
"yyem Buijses ay} Buryono} ynouyyM Buluiyoew yuw4ad
| (see

4991109
pasabb
sqi4
eis BoNpas UO!J0}SIP
pasneo
Aq abexyuls
pueiys -uadu0d
UO!}e4}
YO ‘jeJaWpue aZIWIUIW
}Oa1J09U] 4994109 }0u,, $10ds
,,
(2)
pajeuiwija aq pjnoyus snduapuA ule1gO 0} sassog apis}no 1iWC

1991109,

49944109Uj
‘ajqissod 4aAa1ayM “‘yeup iybiesys
(a) (9)
 sjuasaidas
‘uBisap

poob

(4)

Jey}
‘Wid

qny

219
pue

ay}
+o

Ajseau asow aje4 Buljood e Buiney ‘U0!}9aS JUAIDI}yNS JO SW "qnYy

pue sayods ay} ul $9940} UOIssaidw0d Bulpuodsasi09 YyyIM ‘sule 43s
UOISUAI YIM 14a] Aq ||IM WIJ ay} ‘SUO!}DaS UBAGUN ay} $0 Buljoo9
pue UO!IaUUOD paulejyseal 0} aNq “3se| SalyIp!}Os painod uayM pue Vl
‘sayods ay} ueU) JalAeay pauBlsap si! (\¥) |eaYMA} ay2 JO WH BY BHuipaaoxa jade} e aAeY JOU PjNOYs ssaUxD!Y2 ||EM U!
sabueud padeus abpan “| :Z Peaoxe JOU pjNoYs UO!19aS
1981109U| , ae, yee teoul Bululofpe yo ssauyxo142 aYI U! BdUaJa}}Ip B42 ‘Guin 4o
\ g]n4 e S\v *SUO!}e1]UG9U09 ssauzs pue stods oY Bulzjnse.
(Vv) WO}IUN-UOU
Buljooo
pue JUAAAIA |]IMeII
SUO!IISU
yjOowS ‘sabueyd uoljoas Jdnige pue sajbue dieus pioay
poop se 4 poo peg
7 cane) Ce Say
“Burjseo ay} jeauue O} asim S! }} ‘Uled}s
jewiullW ainsse OJ “sazis UO!}OaS Bulia}yip

3991109U|
a0 7
Ajapim pue $asoo 40 Az191)d13;NwW ‘sajbue JUes}Ua-a)

Figure 4-65 Hints for good casting design. [From Ol 1ver

(NN
‘ww404 40 sabueyd uappns plone pjnoys ubisap 1991109

Smalley: How design is affected by foundry practice. Prod.

Buljseo ay} ‘Uledys AZIWNJUIW O] “Ule4}s Bulysed yo
asned UOWWOD }sOWW 34} S$! BuIlsed e yO syed SS
Vay
Juasasjip Buowe sajze1 Huljood ul uoleueA
Oy

Eng., 117-121 (Feb. 1950).]

 220
Chapter 4: Manufacturing Processes and Design

8. Do not specify tolerances and/or surface qualities more restrictive than

those minimally required for the part to perform functionally. Indeed, the
attitude of the designer should be one of constant questioning of the need for
narrow tolerances or superfine finishes. This attitude is justified in the light
of Figure 4-47 and by the further savings that may be realized from reduced
assembly time, greater interchangeability, and less scrap and waste.

Specific recommendations
Correct design procedure requires constant awareness on the part of the
designer of the fabricating methods associated with making a part or
component. This awareness contributes greatly to the ease with which a part
can be fabricated as well as aiding in the reduction of the fabricating costs
and avoiding factors that may contribute to a reduction of strength or
stiffness.
1. Casting design is as much an art as a science and is so specialized that
the average designer cannot be expected to be an expert. In many cases, it
would be wise to consult with a foundry man or patternmaker before
proceeding with a particularly complex design. However, if the designer
observes a few simple guidelines, he will usually produce a sound casting.
Figure 4-65 illustrates some basic considerations for fabricating good
castings.
For greater detail, the reader is directed to the Steel Castings Handbook
published by the Steel Founders’ Society of America, to the publications of
the Gray Iron Founders’ Society, and to reference [2].
2. Forging design, as in the case of castings, requires specialized knowl-
edge, and expert advice should be sought before releasing the design to the
die shop. However, by adhering to some fundamental requirements, the
designer can contribute much toward the production of a sound forging.
Some of the important conditions he must consider are (1) the draft angle,
(2) the location ofthe parting line and forging plane, (3) the fillet and corner
radii, (4) the rib heights and rib section thicknesses, and (5) the metal grain
flow.
The normal draft angle for ferrous metals is 7 deg for outside surfaces and
10 deg for inside surfaces. Aluminum and magnesium can be forged with
draft angles of 5 deg or less, depending upon the shape of the part to be
forged (see Figure 4-66).
If possible, a straight parting line should be used so that one-half of the
die has all the impressions of the part to be shaped. On the other hand, if
other than a straight parting line is required, the forging should be inclined
with respect to the forging plane (see Figure 4-66).
Fillets and radii should be as large as possible, whereas rib heights should
be as small as possible. However, rib widths should be generous. Recom-
mended sizes for fillets, radii, rib heights, and widths can be found in
reference [7], or in the Tool Engineers’ Handbook of The American Society of
Tool Engineers.
 Za)
Section 4-16: Design Hints

Locating Draft Angle Natural Draft

a7 draft Natural Si Draft must

{|angle ec be machined
|} a angle off

y—Greater than 7°
This Not these

Steel and Aluminum Draft Draft angle—Where normal drafts do not match, increase
> 5° draft in one die so both sides meet at the parting line. Us-
H a be" ing natural draft angles can minimize weight and subse-
| quent machining. Standard draft angle for forgings is 7
Forging Parting deg. U-shaped aluminum fittings should have 5 deg.
plane line
Bie =

on) End radius—Normal draft should be

H maintained on circular sections. Ends of
] forgings should have a radius equal to
1 } twice the sectional diameter.
Radius (2D)

Desirable Undesirable

Parting line — Use a

When a pala When a recess
straight parting line, where When neither exists
exists exists
possible, so one die-half
contains all the impression
to shape the part. To sim-
plify die sinking and trim- Place dil Place ED)
Place P.L. here with
ming, external webs less line here line here option for P.L. at
than 5/32 in. thick should center of web
have parting line coincide
with one web surface.

Forging plane — Should

Ei
AZZ Lilie Ui a
DpR—ORCS EE Kee Ses
other than a straight part-
ing line be required, the
forging should be inclined
with respect to the forging
plane. Dies with counter- line Die
locks to compensate for Preferred Not recommended
lock
side thrust are expensive to
build and troublesome to
maintain.

Figure 4-66 Limitations for product and die design to insure strong and sound
forgings. [From Checkpoints for practical forging design. Annual Handbook of
Product Design, Product Engineering, 1954.]
 222
Chapter 4: Manufacturing Processes and Design

Figure 4-67 Design tips which lead to easier machining and lower costs. [From
Ernest Geiger: The rights and wrongs of details. Prod. Eng., 12 : 72, 122, 148
(1941).]

Flange joint
Hole must be drilled at left involves
through to permit more difficult
knockout of pin and machining than
for easy chip removal at right

Wrong

Accessibility of
bolts speeds
assembly ——~ —

Wrong

Wrong Right

Always design—~—
an undercut for
tool runout when Clearance grooves \
threading up to or undercuts avoid =)
shoulders Nee impracticability —~ =
of machining or
grinding into
sharp corners cs
\ eo 2 i

_ ~ Difficult to lay out

Lc aid hole center and
a start drill. Flat
surface necessary

Clearance for
Be pS SSS Drill may snag when eae taps
should be checked
breaking through
 223
Section 4-16: Design Hints
Builtup Welded Constructions

Li
My 7 ==
hell“~
She Drop-forged
spud or forged boiler
pits ; > Bent and flame
H ;
H cut to size
i Se :
H pipe section
Of
SS —-Steel base plate
Flame-cut
circle <

Sas
Half coupling flame
cut to shape

~ >>~Pipe section
——-— Steel base plate
Extends inside tank
wall, allowing two weld
points for additional :
strength Builtup Supports
Builtup Inlets and Outlets

Shaft
._

plate Thrust bearing

-—
c
i "ia 44 P2722

A i Pipe section --—

i
Shell~
Q 4 a Heavy gage
p - Flanged chan- SS plate sheared
YSN a
Dished nel not welded and bent to

H
"a, ottom /
Flame — Yo 77>
to boss : / shape
LZIZZZA fume’ -- Square supports

>> Sheared GF
ribs N
‘Steel base plate

Tank Support Shaft Support Simple Tank Support

Figure 4-68 Typical welded joint designs for unfired pressure vessels. [From
G. F. Nordenhold, J. Kerr, and J. Sasso: Handbook of Mechanical Design.
McGraw-Hill Book Company, New York, 1942.]

One of the most important factors in fabricating a sound forging is to

design the die so that the grain (that is, fiber) flow of the metal to be forged
is maintained (see Figure 4-9).
For a detailed and in-depth study of forging design, the reader is directed
tothe publications of the Drop Forging Association and the Metals Handbook
of the American Society for Metals.
3. Some planning and forethought by the designer can contribute greatly
to the reduction of both the cost and any fabricating difficulties associated
 224
Chapter 4: Manufacturing Processes and Design
Welding Fittings

Flat-ended end
(h)

(i)

Figure 4-69 Typical weld tube and welded fitting connections. [From W. J.
Van Nattan: How service requirements govern design of brazed joints, Sect. G.
Annual Handbook of Product Design, Product Engineering, 1954.]

with producing a component part in the machine shop. Careful consideration

of how a part will be machined should be of primary importance. Calling
for machining operations or tool room equipment that is beyond the scope
and capabilities of the available facilities will lead to having to ‘‘farm out”’
the work. Naturally, this will lead to higher costs. Such obvious faults as
calling for drilling excessively long holes, machining after heat treatment
(unless required), machining unnecessary surfaces, not specifying grinding or
thread “‘runouts,” and so on, can readily be avoided. As the designer acquires
more skill and experience, spends more time at the drafting board and has
greater contact with the shop, his expertise will grow along with his knowledge
of good design practice.
 225
Section 4-16: Design Hints

Eo=
eet | ie
(a) Be aring

Y/CL, ;
TC AMIN

AON MOM

Z
(c)
Y (d)

Bearings May be rounded

or corners beveled

Bell Crank

Foundation Bolt Bosses for Bases

Figure 4-70 Several applications of welding for machinery. [From W. J. Van

Nattan: How service requirements govern design of brazed joints, Sect. G. Annual
Handbook of Product Design, Product Engineering, 1954.]

Figure 4-67 illustrates some of the design mistakes that can be avoided.
4. In keeping with the previous discussion, Figures 4-68, 4-69, and 4-70,
respectively, suggest ideas and hints of good practice in welding pressure
vessel joints and supports, tubes, and machine parts.
Figures 4-71, 4-72, and 4-73, respectively, offer design suggestions for
producing good brazing joints, techniques for reducing corrosion, and heat
treating considerations that will contribute to the avoidance of cracking and
warping.
 226

deq yng
uy dejsano
JO aasy) sai}ay)
ysauuiyr saquiaw
[IMA Ayjensn
Plaid %OOLde) 1UIOf -u9!oi449a
awog suoneuiquios
yo de; pue 13nq “squlof
“Ad yng siaquiaw
ae paz) | suO!IOasS
pue 19U409 sjulol
ase Pasap!suod
OF ay] SsO1D jeUO!IDAS
Ba1e$0 0} aq suoNeueA
jo 13Nq *squlol
auo “yaquiatw
Buneay uoroas Aneay-

14617.
(v7) peg
uoloas (g) pood
eas rea
Chapter 4: Manufacturing Processes and Design

f (5) poo
peo7Uo 1U/o[(v7) sasneo
e yBiy ssasispue Bursea) (Vv) peg (g) poop (9) pood
jo ay) ‘aze1q
Ag Buideys
ayy Aneay uoltoas
se ul
“(g)40 BuIseasoU!
ay2 1YB1) uo!9as
(9) aya Burseay ajdnog sasneo quiof Buea} *(\7) Bulseasouy
ayi 3y61|
si “paddojs Uo!oas
(g) 10 Bulublsapa
(Dd) s saonpa
S14} “uolq9e

SA Sar
: !

(Vv) peg (g) pood (3) Peo)

yoedwy|peo) sasneo1UIO! aanyiey“(\y) Bulseasouy

ay) '
azesq ease (g) 40 Bulsn
e yes9s 1UIO!(D) ON say
Sanoidius
ayy joedwi soisiiaioeseyo
40 ay} pulol Japun “peo
A UI! (y) |)IM jie} sapuUN BuIpeo|
Ul Jaya UO!D941p
ajIUM 4aUIO9UIO!(g) UeD axe} BUO UO!9A4IP
Peo]
pue (Dd) YIIM PaUAy}Ns
ssO1D UO!IDAS
OM} UOND941P
anbijey
ued uejs
az asay = ‘
“Buipeo}

peg poo poop

qny pue a)xe (v7) jj rey Aq aNBiyey
Je “WIO!-2Q
suBis(g) pue (9) JiwadGny0} MO}|O} AdNXA9}J
$0 au BuIVelO4
Weaq(\y) {JIM aney
e Mo] aNBIWe}
941]
ajxe asnedaq
jo ynd4sapuN
10 snosauab
qny “snipes afLYM aNaa}s
(Gg) |]IM aseas9U!
SIYI “O}!)
 [ps6] ‘Suaaui8uq Jonporg ‘usisag Jonposgfo yooqpun}y JONUUp “f) “92S
Section 4-16: Design Hints

‘sjutof pozeiq JO USISOp UIJOAOS S}USWIAIINbas dd1AIOS MOH : UPNEN UPA ‘f “A

wos] ‘(aded xou os]e 998) sjutof pazesq JOJ SUOT}BJOpIsuod usIsOq LZ-p e4nbig
227

ease azeiq sabse| e sapiAoud pue

quiof ysabuoys ay) sani6 siy
‘ajqissod uaym adA} de} ayy
JO aq pynoys s}usol aunssaig

;j/ p
“pasn s! uBisap iy} UayM pancds6
“yiBuasys jed1UeYydaW AY} a]NdWOd O} payoesIqns
sAemye ase SUOIIIAS JAIABAY AY} JEU) BION
SI PIOA “POA e Bulneay INO SMO}} BACOIB BY) UI A4IM ay
“syed 4e)N9419 Ul asim Buloejd yo spoyyaw aWOs
“suoiqiod payeauue Jo pazeiq ay} puodAag Ajjeoiueysaw
payioddns ase Aay) ‘si0yoe} ase anBijey 10 UONeuqIA a1a4yM,
ahaa|s 10 ajnsse4
am |
ease payeauu --- 7 Weuywssa|
ry Y Ppapuedsys 10 PI|OS i

Y
SNS
eae pazesig =

Z
YUMMY,
=
uonensu|. \

S999
= ease pazesg
== ease pajeauuy
 228
Chapter 4: Manufacturing Processes and Design

Atmospheric Corrosion

Moisture
Rounded Corners and smooth contours should be used when-
and dirt
collect in ever possible to prevent the accumulation of moisture, liquids and
-pockets. solid matter. Using corrosion resistant materials is often found to
be more economical due to greater service life.
Poor Good

Concentration Cell Corrosion

Bottom
end open

Scale and
precipitates
/ collect in

All Liquid Containers should be designed with smooth and rounded corners. All Baffles and internal stiffeners cs
Sharp corners, stagnant areas and other such conditions are favorable to the in tanks should have openings ar-
accumulation of precipitates, solids and scale which promote concentration- ranged to avoid liquid pockets and
cell attack. Sloping bottoms should be used with valves arranged for com- permit the free drainage of fluids.
plete drainage.

Galvanic Corrosion

Weld metal. _ Bracket LN

TD YZ jlnsuation
= =

cathodic cathodic ——a a,

PSE SSS
>Paint

Rivets, bolt and nuts cathodic metal

Deere Fair Do
In Joints and connections, the pro- When Possible the connection of Painted Coatings should be ap-
portions of dissimilar metals should be dissimilar metals should be separated plied with caution. Do not paint the less
chosen so the anodic or less noble metal by an insulating material to reduce or noble material, otherwise greatly accel-
has the greater exposed area. If fasten- prevent the current flow in the galvanic erated corrosion will occur at imperfec-
ings such as bolts and rivets are re- circuit. Paint or plastic coatings serve to tions in the coating. If possible, exposed
quired, they should be made of the more reduce the galvanic current by increas- surfaces should be painted. Commercial
noble or cathodic type material. ing the circuit resistance. protectives are now available.

Figure 4-72 Detail design considerations to reduce corrosion (see also next
page). [From F. M. Reinhart: Recommended design details to reduce corrosion,
Sect. G. Annual Handbook of Product Design, Product Engineering, 1953.]
 229
Section 4-16: Design Hints

Crevices
we el NS Aye
collect Solder, weld, peen ’
7” liquids or fill with caulk-

Poor
Better

Best Poor Good

Butt Welded Joints are less likely to corrode and Storage Tanks and other containers should be sup-
should be used. If lap joints are required, all crevices should ported on legs to allow a free circulation of air under-
be filled with a non-absorbing caulking compound or welded neath. This prevents the possibility of any condensation
to prevent retention of liquids in crevices. and collection of moisture under the tank.

Current
-
flow

Cathodic Protection of containers with corrosive liquids can

In The Design of liquid passages, all pipes and
be done by immersing a rod of a more anodic material inside.
connections should be constructed to insure uni-
This reverses the galvanic current and the container becomes
form flow, with a minimum turbulence and air
cathodic and is less likely to corrode. Magnesium and zinc
entrainment. This also reduces settling of solids.
rods are often used.

Coil
g Avoid threaded
/ (cathodic) Oversize
sory
.8 Posessrsytesssssssrssy & WY connections
if
y ieee eee Z replaceable
Y
g
Y
Y
4
4
y
Y
Y
Y
Y
Y
Y
4
Y
Y
Y
y
4
Y
Y
Y
Y
4
4
y
Y
Y
Y
g IWC
Wa
AWAVALAVETLRSTSSSSSSGSTESSSSEEEEEST'
Y esses,
~ Tank (anodic)
In Designing Equipment, keep In Joining Dissimilar Mate- At Connections of dissimilar ma-
different metals far apart in the solu- rials, well apart in the galvanic series, terials, consider using small replaceable
tion; this increases the resistance of the avoid threaded connections since the sections made of the less noble metal.
electrolytic path. Chemical inhibitors are threads deteriorate rapidly. Brazed joints These expendable parts should be easy
often added to corrosive solutions. are preferred using a brazing alloy more to replace and made oversize to increase
Some bare zinc, magnesium or steel noble than at least one of the metals their corrosive life. Nonmetallic gaskets
in the liquid will counteract corrosion. that are being joined together. increases the circuit resistance.

Figure 4-72 (continued)

 230
Chapter 4: Manufacturing Processes and Design

No e/ Small Be) Large a)

Best

Wrong Right Wrong Right

Working face

SMU

Wrong

Xi
Duplicate s
(but dummy) Right
working-face

Thin Key slot moved

section half a pitch Balanced Slots and Keyways

4 |

tt +
Wrong Right
Wrong Right Best

Hole

Deep, blind hole poor

z. for quenching

Wrong Right
Wrong Right

Figure 4-73 The ideal shape for a part to be heat treated is a shape in which
every point of any section or surface receives and gives back the same amount of
heat with the same speed. Such a shape, of course, does not exist, but it is the
designer’s task to come as near to it as possible. To do this, keep the workpiece
body simple, uniform, and symmetrical. For example, the first figure at the top
shows how changes in cross section must be made gradually to minimize stress
concentrations during heat treatment. The other figures show further specific
ways to keep out of trouble when subjecting parts to heat treatment. Holes, for
example, should be correctly located. [Courtesy Product Engineering.]
 Tis|
Problems

PROBLEMS
1. Into what general categories can manufacturing be classified?
2. List ten ways of making castings. Which of these methods (a) would you recommend
for producing large parts in quantity, (b) requires the least amount of finish
machining, (c) produces smoother surface finishes, and (d) produces the narrower
tolerances?
. What is draft and why is it required on casting patterns?
4. What are the principal types of metal alloys used to make investment castings?
. What are some of the advantages to be gained in making parts of powdered metals?
What are some of the disadvantages?
. Explain how the mechanical properties are improved by forging.
. Compare the advantages and disadvantages of open die forging, closed die forging,
and upset forging.
. List the different methods used in forming plastic products. Which of these are
most suitable for quantity production?
. Name the three types of gas welding. Which of these types is used primarily to
join nonferrous metals?
. Name the eight methods of arc welding. Of these eight methods, which is the most
widely used?
. What metals can be joined by shielded arc welding?
. What method of arc welding is used for automatic production? Name the metals
that can be joined by this method of welding.
- How is atmospheric shielding obtained in shielded arc welding? What are some
of the most important advantages gained in using this type of welding?
. Name six types of resistance welding methods. For what kind of production is
resistance welding mainly employed?
. What metals may be spot welded? Can dissimilar metals be spot welded?

. How does spot welding differ from seam welding?

. What is the difference between flash and upset welding?
. Which method of resistance welding is used to join dissimilar metals?
. What are the differences between welding and brazing?
. What are the differences between brazing and silver soldering?
. Name the three secondary production processes. Which ofthese is the most widely
used?

22. Define the following : (a) nominal size, (b) basic dimension, (c) tolerance, (d) clearance,
(e) allowance, (f) unilateral tolerance, (g) bilateral tolerance, and (h) limit dimension-
ing.
23. (a) Describe the basic hole system for assigning tolerances. (b) Name the eight
classes of fits categorized in this system. (c) What are some of the objections to
this system?
24. Explain the differences between selective and interchangeable assembly.
 oe
Chapter 4: Manufacturing Processes and Design

25, What must the designer guard against in assigning tolerances to component
parts that are to be assembled in one unit?
26. What is the primary purpose in using preferred basic sizes and preferred series
for tolerances and allowances?
27. What is geometric dimensioning and tolerancing? Why is it employed?
28. A shaft has a 24-in, diameter. Determine the shaft and hole dimensions, allowance
and maximum clearance for each ofthe listed conditions (use the basic hole system) :
(a) a close sliding fit, (b) a precision running fit, (c) a medium running fit, and (d) a
loose running fit.
29. Solve Problem 28 using the basic shaft system.
30. A number 306!> ball bearing with ABEC 1° tolerances is to be used for a revolving
shaft. The housing for the bearing’s outer ring is stationary. A bearing catalog
specifies the following dimensions: The maximum bearing bore is 1.1811 in.;
the minimum bearing bore is 1.1807 in.; the shaft diameter is to be no greater than
1.1816 in., or smaller than 1.1812 in. Determine (a) the bore and shaft tolerances,
(b) the allowance, (c) the maximum and minimum metal interference, and (d) the
class of fit.
31. The same catalog as used for Problem 30 gives the following dimensions for
mounting the outer ring in the stationary housing : maximum and minimum bearing
outer diameter, respectively, is 2.8346 and 2.8341 in.; maximum and minimum
housing bore diameter, respectively, is 2.8353 and 2.8346 in. Determine (a) the
outer diameter bearing and housing bore tolerances, (b) the allowance, (ec) the
maximum and minimum metal interference, and (d) the class fit.
32. Specify the dimensions, allowance, and maximum clearance for each ofthe following
conditions (use the basic hole system) : (a) a 5-in. diameter sleeve bearing and journal
for a vacuum cleaner, (b) a 2-in. diameter sleeve bearing and journal for a bakery
dough mixer, (c) a 14-in. diameter sleeve bearing and journal for a motor generator
shaft, (d) a g-in. diameter shaft assembled into a hub by means of an arbor press,
(e) a 3-in. diameter dowel pin in a hole in die plate, (f) an 8-in. diameter rim for a
railroad car wheel, (g) a 45 in. diameter crankpin and bearing of a diesel engine,
and (h) a 4-in. diameter shaft and sleeve bearing of a 1200 rpm gas turbine.
33. A shaft and hole have a nominal diameter of 2 in. The shaft has a tolerance of
0.003 in., the hole has a tolerance of 0.004 in., and the allowance is set at 0.001 in.
The dimensions are based on the basic hole system. The parts are made in quantity
by screw machines which give a bilateral tolerance of 0.001 in. for the pin and
0.002 in. for the hole. Determine (a) the most likely (that is, average) clearance,
and (b) the expected maximum and minimum clearance.
34. A 3-in. diameter journal and bearing are made to a close running fit. Ifit is expected
that the method of manufacture will produce a natural spread equal to 80°% of
the tolerance, determine the likely maximum and minimum clearan¢es. Compare
your answer with the allowance.
35. The piston wrist pin shown is one of the assembled components of a model airplane
engine produced in quantity. The method of manufacturing can maintain the
linear dimensions shown. (a) What is the tolerance on the overall length? (b) What

'S Chapter 9 explains these terms. However, their exact meaning is not needed here to solve
the problem: they are used to introduce the reader to their usage.
 233,
Problems

Problem Figure 35

are the maximum and minimum lengths (that is, limit dimensions) of the pin?
(c) If only ten such pins were made, what would be your answer?
36. For each ofthe tolerances required on various parts, specify a secondary production
process.

Final Size Required Tolerance, in.

(a) 2-in. O.D. 0.007

(b) 2-in. O.D. 0.0005

(c) 1/2-in. diameter hole 0.005

(d) 1/2-in. diameter hole 0.0005

(e) 3-in. thick flat plate 0.014

(f) 3-in. thick flat plate 0.0007

37. For each of the pairs (that is, a,b; c,d; e,f) of Problem 36, determine the expected
surface finish in microinches (rms) and compare the relative cost for each pair.
(Note. Do not use Figure 4-47 to solve this problem because it is included herein
for illustrative purposes only.)
38. Make an engineering sketch of a shaft showing all dimensions, tolerances, and
surface finishes. The specifications are as follows: The overall length of the shaft
is 15.750 + 0.005 in. Each end of the shaft is nominally 23 in. in diameter. One of
these ends must be turned to a free running fit 5.250 + 0.002 in. long. This same
end contains a by 14 in. long square sled-runner keyway. The other 23-in. diameter
end is 3.000 + 0.005 in. long and is to be turned to a free running fit. The middle
portion of the shaft is 2gin. in diameter and contains a 3 by 3in. long square
profiled keyway. This keyway is centered in the middle portion of the shaft. The
transition planes between the middle larger diameter and the smaller diameters
act as location shoulders for helical gears.
39. At times many of us have exclaimed ““What a terrible design. I could have done
better than that!” In this regard, select some simple item or device with which you
are familiar. Indicate what, in your opinion, are its bad features. Proceed to show
how these negative features can be improved or eliminated. Keep in mind that
safety, materials, ease of use and/or maintenance, function, and overall cost are
but some of the parameters that should influence your thinking and final decision.
 234
Chapter 4: Manufacturing Processes and Design

REFERENCES
[1] 1970 Metals Reference Issue, Machine Design. Penton Publishing Co., Cleveland,
Ohio.
[2] E. Paul De Garmo: Materials and Processes in Manufacturing, 3rd ed. The Mac-
millan Company, New York, 1971.
[3] Roger W. Bolz: Production Processes—Their Influence on Design. The Industrial
Press, New York, 1956.
[4] S. W. Bernard, E. O. Waters, and C. W. Phelps: Principles of Machine Design. The
Ronald Press Company, New York, 1955.
[5] Oliver Smalley: How design is affected by foundry practice. Prod. Eng., 117-121
(Feb., 1950).
[6] Checkpoints for practical forging design, Sect. D. Annual Handbook of Product
Design, Product Engineering . McGraw-Hill Publishing Co., New York. 1954.
[7] Ernest Geiger: The rights and wrongs of details: Prod. Eng., 12: 72, 122, 148 (1941).
[8] G.F. Nordenholt, J. Kerr, and J. Sasso: Handbook of Mechanical Design. McGraw-
Hill Book Co., New York, 1942.
[9] W. J. Van Nattan: How service requirements govern design of brazed joints,
Sect. G. Annual Handbook of Product Design, Product Engineering. McGraw-Hill
Publishing Company, New York, 1954.
[10] F. M. Reinhart: Recommended design details to reduce corrosion, Sect. G.
Annual Handbook of Product Design, Product Engineering. McGraw-Hill Publishing
Company, New York, 1953.
Analysis of Stress
and Displacement

SYMBOLS

c = outer fiber distance, in. R reaction force, lb

dw/dx = slope, rad. yp
yield point, psi
E = Young’s modulus, psi T = torque, in.-lb
F = load, Ib U strain energy, lb-in.
G = shear modulus, psi V= shear force, lb
I = moment of intertia, in.* Ww deflection, in.
J =a geometric constant; the a coefficient of thermal
polar moment of inertia for expansion, (in./in.)/F°
a circular section, in.*+ shear strain, rad.
k, torsional spring constant, displacement, in.
in.-lb/rad strain, in./in.
M = moment, in.-lb slope, rad
N = factor of safety normal stress, psi
[? = loge, No) shear stress, psi
P.,cr = critical load, lb Vr
Dm
DD
~<APoisson’s ratio

The most important criteria in machine design are that the machine or
machine element must function properly and that it be safe. In most cases,
the dimensions of machine elements are based on the strength of the material
and predicted maximum stresses. Some design dimensions are based on
maximum permissible deflections. In either case, a major difficulty arises in
selecting an analytical model to represent the actual proposed design and
the predicted system of loading. This step requires considerable engineering
judgement, since, if the model is not representative of actual conditions, any
further analysis would be meaningless.
230
 236
Chapter 5: Analysis of Stress and Displacement

SECTION 5-1
The Analytical Model
The analytical model is a compromise, attempting to idealize the system
of loading and the load-time relationship as well as the design geometry and
the materials used so that a reasonably simple solution is possible. Most en-
gineering materials may be considered homogeneous, continuous, isotropic,
and linearly elastic, although no actual material truly meets these specifica-
tions. Continuity assumes a lack of structure in a material, but a crystalline
substance may be treated as continuous, homogeneous, and isotropic if
(1) the part dimensions are large in comparison with the individual crystals
and (2) the individual crystals are distributed at random so that the properties
at a given section in the part are essentially the same as at any other section
in the part, regardless of orientation.
Steel and some other metals have a nearly constant elastic modulus for
stresses up to the proportional limit. Beyond this stress level, however, we
do not observe linear elasticity. The behavior of a ductile material becomes
plastic or elastoplastic beyond its elastic limit. Some materials, particularly
nonmetals, fail to exhibit a linear stress-strain relationship. Nevertheless,
design is based on linear relationships in almost every case. An average
value of elastic modulus for the expected load range may be used. These
assumptions are discussed in reference [1].
Most machine parts include discontinuities in section due to fasteners,
bearings, or other design considerations. For calculation of stress and
deflection, however, the geometry of the part is often simplified ; for example,
a constant cross section may be assumed, and the results corrected by using
stress concentration factors.

SECTION 5-2

The Relationship between ‘‘Exact Theory”

and the “Strength of Materials’’ Approach
There is no fine line separating the “theory of elasticity”’ (see, for example,
reference [1]) from the topics and methods of “strength of materials”’
(references [2], [3], and [4]), and the value of one approach over the other
depends on the particular application. In most engineering problems, both
methods assume a homogeneous, isotropic, linearly elastic material. Both
methods require that equilibrium of force be satisfied. To determine a three-
dimensional stress distribution using the theory of elasticity, six stress-strain
equations and six strain displacement equations are used in addition to the
three equilibrium equations. The unknowns, six stress components, six strain
components, and three displacements may then be found for given loading
and boundary conditions. If the problem is formulated in such a way that
the displacements are not explicitly included, it is necessary to establish
 27
Section 5-3: The Nature of Stress and Strain

compatibility of strains, that is, we must show the material to be continuous

in the stressed as well as the unstressed state.
An alternative problem formulation assumes a state of stress that satisfies
equilibrium of forces and corresponds to the loading and boundary condi-
tions. Compatibility of strains is not necessarily satisfied. This method, the
strength of materials approach, permits the solution of problems that would
be very unwieldy by elasticity methods. Because the stress distribution is
assumed beforehand, it is apparent that this approach would be meaningless
if we were required to find the stress concentration due to a hole in a tension
member or for any problem where we have no rational basis for assuming
a certain stress distribution.
Where the theory
elasticity
of results in a tractable formulation, the solution
is accurate to the degree of accuracy of description of loading and boundary
conditions and to the degree to which the material approaches the ideal
assumed homogeneous, isotropic solid. These conditions being met to a
reasonable degree, we would expect the elasticity solution to be superior to
the strength of materials solution. On the other hand, an assumed stress
distribution may accurately portray the system due to factors such as local
yielding and, in such cases, the strength of materials method may be favored.

SECTION 5-3
The Nature of Stress and Strain
Consider a finite plane area on the surface of a body or within the body.
Average stress is defined as the force on that area divided by the area. If the
force is normal to the plane, the stress is called a normal or tensile stress;
, the resulting stress is a shear stress. The
word stress will not be used to refer to total force in this work, although
that usage is common in structural analysis.
The nature of most engineering materials is such that a small fracture
anywhere in a load-carrying machine element may be expected to result in
a total failure. Therefore, average stress is of little value; we are most
concerned with maximum stress. Thus, we define stress over an infinitesimal
area : the limiting value of force divided by the area over which it acts where
the area approaches zero
AP
t
stress = en
himit=—
A Soll
(5-1)

Stress concentration
Following the strength of materials method, relatively simple stress
distributions are assumed in machine members (for example, uniform stress
in tension members ; a linear stress distribution for bending). However, holes,
fillets, and other section changes cause high local stresses which may be
accounted for by using a stress concentration factor. Stress concentration
 238
Chapter 5: Analysis of Stress and Displacement

factors and methods of reducing stress concentration are discussed in

Chapter 3.

Combined stress
When finding the effect of a combination of loads on a machine member,
we must combine stress. In so doing, it is necessary to note the direction of
each stress and the plane on which it acts. Failure due to combined stress is
discussed in Chapter 6.

Strain
Tensile or normal strain represents the extensional change in length of
an element per unit length. In theoretical studies, strain, like stress, ordinarily
refers to an infinitesimal element. However, in experimental work, strain
may be measured over a finite distance. Shear strain is the change in angle
(radians) from a right angle.

Limitations and effect of assumptions

As noted above, we assume that a homogeneous, isotropic, linearly
elastic solid will be an appropriate model, applicable to the design of most
machine elements. However, as a result of rolling and other processes, there
may be preferential grain orientation in some materials so that isotropy
does not strictly hold. The linearity and even the elasticity may be in doubt
as the loading of a part approaches the ultimate strength of the part; thus,
we will ordinarily design for stresses well below the yield strength. One of
the most serious questions is whether the mathematical model, which is the
basis of our stress analysis, is actually representative of the part that we are
attempting to design and of the expected loading. For example, we must
design an elevator on the basis of the greatest load that it could reasonably
be expected to carry, but we cannot control its use after it is manufactured.
There are many additional assumptions necessary to provide a simple
basis for analysis. In the bending of beams (including shafting and other
machine elements) the elementary theory is based on the assumption that
the slope of the deflected beam is small compared with unity; and in plate
theory, accurate results require a stricter limit on deflections: the deflection
should be small compared to plate thickness.
If our design is to be based on the yielding or fracture of a material, we
must have reliable data on the strength of the material. Yield and tensile
strength, however, can only be obtained through destructive testing, and
test results vary considerably from sample to sample—even among samples
produced from the same heat. Thus, we may approach the problem statis-
tically, accepting strength values corresponding to a satisfactory degree of
reliability.
The time history of the loading is critical as well as the (maximum)
magnitude of the loading. Even in this respect, however, the actual condition
is idealized. That is, we assume that the’static vald@ oPloadingcssincreased™
 239
Section 5-3: The Nature of Stress and Strain

; or, if some repeated

loading is known to occur, we may assume a harmonic component of
loading superposed on a constant value of loading.
After we have accounted for all the loading effects that we could reasonably
anticipate, it would still be imprudent to allow maximum predicted stresses
in a part to approach the failure level closely. For this reason, it is customary
to design on the basis of maximum stresses that do not exceed a working
stress. Working stress is given by the failure strength (for example, yield
point or ultimate strength) divided by a factor of safety. The factor of safety,
a number exceeding unity, represents our degree of uncertainty or ignorance
concerning loading, material properties, and so on. Its value is influenced
by the probable hazard to life and limb and the monetary cost in the event
of a part failure.

Measurement of stress and strain

For some machine members, stress can be calculated directly if the loads
are given. Experimental stress analysis is used when loading and part
geometry are complicated.
The most common method ofverifying analytical design methods involves
the use of strain gages. Thin wire or foil may be bonded to paper or plastic
backing, as in Figure 5-la, forming a resistance strain gage. The gage is

Load

Gage
length

(b)

—-_
Figure 5-1 Strain gages. (a) Schematic of a strain gage mounted on a beam.
(b) A rosette gage.
 240
Chapter 5: Analysis of Stress and Displacement

Figure 5-2 A rosette gage mounted at the clamped end of a cantilever beam.
[Courtesy Vishay Research & Education, Romulus, Mich., Division of Vishay
Intertechnology, Inc.]

Figure 5-3 (a) A photoelastic model of a beam (see also next page).
[Courtesy Vishay Research & Education, Romulus, Mich., Division of Vishay
Intertechnology, Inc.]
Figure 5-3 (continued) (b) through (f)
Photoelastic models illustrating stress
distributions for various geometries.
[Courtesy Vishay Research & Education,
Romulus, Mich., Division of Vishay
Intertechnology Inc.]

241
 242
Chapter 5: Analysis of Stress and Displacement

bonded to the part that is to be tested so that the gage is strained when the
part is loaded. Strain in the part is determined from the resistance change in
the strain gage and the known gage characteristics. Both static and dynamic
loads may be measured. Strain is given by the unit change in resistance
divided by the gage factor (supplied by the strain gage manufacturer). Since
resistance change is small, a bridge circuit is usually required. A strain gage
measures average strain over its gage length (usually between ¢; and jin.).
If strain changes rapidly with location, a strain gage with a small gage length
should be used.
When the principal directions of strain on a part are unknown, a rosette
gage is used (Figure 5-1b). It consists of gages at different orientations.
Figure 5-2 (page 240) shows a demonstration model of a cantilever beam
instrumented with a rosette gage.
Photoelasticity is a method of experimental stress analysis in which a stress
sensitive transparent plastic model is used to represent a machine part
(Figure 5-3a, page 240). Loads applied to the model are proportional to loads
encountered by the actual part. Using special optical instrumentation (a
polariscope), observed patterns can be interpreted to indicate stress in the
part. Stress concentration can be examined at holes, notches, and fillets
under various types ofloading (Figures 5-3b, c, and d, page 241). In addition,
shapes that would be difficult to treat analytically may be studied by photo-
elasticity methods (Figures 5-3e and f). Tests of this type sometimes lead to
redesign of a part to reduced stresses.
Vibrating members and other dynamic problems may be studied by
observing a photoelastic pattern using stroboscopic light. Three-dimensional
stress patterns may be examined by slicing a “frozen” stress pattern or by
using special lateral illumination of the model. Instead of using a model,
it is also possible to bond a photoelastic coating directly to a machine part.
Brittle lacquer techniques are also used to provide experimental data. A
brittle coating is sprayed on a machine part. Then, as the part is loaded in
proportion to loads that would be encountered in service, cracks begin to
appear in the coating. The extent of the cracks is noted for each increment
of load. The coating is “‘calibrated’’ by spraying it on a simple beam and
observing the strain at which cracks appear. Brittle lacquer techniques alone
may be used to obtain quantitative results. Frequently, however, the results
are used to aid in placing strain gages for further measurements.

SECTION 5-4
Bending Stresses in Machine Members
In the design of machine members, bending stresses are frequently the
major criterion. Let us consider a machine member with one dimension
much larger than the others and let there be a vertical plane of symmetry
through the longitudinal axis (see Figure 5-4). In that figure, the xz plane is
 243
Section 5-4: Bending Stresses in Machine Members

XZ plane is plane of symmetry

Figure 5-4 Bending loads in machine members.

Neutral surface (x axis)

in unloaded position

Deflected
neutral surface
(x axis)

Change in slope between

two adjacent sections

Figure 5-5 Deformation due to bending. The axial strain in element CD is

d?w

and the tensile stress is

d?w
 244
Chapter 5: Analysis of Stress and Displacement

the plane of symmetry and the xy plane is the neutral surface before deflection.
If there is to be no twisting about the longitudinal axis, forces may be
applied in the plane of symmetry and moment vectors may be applied
perpendicular to the plane of symmetry. Then the (originally straight)
longitudinal axis will deflect within the plane of symmetry. Bending effects
will predominate and the usual assumptions of beam theory apply.
Let us consider a plane section perpendicular to the longitudinal axis
ofa member. Ifthat section remains plane and perpendicular to the (deformed)
longitudinal axis, then the stress field in the member is readily determined.
In Figure 5-5 (page 243) bending stress a, is related to the second derivative
of deflection w, whereas in Figure 5-6, bending moment M is related to the
same quantity. Combining the results, we obtain
Mz
0. ————; (5-2)
Sed)

which reaches a maximum at the surface z = c where

Mc
Ox(max) == bd Bd : (5-3)

SS
pee

Figure 5-6 Stress due to bending.

Resultant moment

M =| podA = = pre
section
where

r= | z?7dA
cross
section
In this sketch, bending is produced by a positive moment, whereas
in the preceding sketch the value of d?w/dx? was positive.
 245
Section 5-4: Bending Stresses in Machine Members

This value of stress may be a basis for design when bending moment is
computed from the loading. The determination of bending moment due to
various forms of lateral loading is illustrated in Section 5.5.
For a rectangular
cross section, c = h/2 and J = bh?/12. Thus

6M
0 x(max) =. bh2 (S-3a)

For a circular
cross section, | = nD*/64 and
32M
O x(max) = xD? (5-3b)

The above results assume linear elasticity, that is, 0, does not exceed the
proportional limit. Furthermore, they are based on the assumption that
stress is independent of the y coordinate and that y direction stress, 0, = 0.
These conditions are generally met if the loads are applied in the axial
plane of symmetry, if the dimension b is small compared with the length of
the member, and if computed stresses are well below the yield point of the
material. Analysis of beams having considerable initial curvature is illustrated
in Chapter 18.

Composite sections—center of gravity

Some machine members are made in the form of T or I or channel cross
sections. In some cases, the center of gravity of the cross section cannot be
located by inspection. Consider, for example, the T section, Figure 5-7a,
with loads in the plane of symmetry (the vertical plane). In general, the
distance from some arbitrary reference to the center of gravity for any cross
section is given by

(5-4)

Integration is carried out over the entire area A. Both z’ and z are measured
from the arbitrary reference, which could be, for example, the top of the
section.
When it is convenient to separate a cross section into two or more parts,
the expression for the ity may be given in terms
of the individual cross section areas and the distances from the arbitrary
reference to their centers of gravity: For example, the center of gravity of
the T section in Figure 5-7b is located by the equation

2A, + 253A,
i
Ale A;
Using the dimensions in the Figure 5-7a, we obtain z = 0.72 in.
 246

Cee
Chapter 5: Analysis of Stress and Displacement

(a)

(b)

Cc,

Tae
2 4 - Sp Neuteal axis (c)
do

Cy

pe BEd Be
Figure 5-7 Bending of a composite section. (a) T section loaded in plane of
symmetry. (b) Location of center of gravity. (c) Moment of inertia.

Moment of inertia for composite sections

In general, the moment of inertia of a cross section is given by

I= { z7dA
cross section

For machine members in bending, we are interested in the moment of

inertia about the neutral axis—for example, the horizontal axis through
the center of gravity G in Figure 5-7c. The moment of inertia of acomposite
section about its neutral axis may be expressed as the sum of the moments
of inertia of its components about that axis. Furthermore, the moment of
inertia of each component area may be expressed as the sum of the moment
 247
Section 5-4: Bending Stresses in Machine Members

of inertia J, of that area about an axis through its center of gravity parallel
to the neutral axis plus the transfer effect Ad?, where A is the individual
area and d is the distance from the neutral axis to the center of gravity of
that area. Thus, the total moment of inertia about the neutral axis is given
by
SOIR 28 LE) (5-5)
summed over all area components. For the T section shown in Figure 5-7,
we find J = 0.291 in‘.
In this example, the bending stresses at the extreme (top and bottom)
fibers are not equal. Normal stress (bending stress) at the top is

~ Me,
|
where c, = 0.72 in. and at the bottom,

Mc,
(6h

J I
where c, = 1.28 in. One stress will be tensile, the other compressive, depend-
ing on loading and support.
When a member is loaded in a vertical plane of symmetry and has a
horizontal plane of symmetry as well, the horizontal plane of symmetry
forms the neutral surface (the neutral axis on a cross section) and can be
located by inspection. In many cases, we may take advantage of the symmetry
when determining the moment of inertia. In Figure 5-8, for example, we
note that the moment of inertia of the I section with equal flanges is the

| Loading =

ea
72 a
2
hy) |hg —-— — ——Neutral axis

Figure 5-8 [| section with equal flanges.

 248
Chapter 5: Analysis of Stress and Displacement

moment of inertia of the outer rectangle less that of the void

baht _bah}
12 12
Because standard rolled steel sections are not actually made up of rectangular
shapes,
the
manufacturer.
St. Venant’s principle
In effect, St. Venant’s principle states that

thereby enabled to solve many problems without undue rigor. For example,
by treating gear loads and bearing reactions which are distributed over a
small area as their resultant point loads, we may analyze stress and deflection
of shafting. Obviously, we must use caution in applying this simplifying
device, particularly when we are interested in determining stresses. As an
extreme example of improper application, consider a slender
rod’ subjected
to a tensile force P at each end. If the forces are reversed at each end, it
could be argued that the force system is statically equivalent to the first
case, but we note that the stress situation is not the same and, even if the
first case represented a safe part, the second case could be unsafe on the
basis of elastic stability.

An elementary design problem

Consider the design of a handle for a ratchet wrench (Figure 5-9). There
are many design decisions, and many different acceptable designs are possible.
The length of the handle is based on the torque required for tightening.
We set the required torque equal to Pe, where e is the distance from the
center of the socket to the line of action of the approximate resultant force
P applied by a user. The cross section of the wrench handle would be based
on a higher load, however. In removing rusted bolts, it is not unusual to
extend the wrench handle by putting a piece of pipe over it, thereby increasing
the moment arm, or to hit the handle with a hammer. Although we cannot
design an “‘unbreakable’’ wrench, we must base our design on the worst
loading that could reasonably be expected.
Let us specify an I section for the wrench handle and assume that the
critical (most highly stressed) point lies on section A. The moment there is
M, = Pa, where a larger value of P is used than required for tightening.
A first approximation of moment of inertia about the neutral axis in bending
may be made by neglecting the contribution of the web and the contribution
of the flanges about their central axes. Then, we have simply I = 2bhd?, and
a maximum tensile stress on the outer fiber at section A of
_ M4(d + h/2)
bi a ae)
 249
Section 5-4: Bending Stresses in Machine Members

foeee
a
d
Neutral axis
in bending

Section A

Figure 5-9 Ratchet wrench handle.

We may set o, equal to a working strength o, = S,,/N, where S,, is the

yield point of the material selected and N is the factor of safety. We would
probably select a ductile steel for the design, in which case the first evidence
of failure would be yielding. Since this presents no danger to the user, a low
factor of safety would be adequate.
In order to complete the problem of designing the section, we may reduce
the number of unknowns by arbitrarily relating the section dimensions to
one another. Then, we will have a single equation with one unknown
dimension to solve for. The handle may be tapered to a smaller section near
the end where the force P is applied, provided that the working stress is
nowhere exceeded.
As an example, let the overall length of the wrench handle be about 11 in.
and let the distance from the applied force to the critical section be 7 in.
The applied force would actually be distributed over the part of the handle
near the end. However, we would, according to St. Venant’s principle,
assume a point load near the center of the actual distributed load because
the critical section is some distance from the applied load.
Let us arbitrarily specify the proportions of section A (Figure 5-9) by the
relationships b = 4h and d = 4h. Assuming that a user could apply a
maximum force of 100-150 lb to the wrench handle, we might design for an
 250
Chapter 5: Analysis of Stress and Displacement

applied force P = 300 1b to account for shock or misuse. Let the part be
made of steel with a yield point S,, = 75,000 psi and, because part failure
would not have serious consequences, use a factor of safety N = 1.5.
Then, using equation (5-6) we have

_ 75,000 psi _ (300 Ib)(7 in.)(4.5Sh in.)

Ox 15 2(4hin.)(hin.)(4h in.)
from which h = 0.114 in. In round numbers, we might use gin. thick web
and flanges making the I section 4 in. wide by 17 in. deep. At this point, we
could recheck the design, computing an “‘exact’” moment of inertia based on
final dimensions. This is unnecessary, however, since the approximate
calculation was conservative (that is, on the safe side).

SECTION 5-5

Deflection of Machine Members

Due to Bending
Machine members are usually subjected to loads that can be represented
by concentrated forces and moments. Distributed loads, especially those
arising from the mass of a member, are seldom important in the design of
machines. Inertia loading is an exception and, of course, large structures are
subjected to significant distributed loads.
The relationship between loading of a machine member and its bending
deflection is treated in detail in strength of materials texts (references [2],
[3] and [4]). Some of the results are summarized in Figure 5-10 along with
the usual sign conventions.
As an application of the expressions in Figure 5-10, let us determine
bending deflection in the horizontal part of the machine member shown in
Figure 5-11. Let a load of 707 lb be applied at point E as shown. The member
is simply supported at points B and C; the support at B resists the horizontal
component of load P.
In addition, load P causes a bending moment of 1500 lb-in. and a vertical
force of 500 1b at point A. The solution to this problem is illustrated in
Figure 5-12. Ignoring the weight of the member itself, reaction R, is deter-
mined by equating the sum of the moments to zero at the simple support C.
Then equilibrium of vertical forces yields R, and the shear and moment
diagrams may be constructed. Curvature is given by

—_ =-— 5-7
dx? EI OD)
but it is not necessary to make an additional plot if E and J are constant.
If we have not yet specified the section dimensions, the deflection variable
EIw may be determined by integrating twice with respect to x. Constants
 Zod
Section 5-5: Deflection of Machine Members Due to Bending

Distributed load Positive distributed load

+q(x) |b/in. decreases shear

Concentrated load Shear force V = —fqdx

ue (Aye
Positive concentrated load causes
R, Reaction Ro negative jump in shear. Reactions
(as shown) cause a positive jump

M, Bending
(C0)
Bending
a— moment moment M = {Vdx

Concentrated moment (as shown)

Causes a positive jump in bending
moment. Positive bending moment
c Me: 2 —M
déw_ —M M corresponds to negative curvature
urvature is given by ae = 5) d2w/dx2

Deflection of neutral surface

w = f0dx

Figure 5-10 Sign conventions and relationships between loading and deflection.

of integration are determined by maintaining continuity of slope and deflec-

tion and by setting the deflection equal to zero at B and C.
Although the slope variable curve was determined analytically in Figure
5-12, graphic integration would produce identical results. If the results are
to be the basis for an actual design, it is recommended that the problem be
solved both graphically and analytically and the solutions compared.
If we wish to complete the solution graphically, beginning with slope,
we may temporarily set 0, =0. Then integrate graphically, using the
trapezoidal rule. In this problem, + in. increments in x will produce results
within “engineering accuracy.” The resulting plot is shown in Figure 5-13.
Since the neglected part, 0,, is a constant, its integral @,x is a straight line.
Thus, to obtain the w = 0 line, we draw a straight line intersecting the Elw
 22
Chapter 5: Analysis of Stress and Displacement

| 2in. — > 3in, ——_—$o} 1.5 in. —o

thickness b

|
45°
P=707 Ib

Figure 5-11 Machine member subject to bending loads.

curve at the bearing locations, B and C. The true displacement variable is

given by the vertical distance from the w = 0 line to the Ew curve.
The value of slope 0, may now be found by locating the point where
dw/dx = 0 on the Elw curve. A line is drawn parallel to the w = 0 line and
tangent to the EJw curve. The point of tangency is the point of zero slope.
Projecting that location upward to the EJ@ curve, we may now dimension
the slope.

Superposition
Most problems in the bending of machine members may be considered
linear problems. Because reactions and shear force are proportional to
loading, we see from Figure 5-10 that moment, slope, and deflection are
linear functions of the applied load (provided, of course, that the usual
—are valid). Then,
the combined effect of two or more loads acting simultaneously is given by
the sum of the effects of each load acting separately. The last sentence is a
statement of the principle of superposition. Superposition is a valuable
technique for handling statically indeterminate problems and statically de-
terminate problems where two or more loads act on a member.
For convenience in solving such problems, we will begin by collecting a
set of solutions to basic statically determinate problems (Figure 5-14).
These problems are straightforward in that reactions can be determined by
statics (that is, equilibrium relationships) alone. Since dimensions will be
expressed in general terms, solutions of this type are obtained analytically.
As an example of superposition, consider a machine member represented
by Figure 5-15. The elastic modulus and moment of inertia of region AB
 253
Section 5-5: Deflection of Machine Members Due to Bending

Figure 5-12 Loading and deflection of a machine member

M, = 1500 Ib-in.

Ly =3in.

fz w Loads in vertical plane

Axial location x
0 i 2 3 4 5 6 in.
+500 Ib
Shear 333 Ib
Force 167 Ib
V 0
x4 o—__—__>
a a ee WD)
0

— 500Ib in
EI6
Bending +500
Moment
M
0)
M =-—1500 + 500x
: O<x<2
—1500 in —500
Ib
x2
EI9 ==E/0, + 1500x — 500 — —1000
O0<x<2 Z

Slope E10 = E/0g + 500x, — 167 =

i
—1500
Variable 0<x, <3 2
£10, == El0g =+ 750 == E/0, + 2750 —2000

—2500

—1000
Deflection
Variable
0
Elw
= Elw, + E10,x + 750x? — 83x3
+1000 O0<x<2

Elw = Elwg + El0gx, + 250x* — 27.8x?

+2000 0<x, <3

Elw = Elwe
+ El ox>
+3000 0<x<15

are given by E, and J, and those of region BC by E, and J,. The applied
force F lies in plane ABC and moments of inertia are to be calculated based
on bending in the plane ABC. The connection at B is rigid so that moment M
and slope @ are continuous there.
For convenience in solving this example, force F is replaced by its horizontal
and vertical components. The problem is solved by superposing the slopes
and deflection given in Figure 5-14a and b. To avoid ambiguity, subscripts
and coordinate axes are changed where necessary.
 254
Chapter 5: Analysis of Stress and Displacement
Corrected E/0
EI(0 — 0a)
(for plotting) |

3000 +500

2000 —500

E/@ scale
1000 Ib in.? —1000
(a)
i Construction
1000 to locate —1500
zero slope

—2000

a | 5 6
i 2500
e 0 1 2 3 4

Construction to
locate zero slope

0 Measure E/w vertically

from w = O line
Parallel to w = 0
Figure 5-13 Graphical solution for
wt— line, tangent to
Deflection SK deflection curve
slope and deflection in the vertical plane.
2000 (a) Slope. (b) Deflection.
curve

4000

6000

Elw scale
ie) for plotting

8000

10000
 sy)
Section 5-5: Deflection of Machine Members Due to Bending

(a) “ (b) 4
x

[5 < = —P(L — x)
2 2
E10 = P(x ~7>) e= Se

Elw == P (Lx2 ——x3 =aus

ieee el eee) Veer]

6 IC Rox}
Clty aie 2E/

LiLo Ly
Wes (i eueid)
6E/ L

ae a Rox
(c) irene 7 2 6E/

=L
R,x? PL PLSL3
Way eee ee ees aa
Ox je, 9 ET Gail 2 8 3E/L
1
hefhapf pfle |x
ifL,>Llo Wrngx OCCUrs
at xX, = aw

+M y Clamped end
(Sao. —
6 =0

3 iy, 2 2 Simple
oe support
A w=0
Support does
Sign conventions
not apply moment
for Moment, Slope,
or axial restraint.
and Deflection
6 is given in radians.

Figure 5-14 Bending of members with constant cross section (see also next page).

Bending stress reaches its maximum value Mc/I at the outer fiber of a
given cross section. To this, we add direct tensile stress. In region AB,
bending moment magnitude may be maximum at A or B depending on the
direction of F. Maximum tensile stress is given by the sum of bending stress
and direct tension

a |M |(maxy© F,
O x1 (max) 7 I Bc A
1 1
 256
Chapter 5: Analysis of Stress and Displacement

M,x?
mera
O<x,<L,
ML ML M,x3 Milbee=Ol>)
A GEI BS Sar] 6
veveu =0, 2EIL im 6E/L
x3
w mes ae
? Myx? =M,x,(L2
—343 —x?)
w
M,L?2 L Get nee OC menre 6E/L
Wmex ~ 9 Ee) * 7B

forO0<X <L
pie
see Figure 5-14b pee O0<x9<L9o
=—P(Ly —x5)
2 z

OK i) eh, =0,=—s>
8 BF
es eee SPL ks
Seu ene Et Ae REET
= MPEZa mea
Wo <xq Ly Wet! BX = Fy (5 = te x 22
2 _ Phils
PERE
Ox cn ee (4ox2 7) 8a ae
~PL P|5x x421 PL Pe
w 8 i)
oe <r; GET (4; ia) CCE) a2)
_ Pxo [ese Lox |
Cex cin Eales 2 6
PLL3
We ~ 3E]

ForO0< x <L see Figure 5-14d

(7) alee
O<ky <=, !
Mx,{L 4
w = ——_ J]-—-4+ —
O<x, <L, fal) \\Sp

Figure 5-14 (continued)

 D7,
Section 5-5: Deflection of Machine Members Due to Bending

Fy =Fsino
F (resultant) Fol5

08 -F Ey
B

Loading of
region AB

2E4/;
Fol3
seca 2E oI5

Fok 3
B
arr _ Fill yxy2
FACE
— 3) —Foboxsw
M = ay ia Ey =e) O0<x,<L 2E,f
0<x,<L, ou nee AB ‘eae ie
Ripls/S
=Fok Eal2.
W1B
Eyl,

Wo adOxo + F 229x3/3
— Fol.
2h ox2x3
0<xy <L, 2Eo/>
BC
Fililo—eFolahs F5L3
M A WIC
Patieenenmeea 2E4/, 3F 5/5

Foly—FyL, “Horizontal deflection of BC is Wig’

Figure 5-15 Superposition.

In region BC, M, = F,L, is obviously the maximum value and the corre-
sponding tensile stress is

fa _ Myc n le
x2(max) ~ ie A,

The results are based on the assumption that slope @ is everywhere small
compared to 1rad. If this condition is not met, the loading then depends
upon the deflection. In the example just described, the axial loading was
tensile. When compressive loading occurs, we must also examine the problem
of elastic stability to insure that buckling failure will not occur.
 258
Chapter 5: Analysis of Stress and Displacement

Statically indeterminate problems in bending

In the preceding examples, the support reactions were found by using the
equations of equilibrium. A machine member clamped at one end and
simply supported elsewhere (Figure 5-16a) is statically indeterminate in that
the equilibrium equations are not sufficient to yield the support reactions.

y fg 2 —
A
ae Z c DB — = _ (a) Actual loading.

{" tr ,

(b) Reaction Ry removed.

(c) Load P removed.

tr
: _ PL4(2L,
2
+ 3k)
2 2(L, +L£5)3
Mig —P\En x) ih (eee X)
O<x<l,

M = faiglbn 4b(hs
=59)
fi SQL tlh,
i” _ PLF3x
~—Ly) — px? (3L, + 3Ly —x)
Ly <x <L,+L, 6E/
Figure 5-16 Statically indeterminate member (redundant end
support).

This problem is solved by superposing the effects of load P alone (without

reaction R,) upon the effects of R, alone (without load P).
The deflection at C due to load P (Figure 5-16b) is
_PLi[Ly | Ly
Sie, HG 3]
calculated from the equations in Figure S-14f with the necessary changes in
symbols. Similarly, for the same point when reaction R, is applied (Figure
5-16c), we have

—R,(L, + L,)°
Wo —
SEI
 259
Section 5-5: Deflection of Machine Members Due to Bending

Superposing the two effects, we use the condition

Mie = Mii + w,. =0

to obtain support reaction R,. Using this value of R, and the equations in
Figure 5-14f, we obtain deflection and slope at any point.
The location of the point of maximum deflection is found by differentiating
the deflection equation with respect to x and setting the result equal to

(a)
Three
Support
Member

(b) 7» Ae
Intermediate 58 PL ok L aE
Support eee 4 =e $4) PL gl
Removed (ie = 6E/ 6E/ L Wp = 2 (2-22 — 12)

m' La<xsl
MRL ei Rx?
NDE

Effect oeee Dax 6E/ 3E/L Sign Convention:

of R, w, and 0
Intermediate OP Roly fs fi£3) positive as shown;
Support A 6E/ je

M ;
Mm" ale,
=—Rix MM" weinr e
=—R(L—x) C20 +

(d) PL PL Rol
Combined + (03 +.2045-22) ee
Moment L
Diagram

Figure 5-17 Statically indeterminate member (three supports).

 260
Chapter 5: Analysis of Stress and Displacement

zero. In the region from B to C (Figure 5-16a) maximum deflection occurs at

PL?
x= L, ets + L,)*?-
R,

If, for example, load P is equidistant from A and C, then R, = 7¢P and the
greatest bending stress occurs at A where moment is given by M, = —3PL,.
For that case, the maximum deflection occurs at x = 1.106L,. In general,
we should check each region of a given problem. Solutions that fall outside
of the region being checked are rejected, and deflections are examined at
end points of the region.
The solution to the three-support member, Figure 5-17 (page 259), is
similar. In this case, we superpose the effects of load P (without the inter-
mediate reaction) on the effects of the intermediate reaction alone, treated as a
load. These values may be obtained from Figure 5-14c. The magnitude of the
reaction is found by setting the displacement at the reaction equal to zero.
Parts b and c of Figure 5-17 illustrate the superposition. Because the true
deflection w, = wz — wz = 0, we have

R,
PLL?
=? 13)
21,2
For the special case of two equal spans with load P at the midpoint of
one span, the center reaction is simply R, = +4P. In that case, the maximum
bending moment, M = ;3;PL, occurs at load P.

SECTION 5-6

Shear Stress and Shear Deflection Due to

Lateral Load—Shear Center
Bending stresses and bending deflection are usually of primary interest
in the design of laterally loaded machine members. However, shear stress
and shear deflection due to transverse shear forces are sometimes important
as well.
Consider a machine member subjected to lateral loads that lie in a plane
of symmetry of the member (Figure 5-18). At some location x, the bending
moment has a value M, and there is a linear distribution of axial stress given
by

0, =— (5-8)

where z is measured from the neutral axis. Moving to the right a distance dx,
the moment increases by (dM/dx) dx or V dx and axial stresses increase by
[(V dx)z]/I where V is the shear force.
 261
Section 5-6: Shear Stress and Shear Deflection Due to Lateral Load—Shear Center

PIAE
as SE

(a)

at H (b)
Vv

aaa (ef NS \ yaa

| Pik ® =| Moy dx
rae j — — (—neutral axis 6
c

| | ae
ee
x dx
aN

pp DONS

Figure 5-18 Shear stress due to lateral loads. (a) Machine

member with lateral loads. (b) Shear force on cross section.
(c) Moment and stress distribution. (d) Equilibrium of axial force.

Now, consider the element that lies below z = z, (Figure 5-18c and d).
Transverse shear stress t,, on the top of the element produces shear force
T,,b dx which is balanced by the net tensile force
Vea xem
zdA
I és
Z=21

Solving for transverse shear stress, we get

"xe =al_
= BT ere dA 5-9
(5-9)

where dA is an element of cross section area and b is the width of the section
 262
Chapter 5: Analysis of Stress and Displacement

normal to the page at z = z, where we wish to find t,,. Shear stress t,, = 0
at z = +c, the outer fibers. The maximum value occurs at the neutral axis
V Wi

{ee
( ) bI nes
zdA (5-10)

For a of area A, dA = bdz and the equation

for shear stress becomes
15 “| (=1
Tx. eee ral ba
A C

the
and for an oaded in the usual
fashion (that is, with load resultants through the plane of the web), transverse

We note that for a member subjected to vertical forces in its plane of

symmetry, maximum transverse shear stress occurs at the neutral axis where
tensile stress due to bending is zero. Transverse shear stress is zero at-the
extreme (top and bottom) fibers where tensile stress is a maximum. Bending
stress at the extreme fiber is of greatest importance in most solid members
of rectangular or circular cross section subject to lateral load. In very short
members, transverse shear may govern the design. In
shafting,
shear stress
The effect of combined shear and
tensile stress will be considered in Chapter 6. In I-type sections, the combined
stress may be critical where the flanges join the web (see Section 6-3, The
. Maximum Shear Theory).

Shear deflection
Shear deflection arises from shear strain and, in the case of most laterally
loaded machine members, gta COERCION tree
effects. The additional slope due to shear is the shear strain at the neutra
axis

noeG (5-11)
where G is the shear modulus. Thus, for a rectangular section, yy = 1.5V/(AG);
and for a circular section, yy = {V/(AG) where A is cross sectional area. For
an I section yy = V/(GA,,.,) (approximately), which may be 2 or more times
V (AG).
The deflection change between two points on a member is given by the
integral of the shear strain between those points. For example, in the case
of a simply supported rectangular member of length L with central load P,
the shear deflection at the center is

(ie Dfee 3a BE
ee ee AG
 263
Section 5-6: Shear Stress and Shear Deflection Due to Lateral Load—Shear Center

deflection. At the center, the bending deflection is PL?/(48EI), which in the

case of most rectangular members is much larger than the shear deflection
so that the latter can be ignored. However,
and honeycomb and other sandwich sections may
be of the same order of

Shear center
In the laterally loaded machine members considered previously, we have
assumed that load resultants lie in a plane of symmetry through the longi-
tudinal axis of the member. Thus, we assumed that load resultants go
through the center of gravity in the case of a symmetric cross section. Under
these conditions, the member should deflect in the plane of symmetry without
twisting. This arrangement, however, is not always feasible.
As an example, consider the channel section in Figure 5-19a subject to
vertical loads. It is not symmetric about a vertical plane and will tend to
twist unless load resultants lie in a certain plane defined by the shear center.
Let the shear center lie an unknown distance e from the center of the vertical
wall of the channel and assume load resultants do go through the shear
center. In the case of a cantilever with load P between the free end and the
cut section shown, there is a vertical shear V = P in the vertical wall of the
channel. The vertical wall is assumed to take the entire vertical shear force
V, whereas the horizontal walls are assumed to have only horizontal shear
resultants. The actual shear stress distribution may be very complicated,
but we will treat thin sections in a simplified manner, neglecting any variation
in stress across the thickness t. Equation (5-9) may be applied to this problem
if we integrate along the centerline of the web and then along the centerline
of the flange to the free edge of the flange. We may assume continuity of
shear stresses where the web joins the flange.
Thus, our approximate shear stress distribution has the parabolic form
V i_ al |
a o= I > Oz | ay,
| Zi1 2 10)

V[h? hd
a it ar ay — | (5-12)

for location z = z, in the web. For location y’ = y’ in the flange, shear stress
distribution has the linear form

Vig?
RY

ad ——1)y,)
Vh
=——(d (5-13)
5-13

For this example, shear stress direction is sketched in Figure 5-19b.

 Free end

Channel
Section h/2

Neutral
) axis
Zz

t (typical) h/2

(b) Tas

ix

(c)
e h

Vv

@ Center of gravity

(d)

Figure 5-19 Shear center. (a), (b), and (c)

Channel section. (d) Angle section.

264
 265
Section 5-7: Torsion

If we integrate equation (5-13) over the entire flange, we obtain the shear
resultant in the lower flange (on a cut perpendicular to the beam axis)
F = Vhtd?/(41). An equal and opposite force appears in the upper flange.
The moment Fh due to the shear resultants in the flanges is balanced by the
moment Ve due to vertical load and vertical shear, provided resultant loads
and reactions are applied at the shear center. Equating moments, we may
solve for the location of the shear center
h? td?
é =———
4]
(5-14)
The value of e found from equation (5-14) locates the plane of loading if
the channel is to be free of twisting. This location could seldom be realized
in an actual design; therefore, we can expect torsion effects, particularly in
nonsymmetric sections.
For the case of thin sections that join at a single point (for example,
point A in the angle section, Figure 5-19d), the shear center is that point.
For vertical loads on the angle section, there will be no twisting if the loads
go through the centerline of the vertical leg. As noted previously, if a section
has a vertical plane of symmetry, vertical loads in that plane will not cause
twisting.

SECTION 5-7
Torsion
A moment that tends to twist a bar about its own axis is called a torsional
moment, or simply a torque. The same force system may cause bending in
one member of a machine and torque in another.

Shear stress due to torque in a circular shaft

When a solid circular shaft is subjected to torque T, we assume a linear
shear stress distribution within the shaft.

(Figure 5-20). A more rigorous analysis of the problem

shows this assumption to be valid for solid and hollow circular shafts, but
not valid for other shapes.
We write the linear shear stress—torque relationship inthe form
eG
1G ; (5-15)

Observing that the shear stress on each element of the cross section contrib-
utes to the torque according to the integral

ipe= il trdA
cross section
 266
Chapter 5: Analysis of Stress and Displacement

4S e\

Figure 5-20 Torsion ofa circular shaft.

and substituting equation 5-15 we obtain

Fie mD*
ete)
for the solid circular section of diameter D. The constant J is, of course,
the polar moment of inertia of the cross section. For the shaft with a central
hole of diameter D;
m(D* — D?)
ike
32
Ain SRT RETA in solid and hollow
circular shafts. Thus, from equation 5-15, we have
TD
iG ——
max Ou

Tmax = ~H3 (5-16)

We will see later, however, that for noncircular sections, stiffness is not
proportional to the polar moment of inertia and that shear stress is not
proportional to radius.

Deflection due to torque

The modulus of elasticity in shear, or simply shear modulus is defined by

G=— (5-17)

where shear strain y is dimensionless. Consider an element of shaft of

diameter D (Figure 5-21). Let one face rotate through an angle d@ with
respect to the other face. For small strain, the arc of the circumference
 267
Section 5-7: Torsion

Figure 5-21 Shear strain and

deflection.

(D/2) dd is approximated by y dx. Equating the two and using equations

(5-15) and (5-17), the result is the angular deflection per unit axial length
0 = d¢/dx = T/(GJ). For constant torque T, relative rotation between two
faces at a distance L is given by
(WE
o (rad) = (5-18)
GJ
It is sometimes convenient to define a torsional spring constant k, for a
finite length of shaft as the torque per unit rotation. Thus, from equation
(5-18) we have
Te Gu ;
= D hye (lb-in./rad) (5-19)

In the case of a stepped shaft, we may subscript k,, L, and J for each section.
If the torque is constant over the entire length, the equivalent spring constant
is given by the reciprocal of the sum of the reciprocals of the constants for
each diameter

Noncircular sections in torsion

Due to the symmetry of a circular shaft, it is quite logical to assume that
there is no warping of cross sections (planes perpendicular to the shaft axis).
However, a plane cross section of a noncircular shaft will warp to a non-
planar shape when subject to torsion.
Consider a noncircular cylindrical section (that is, a shaft or other member
of constant but noncircular cross section) subject to a constant torque over
 268
Chapter 5: Analysis of Stress and Displacement

egy

zx

Figure 5-22 (a) Torsion of a non-

circular section. (b) Shear on a lateral
surface element. (c) Elliptical cross
section. Maximum shear stress occurs at

—b

(c)

its length (Figure 5-22a). Because the section does not change, we would
have a constant rate of twist (in radians per unit length), which we will
designate 6. Let u, v, and w represent displacements of a point in the x, y,
and z directions, respectively. Then, for a small total rotation 0z, we have

u= —Ozy
b= AG (5-20)
 269
Section 5-7: Torsion

where W = W(x, y) is, as yet, unknown and represents the warping of cross
section planes.
In order to describe the general problem of noncircular sections, we now
turn to the shear strains and stresses on a face perpendicular to the shaft
axis. The shear strain-displacement relationships are given by
_ ou ceow
ex OZOXs
(5-21)
= ov Ow
(Ese Gy,
Substituting equations (5-20) into equation (5-21) and using the stress-
strain relationships

we obtain

T,, = GO| -—y + a

k x (5-22)
“|
to
Z)
EGU Oy

The sum of the contributions of the shear stresses on each element of the
cross section times the moment arm of the element makes up the total
torque. Measuring the x and y coordinates from the center of gravity of
the cross section, we have the torque

ie { (Tete iA (5-23)
A

where A is the entire area of the cross section. Substituting equations (5-22)
into (5-23), we obtain T = GJ@ where

a) =| [ + y* + Bs ~ i dx dy (5-24)
a oy Ox

An inverse or semiinverse method is sometimes useful in problems of

this type. That is, we assume the form of the solution and then find what
problem we have solved. The assumption must satisfy the equilibrium
equations. For example, we may assume displacement w to be proportional
to the product of x and y, that is,

W = cxy
where c is constant. We now examine the boundary conditions to find the
nature of the boundary for which this assumption could be valid. With no
normal stress in the z direction, the net force is zero on an element of the
 270
Chapter 5: Analysis of Stress and Displacement

lateral surface. Balancing axial shear forces on a surface element (Figure

5-22b) the result is
T,,(dy dz) + t,,(—dx dz) = 0
The negative sign preceding the differential dx does not imply negative
area, but accounts for the negative change in the x coordinate as sketched
in the figure. Noting that t,, = t,,, and so on, and substituting equations
(5-22), we obtain the boundary condition
7) A
| yok | d x + “)226 (5-25)
OX ey
for a general cylindrical boundary. If we substitute y = cxy for this example,
equation (5-25) becomes (c — l)ydy + (c + 1)x dx = 0. Integrating and
rearranging terms, we obtain

the equation of an ellipse. That is, we have solved the problem of torsion of
an elliptical shaft with semiaxes a and b, where
G\ alive b?
— a?
= —
l+ec ba erae
(See Figure 5-22c.) Thus,

> Xxy (5-26)

Evaluating equation (5-24), we obtain

J=(1 + af x?dxdy + (1 — of y? dx dy (5-27)

where the integrals are moments ofinertia/ ,and I,, respectively. Substituting
for c and using J, = na*b/4 and I, = nab? 4, we obtain

joe
ae
5-28
coe
for the elliptical shaft. Using this value of J, we obtain total rotation
_ (a? + b*)LT
(5-29)
~-a3b3G
We see from equation (5-27) that the constant J represents the polar moment
of inertia only when c = 0; that is, when the section is circular.
Shear stress is given by

= 2Ty a él MAES
ay nab? 2 mta°b
 JA
Section 5-7: Torsion

The maximum value of shear stress occurs on the surface at the least radius.
It is
2T
Tmax GAG; 0) = (S-30)
ma*b
where b > a for the elliptical shaft oriented as in Figure 5-22c.
The preceding analysis follows the work of St. Venant. The work of
Prandtl is also of importance in the study of torsion. The work of both
investigators is described by Timoshenko and Goodier [1].

Example 5-1. Compare the torsional stiffness and strength of a solid circular section
with a solid elliptical section having a semimajor axis twice the semiminor
axis. Let both torsion members be made of the same material and have the
same length and the same cross-sectional area.
Solution: Torsional stiffness of the two members can be compared by
examining the torsional spring constants of each member (the torque divided
by the angular deflection) where

as given by equation (5-19).

For two members of the same material and equal length, stiffness is simply
proportional to the J’s. Using equation (5-28) for the ellipse where b = 2a,
Reb 8)
= ee es
The area of an ellipse is zab, and thus a circle of the same area has a diameter
1D) =2 ab = 2a,/2
The stiffness of the circular section is proportional to
mD*
J = — = 2na*
3
that is, 25°% greater than the elliptical section.
If we let maximum shear stress equal a working stress in shear, we may
define torsional strength as T/t,,,,. From equation (5-30), for an elliptical
section, we have

where b = 2a.
For the circular section of the same cross-sectional area,
3
J = = s/n?
t max

representing 41% greater strength. @

St. Venant showed that a solid member ofcircular cross section is stronger
and stiffer in torsion than any other simply connected cross-sectional shape
 Dfz
Chapter 5: Analysis of Stress and Displacement

having the same area. The situation with regard to bending is entirely
different. i

Thin rectangular cross sections

For the case of torsion of a bar of length L with a thin rectangular cross
section 2a by 2b (where a « bas in Figure 5-23a), the total angle of twist is
approximated by
cliB
ae BEG
and the maximum shear stress occurs on the surface at the center of the

Points of maximum
shear stress

—b (a)

(b)

Figure 5-23 Torsion of thin sections. (a) Torsion of a thin

strip with cross section dimensions 2a x 2b. (b) Torsion of a
thin open section.
 eS,
Section 5-7: Torsion

long edges. It is approximately

eh
1b
Pee Sa2p
atx = +a, y = 0. (See Timoshenko and Goodier [1].)
The same approximation may be used for other thin open sections subject
to torque (such as angles, channel sections, and tubes that have a seam
which is not securely joined). In such cases, a is replaced by half the wall
thickness and b is replaced by half the length of the cross section measured
along the middle surface.
For the slotted tube of Figure 5-23b we have an approximate angle of
twist
je a 1
~ 2nGRt?
and approximate maximum shear stress
owl
Umaxs aecaiki>:
2nRt*

Example 5-2. Compare the torsional strength and stiffness of a hollow shaft with
that of a hollow slotted shaft of the same length and material. Both have a
2-in. O.D. and a 1.75-in. I.D.
Solution: As in Example 5-1, for the same material and length, torsional
stiffness of the sections is simply proportional to the values of J, and we will
again define torsional strength as the ratio of torque to maximum shear
stress.
For the closed section,
TD 16TD
gle WO gu DiesDA)
For torsional strength
if m(D* — D?) Ib-in.
= “ = 0.65
Ta 16D pe Ib/in.?
and torsional rigidity
IDG m(D* — D?) ate
Gb J) 39 = 0.65 in.

The slotted shaft cross section has the same polar moment of inertia, but,
due to warping, a much lower stiffness and strength. For the open section,
approximate values are
T = 2nR?? = 010307ic
ee 3 ee Ib/in.?
and
3

ie a = 0.00383 in.4
It is obvious that open sections ofthis type should be avoided when members
are designed to resist torsional loads. @
 274
Chapter 5: Analysis of Stress and Displacement

SECTION 5-8
Thermal Stress
When a machine member is heated to above room temperature, it will
tend to expand. If the temperature change is uniform, if the member is made
of a single material, and if there are no external restraints, then the unit
dimension change (true strain) will be constant throughout the member.
When these conditions obtain, there will be no stress due to the temperature
change. Expansion joints are commonly employed in piping carrying hot
fluids and in other structures subject to temperature changes for this reason.
On the other hand, stresses may be induced when temperature varies with
location within a member or by a temperature change in a member made of
dissimilar materials, or by a temperature change in a member with external
restraints. Since welding and other manufacturing processes involve non-
uniform heating and cooling, residual stresses often exist in machine members
at room temperatures. Residual stresses are sometimes relieved by annealing.
In some cases, residual stresses are desirable. One example is that of a
shrink fit of a hub on a shaft.
The coefficient of thermal expansion « expresses a rate of thermal strain
per unit temperature rise where a material expands freely. Thus, for a
temperature rise 7, the free strain is given by
&, = aT (5-31)

and the change in a dimension L by

A, =aTL (5-32)
if the part is unrestrained. The dimension L may be replaced by an inside
or outside diameter if we wish to find the change in that dimension. Since
dimension changes are small compared with the original dimension, volume
change per unit volume per unit temperature rise is approximately 3a.
Tables of values for the coefficient of thermal expansion are given in
references [7], [8] and [9]. A few average values are shown in the following
table.

Material Temperature Range, °F a, (in./in.)/°F

Structural steel (0.25% carbon) 68-572 7.22 X 10°

Machine steel (0.40% carbon) 68-572 6.7. X 10-6

Machine steel (0.40% carbon) 68-1112 7.94 X 10°6

High strength
cast iron (3% carbon) 68-572 5.8 X 10-6

Aluminum alloy (2024-T4) 68-212 3, 6 MO

Magnesium alloy (AZ80A-TS5) 68-212 14.4 X 10°6

eS
 275
Section 5-8: Thermal Stress

Values in metric units are obtained by multiplying by 1.8 Fahrenheit degrees

per centigrade degree. For example, the coefficient of thermal expansion for
structural steel is « = 13 x 10° °(cm/cm)/C° in the temperature range of
20-300°C.
Although « may be considered a constant for most engineering work,
careful measurements have shown a nonlinear rate of expansion in most
metals with increasing temperature. For example, reference [8] gives the
free strain in aluminum from 0°C to some temperature t°C as
= (2221 00141) 10"
in the range 10°C < t < 90°C. This will result in a final dimension
ee eee 21x 10 tO O114 10817)
where L, is the dimension at 0°C. To convert from Fahrenheit temperature
ty to Centigrade temperature t, we use
ip — oe
f= ==
1.8

Thermal stress in a restrained member

A temperature rise T in a member of length L will cause a free strain
éy = aT and a length change A, = «TL if the member is free to expand
(Figure 5-24a). Other dimensions increase proportionately. If the member is

a aTL

a (a) (b)
,

SS SSS

Tube (2)

(c) (d)
Figure 5-24 Thermal stress. (a) Free expansion due to temperature rise T.
(b) Restrained member. (c) Rod and tube subject to temperature rise T. (d) Several
elements at different temperatures or several elements of different materials.

subsequently forced back to its original length (Figure 5-24b) while the
elevated temperature is maintained, we have the equivalent of an elastic
strain —e, = —aT. The total effect is a net strain of zero and therefore no
 270
Chapter 5: Analysis of Stress and Displacement

change in length. There is also a resulting stress oy = — ET. Thus, in the

case of an ideally restrained member subject to temperature rise T, the true
strain and length change are zero and the thermal stress (the true stress)

Gz.= —EaT (5-35)

provided the member does not buckle.

Symmetric members of two or more materials

Many thermal stress problems arise in members that are neither ideally
restrained nor free to expand. When we join members with different
coefficients of thermal expansion or have varying temperatures within a body,
significant stresses sometimes develop. Consider a rod and tube combination,
joined at the ends (Figure 5-24c). If the two are subject to different tempera-
tures or if they are made of different materials and subject to a temperature
rise, thermal stresses will, in general, occur. The combination expands freely
but as a unit, having the same total expansion over length L. The true strain
is, therefore, the same for both

Sore ees (5-34)

and in each, the true strain depends on the free strain «T and the true stress

é; = 0,7, + a
Ey
(5-35)
o
Eo = Oy F, + 5

In addition, equilibrium of forces yields

0,A, + 0,A,=0 (5-36)

From equations (5-34) and (5-35), we have the true stress

OR et Sel aes ey)

(5-37)
02 = E,(é im a,7>)

where the true strain is given by substituting equations (5-37) into equation
(5-36) and rearranging.

(5-38)
A,E, + A,E,
If we have n elements joined so that there is no tendency toward bending
(for example, Figure 5-244), then stress is still given in the form of equations
(5-37)
o; = Efe — a;T,) (5-39)
 27
Section 5-8: Thermal Stress

and strain is given by

a! i=1 (AEaT);
5" (AB), oy)
The extension to a symmetric variation in temperature throughout the
thickness is illustrated in Example 5-3.

Example 5-3. Let temperature elevation vary through a uniform machine element
(Figure 5-25) according to the expression

Figure 5-25 A continuous temperature

distribution.

(independent of x and z). Find axial tensile stress, assuming zero tensile
stress in the y and z directions.
Solution: Extending equation (5-40) to integral form, the true strain is
J ExT dA
6 = ———_
JEdA
where we integrate over the cross section. Substituting dA = b dy and noting
that E and « are constant, we obtain

bEa f*_, T dy
bE ft,4y

Substituting the temperature distribution,

——
aT (* 1 —_—
ly|
— )

ves, | 2a d)
e OTs ve a y?

~ 2a [> ab fik4a

Local stress (from equation 5-37) is o = E(e — aT). Again, substituting the
temperature distribution
1
 Ass
Chapter 5: Analysis of Stress and Displacement

which results in a variation from

o = —{EaTy at the center

G=4Eah at the surface @

SECTION 5-9
Energy Methods—Castigliano’s Theorem
Certain types of problems, including those involving the deflection of
frames and curved beams, are conveniently treated by energy methods.
Before applying energy methods to solve design problems, we will examine
the concept of strain energy. Consider a slender bar subject to a tensile
force P. If the force is increased to produce an infinitesimal change in dis-
placement do, then, for an elastic system, the work done on the bar represents
an increase in strain energy: dU = P do. The total strain energy given by
the area below the P-6 curve of Figure 5-26a is U = f P do.

Figure 5-26 Strain energy. (a) Strain energy

in tension. (b) Strain energy for a linearly elastic
material.

The area to the left of the P-d curve will be called the complementary
energy: U, = J 6 dP. An infinitesimal increase in complementary energy is
given by dU, = 6 dP from which we may write 6 = dU,,/dP, which is applic-
able to nonlinear as well as linear problems. For most engineering work, we
would assume a linear force-displacement relationship, Figure 5-26b. Then,
 HD
Section 5-9: Energy Methods—Castigliano’s Theorem

strain energy is equal to complementary energy; U = U, and deflection

due to a tensile load P is given by

5 = —— (5-41)
Equation (5-41), called Castigliano’s theorem, is far more general in
application than indicated by the above discussion. Consider, for example,
strain energy in bending for a beam subjected to uniform moment M. The
relative rotation of the ends is given by 0 = ML/(EI) resulting in internal
energy U = 4$M’L/(EI) due to bending for the linear case. If M varies with
location, the corresponding energy expression is U = ${ (M?/EI) ds, where
integration proceeds over the entire member or structure. If a structure is
subject to bending and direct tension as well, we have
a Fe
U = af ds (5-42)
joypee Jor|
where M and F are local bending moment and tensile force, respectively.
Now, strain energy U may result from a complicated system of loads and
reactions, but if no work is done by the reactions and if the force system is
otherwise unchanged by the application of a given load, then equation (5-41)
becomes
oU
6 = ——
AP 5-43
(5543)

Equation (5-43) gives the deflection at a given load P in the direction of P.

Applying this equation to equation (5-42), the result is
MM*_ FF*
5=f |re | a (5-44)

where integration is carried out over the entire structure, M* = 0M/édP

and F* = 0F/0P. We see that M* and F™* are, respectively, moment and
force in members of the structure due to a unit load (P = 1 |b) applied at
the same location and in the same direction as the desired deflection 6. The
actual loads on the structure may be elsewhere; thus, the unit load, on which
M* and F* are based, may be called a “dummy load”’. Castigliano’s theorem
and other energy methods are treated in detail by Seely and Smith [10].

Example 5-4. Given: Load P, acts at point D as shown in Figure 5-27a. The moment
of inertia J about the neutral axis of each member in bending is the same and
all are made of the same material.
Find: Horizontal deflection 6 at point D.
Solution: We employ equation (5-44), integrating over the three members of
the frame. Moment M is given by Figure 5-27b. F = P, in BC; F = 0 in the
other members.
 280
Chapter 5: Analysis of Stress and Displacement

[-——--
(a) Frame subject to
horizontal load P,

b) M d (2 5 A
a a ae Figure 5-27 Application of
Castigliano’s theorem.

(c) Frame subject to

vertical load OQ

o—> §

For a unit load in the direction of 6 applied at D, the moment and force
distributions are similar: M* = y in AB and DC; M* = L, in BC. F* = 1
in BC
A L
Ay, 2 L
2 PTs 2 {L 2 IP _ P,Li[2
2
nos
5-2] dy + EI dx + , EA** = EI zhi t Lp +

Note that we have assumed small deflections, that is, the moments and
forces may be computed on the basis of the undeflected structure. @

Example 5-5.Given: The same frame as in Example 5-4, except that a load Q is
applied at the center of the top member as in Figure 5-27c. Find: 6, the
deflection of point D.
 281
Section 5-10: Elastic Stability

Solution: Moment due to Q is given by

Ox OL, — x
ae Moo = CE 9
O0<x<L3/2 2 L2/2<x<L2 9)

in member BC. M = Oin AB and CD. The force in AB and CD is F = —Q/2;

F = 0 in BC. The effect of a unit load in the direction of 6, applied at D, is
the same as in Example 5-4: M* = yin AB and CD; M* = L, in BC. Thus

6= Pe OxE; dx i
eee ee!
2 O(L, — x)L, d
——————— IDG SS
QL,L}
0 2EI L/2 2EI 8EI
The compressive force in AB and CD does not affect 6 in this case. Stability
must be examined separately. @

Example 5-6.Given: A frame loaded as in Example 5-5 except that point D is a

fixed pin. Find: The horizontal reaction at D.
Solution: This problem is statically indeterminate to the first degree. Since
6 = 0, we may equate the deflection 6 of Example 5-4 plus that of Example 5-5
to zero, from which we obtain the reaction
—QL,L3
le 273 2
8[3Li + LyL2 + (L21/A)]
There is an equal and opposite reaction at A.
Instead of the above superposition method, however, we may determine
the reaction directly from equation (5-44) by settingd=0. @

The above examples illustrate the effectiveness of Castigliano’s theorem

if we are to find deflection at a single point on a complex frame or machine
part. The penalty associated with this method, however, is that we find
deflection at only one point, whereas the classical method of successive
integrations yields slope and deflection at all points.
In order to find angular deflection (slope) we proceed as in the Examples
but differentiate strain energy with respect to moment, obtaining
MM FF’
g -| mf al gs
Quantities M and F are moment and force due to actual loading as before,
but M’ and F’ are moment and force due to a unit moment applied at the
location where we must find 0.

SECTION 5-10

Elastic Stability
Elastic stability concerns buckling of machine elements due to compressive
load.

Slender columns
Most analysis of machine elements is based on a linear relationship
between loading and deflection and on a linear relationship between loading
 282
Chapter 5: Analysis of Stress and Displacement

and stress. Thus, in most design problems, we do not expect a rapid increase
in deflection with small changes in load, providing we do not exceed the
yield point of the material. Buckling of slender columns is an important
exception to this usually valid relationship. Stability, in the sense of this
discussion, implies that a load increase will result in a redistribution ofstress
so that internal stresses balance external forces and moments. When deflec-
tions cause an increase in the effect of external forces, thereby increasing
moments so that equilibrium cannot be reestablished, we have instability.
The load at which instability occurs will be called the critical load P.,.
Critical load is sensitive to support conditions as well as to the geometry
and elasticity of the member.
Consider a member having constant cross section A and subject to a
compressive load P (Figure 5-28a). The member is simply supported (pinned)

ee
(b) w
Pp Su

Load P
Pp
cr

(c)

Deflection w 6)

Load P

(d)

Deflection w 6)

Figure 5-28 Elastic stability. (a) Simply supported

compression member. (b) Momentum equilibrium. (c) Load-
deflection relationship (small deflection theory). (d) Load-
deflection relationship (large deflection theory).
 283
Section 5-10: Elastic Stability

at both ends, but axial motion of one end is unrestrained. Let the moment
of inertia about the y axis

i= I 2 dA (5-45)
A

be the least moment of inertia. Then, buckling will tend to occur in the xz
plane. Let us assume that the load P has caused a deflection w of the neutral
axis of the member at a distance x from the left end (Figure 5-28b). The part
of the member to the left of this point will be in equilibrium if internal
stresses at x result in a moment

M = Pw (5-46)
on that section. If we utilize the moment-curvature relationship,
d*w
FED
M = —EI—, (5-47)
5-47

and rearrange terms, the stability equation is obtained.

d’w = =Pw
ee 2) 5-48
bee Jom 2)
The general solution to equation (5-48) is

P le
w=C, cos,/"? + C, sin, (5-49)

where C, and C, are arbitrary constants. The boundary conditions that

have not already been used are w(0) = 0 and w(L) = 0. The first yields
C, = 0and the second yields

0 == C,sin./
sin =a Lc
2 EI
The solution C, = 0 would be trivial, so we have

sin
1 ean = 0
EI
==

from which we obtain the critical load, the lowest value of P that will cause
the simply supported column to fail by buckling
nEI
ES a (5-50)
for the simply supported member. This is also called the Euler load.
For an ideal, initially straight member, there is no deflection until the
critical load is reached. Using C, =0 and P = P., (equation 5-50) in
 284
Chapter 5: Analysis of Stress and Displacement

equation (5-49) we obtain

a meer
w = C, sin Ler

Bie
or w = C, sin Ta

that is, the member bends in the shape of half a sine wave at the critical
load. The amplitude C, = w (at x = L/2) can theoretically assume any value
(Figure 5-28c).
The results obtained in the preceding discussion are based on “‘small
deflection” theory, that is, the assumption that the slope of the deflected
neutral axis is small compared to unity. A more rigorous (though not
necessarily better) load-deflection relationship is obtained by admitting the
effects of large deflections (Figure 5-28d). Deflection begins at the same
load, P.,, but large deflections are obtained only by an increase in load.
Loads at or near P.,, however, cause large enough deflections to result in
failure of machine parts in most cases. Thus, the results of small deflection
theory are usually adequate, and we may allow loads up to

ee
N
for a slender column where N is an appropriate factor of safety.
For cross section area A, the Euler load P.,.as determined above is certainly
not valid when P.,/A exceeds S,,,, the yield point of the material. Experi-
mental data have suggested that its use be limited to the range
P cr S yp
a
A (5-51)
2
Slenderness ratio
The ratio of unsupported column length L to least radius of gyration
r, 1s called the slenderness ratio. Radius of gyration is given by r, = ./1/A,
where I is the least moment of inertia. Using inequality 5-51 and equation
(5-50), we limit the range of validity of the Euler load to slender columns for
which

L 2n’°E
aa
iP Sap
(5-52)

Short columns
Simply supported compression members for which

Eee
ie
iat
Ayes
(5-53)
 285
Section 5-10: Elastic Stability

may be called short columns. Tests of short columns suggest the relationship
P,, Cie
Siw eas AR | ae
called the Johnson formula.
Hollow sections and I sections have a greater moment of inertia than
solid circular or rectangular rods of the same area and a correspondingly
greater critical load. However, we must guard against local buckling or
crippling of thin sections.
The results obtained thus far in this section are based on the simply
supported member, which is an idealization based on the condition of no
resistance to bending moment at the ends. This idealization may be used
in cases where each end of a compression member is fastened with one or
more bolts or rivets. If several fasteners are used, the joint will be somewhat
more rigid than the assumed simple support, but not rigid enough to be
considered clamped. Thus, we err on the safe side.

Example 5-7. Find the axial load that may be safely carried by a l-in. x 1-in. x §-in.
thick steel angle, 50 in. long, used as a brace as in Figure 5-29. The yield point
is 50,000 psi and a factor of safety of 2 will be used.
Solution: We will ignore the moment resistance of the end supports and
approximate the actual brace by a simply supported member of unsupported

(a)
Figure 5-29 Angle brace compression member. (a) Angle brace.
(b) The moment of inertia about the n axis is a minimum.
 286
Chapter 5: Analysis of Stress and Displacement

length L = SOin. Using a table of properties for steel sections (for example,
US. Steel Corporation, Hot Rolled Steel Shapes and Plates [5]), we obtain
the cross sectional area of the angle: A = 0.23 in? and the radius of gyration
about the x, y, and n axes: r, = r, = 0.31 in. andr, = 0.19 in., the minimum
radius of gyration. Buckling will tend to occur about the n axis corresponding
to the least moment of inertia
[=I = Ar? = 0.0083 in.*
Using the modulus E = 30 x 10°, the inequality for slender columns
Ie 2n?E
2
Ph Syp
is satisfied.
The critical load is given by equation 5-50
n*EI
= poe 980 Ib

and using the factor of safety, we have a. safe compressive load

P

Eb

For this long, slender member, the buckling strength will undoubtedly govern
if compressive loads are anticipated. In general, however, we must check to
see that the working stress of the member is not exceeded at the least section
(at the holes); we must be sure the fasteners do not shear ;and we must check
our design for failure of the fasteners or the member in bearing. @

Eccentrically loaded columns

The load on the column in Figure 5-30 has an initial eccentricity e and
the column design may be based on the secant formula

ipa ic ay 5-55
=

Toae PN AAE =

Figure 5-30 Eccentrically loaded

column.
 287
Section 5-10: Elastic Stability

where P is the allowable load with factor of safety N; A is cross-sectional

area and radius of gyration r, = ./1/A. If the loading eccentricity would
tend to cause failure in one plane and the minimum moment of inertia
would tend to cause failure in another plane, it is necessary to check both
possibilities. Equation (5-55) may be applied to short as well as long columns,
but it is necessary to determine (or to assume) the initial eccentricity in

_ El
Cra jp. P=

Q
Mo i=
1o

(c) A (d)

p. —4nE/
cr L2 Pp _ 20.19E/
Chae 2

L L

Figure 5-31 Slender columns with various boundary conditions. (a) Simple
supports (pinned-pinned). (b) Clamped-free. (c) Clamped-clamped (axial motion
only). (d) Clamped-pinned. (e) Clamped-clamped (lateral motion at one end, but
no rotation). (f) Pinned-pinned with intermediate, moment-free support.
 288
Chapter 5: Analysis of Stress and Displacement

Rigid member

(f)

4n2E/

Figure 5-31 (continued)

every case. Because of its unwieldy form, column design by this formula
might best be programmed on a computer.

Other boundary conditions

Some compression members cannot be represented by the idealization
of simple supports. Increased restraint results in greater load carrying
capability when elastic stability governs the design. Slender compression
members with other boundary conditions are compared with the simply
 289
Section 5-10: Elastic Stability

supported compression member in Figure 5-31 (page 287). In each case, the
moment of inertia J] represents the minimum moment of inertia. In solving
problems of this type, it is important that all forces and reactions be considered.
The simply supported compression member (Figure 5-31a) could have no
lateral reaction component; if such a force were present at one end, it would
require a compensating moment at the other end. For the slender clamped-
pinned member (Figure 5-31d), a lateral reaction is possible and we must
consider its effect on moment equilibrium. We proceed as with the simply
supported member, measuring the coordinate x from the pinned end. The
governing equation for the clamped-pinned member is
d*w
hi ri O. (5-56)
dx?
The complementary solution to equation (5-56) is the same as the solution
for simple supports, and the particular solution is
Ox
WwW
Pp

Thus, deflection of the clamped-pinned member is given by

IP Mea
w = Acos [ay* + Bsn es (5-57)

In deriving the governing equation, we used the condition of zero moment

at x = 0, the pinned end. The remaining boundary conditions are
w(0) = 0

dw

w(L) = 0
which lead to the transcendental equation
(iain AC = OE (5-58)
where

cea ee
oe EI
Equation (5-58) can be solved by plotting values of tan X and X (in
radians) as in Figure 5-32. The lowest nontrivial solution, X = 4.4934 leads
to the critical load

COED
Choe L?

for the slender clamped-pinned compression member. The roots of

tan X — X = Oare also given by Jahnke and Emde [6].
 290
Chapter 5: Analysis of Stress and Displacement

Figure 5-32 Solution to X = tan X where

Nee (Ea):

SECTION 5-11
Numerical Methods
This section deals with numerical methods, using digital computer
programs. Problems in the design of machines do not always lead to simple
analytical solutions. Two numerical methods for obtaining roots of equations
will be considered at this point. Both methods are effective for digital
computer use and applicable to many different types of problems. The
clamped-pinned compression member described in the previous section will
be used as an example.

Newton’s method
This numerical method, also called the Newton-Raphson method, is a
common technique for finding a root of an equation. Suppose we require a
root of the equation
F(X) =0

F(X)
Line tangent to
F(X) atX

lees Figure 5-33 Newton’s method.

Root of F(X) = 0
 |
Section 5-11: Numerical Methods

where F(X) is plotted in Figure 5-33. (Of course, we do not have the actual
plot; if we did the solution would be obvious.) Selecting a value X for the
first approximation of the root, we calculate F(X), which happens to be
unequal to zero. A tangent to the curve, however, intercepts the X axis at
X NEW, a better approximation of the root. The new approximation is
given by

XNEW = X Ee)
TES
where

dF
G(X) = ax)

the slope of the curve at X. If the new approximation X NEW is not good
enough, we replace the old value of X by X NEW and repeat the process.
Figure 5-34a shows a computer flowchart for Newton’s method. Our
input includes X, the first approximation of the root; ERROR, the difference
between two successive solutions at which we will terminate the program;
and N, the maximum number of trials. For the function sketched in Figure
5-33 we would expect rapid convergence toward the root of F(X) = 0.
Unfortunately, Newton’s method fails us in certain circumstances. These
circumstances include the presence of local extremes (maxima or minima) or
the presence of inflection points in the region of interest. Furthermore, we
must obtain the derivative of the function, and it should not be zero in the
region of interest if Newton’s method is to be applied. Newton’s method
tends to be particularly troublesome with transcendental functions. When a
function has multiple roots there is always the hazard of failing to obtain
the desired root by numerical methods, particularly by machine methods.
Figure 5-34b is a computer program that was used to find the root of
tan X — X = 0 (that is, F(X) = 0) within +0.00001 by Newton’s method.
For a first approximation of X = 4.6, the program computed successive
values of X until the solution X¥ = 4.493409 was reached. This is the solution
to tan X = X of the preceding section where

poms
‘a EI
and P is the critical load for the slender clamped-pinned compression
member. This solution is more precise than necessary for the physical
problem.
We see from Figure 5-32 that a first approximation which is not near the
desired root will not lead to the correct solution. For this equation, use of
a poor first approximation with Newton’s method led to wide oscillations.
This is due to the nature of the derivative of F(X) = tan X — X and due
to the fact that tan X has an infinite number of branches.
 292
Chapter 5: Analysis of Stress and Displacement

READ X,
ERROR, N

IX — XNEW|
<ERROR Figure A-34 (a) Newton’s method (see
also next page).

OUTPUT: OUTPUT
‘NO ROOT IN ‘ROOT OF
N TRIALS’ F(X) =O’

The secant method

While similar to Newton’s method, the secant method utilizes a difference
quotient instead of a differential. We use a secant line, Figure 5-35 instead
of a tangent line. As with the previous example, the figure is used to illustrate
a numerical method; in general, we would not plot F(X). It is not necessary
 293
Section 5-11: Numerical Methods

2 PRINT, 1,FX,X
GOT06
] I=I+]
IF(I-N)3,3,5
3 X=XN
GOT04
5 PRINT,1
6 CONTINUE
SLOP.
END
Figure 5-34 (continued) (b) Newton’s method, computer program
listing.

Secant line Sy Figure 5-35 Secant method.

SM —<———$—<$—

af XNEW X2 x1
Root of F(X) = 0

to code a derivative when using the secant method, but two initial approxi-
mations of the root of F(X) = 0 are required. Let X1 and X2 be the initial
approximations. We will take the secant line intercept with the X axis to
be a better approximation of the root. The new approximate root is
F(X2)
DOING me
G(X2)
 294
Chapter 5: Analysis of Stress and Displacement

where

E(X2y C1)
OOS Saal
the difference quotient. If XNEW is not a satisfactory approximation, we
replace the old value of X2 by XNEW and the old value of X1 by X2 and
then repeat the procedure.
For problems of this type the desired root should be obtained rapidly if
the initial approximations ofX bracket the expected solution. For the solution
of tan X — X = 0, both initial approximations of X should be on the same
branch of tan X.

PROBLEMS
1. The steel I section shown in Problem Figure | has a yield strength of 45,000 psi.
Find the factor of safety based on normal stress at the outer fiber.

1500 Ib 1500 Ib i

aa SS eee eee) Se es 2

Flanges are iinch thick

Web is : inch thick

Problem Figure 1

2. (a) Find the neutral axis and moment ofinertia of the T section of Problem Figure 2
in terms of thickness ¢. (b) Find required thickness t for a yield point of 60,000
psi, maximum moment of 4000 in.-lb, and a factor of safety of 4.

ES Problem Figure 2
 295
Problems

3. Repeat Problem | for the case where only the 15001b load Sin. from the left
support is applied.
4. Repeat Problem 2 for the case where the horizontal part of the T section has a
width of 6t.
‘5. (a) Find tensile stress at point A in Problem Figure 5. (b) Find tensile stress and
transverse shear stress at point B.

5000 Ib
fo>)oO 5

|
N—

Problem Figure 5

6. Repeat Problem 5 for a similar I section with }-in. thick web and flanges.

7. A 20-in. long steel bar is pinned (simply supported) at the ends. A 5000 Ib load is
applied 10 in. from the left end and a 6000 lb load 12 in. from the left end. Both
loads are applied vertically downward. If we use a square bar with a yield point
of 80,000 psi and a safety factor of 2, find the sectional dimension of the bar.
8. Ifa round bar is to be used in Problem 7, find the required diameter.

9. A nylon machine element is subject to a maximum moment of45 in.-lb. (a) Using
a safety factor of 3 and a tensile strength of 7100 psi, find the sectional dimension
if the member is to be square. (b) Find the required diameter if a round member is
used.
10. (a) Find the maximum deflection due to bending in the steel I section shown in
Problem Figure |. (b) Find bending deflection 5 in. from the left support.
11. A steel machine member is subject to a 1000-Ib load as shown in Problem Figure 11.

Problem Figure 11
 296
Chapter 5: Analysis of Stress and Displacement

The vertical part may be considered rigid. (a) Draw and dimension shear and
moment diagrams. Find bending stress o,, at B and C. Neglect stress concentration.
(b) Draw the curvature diagram. Dimension it. (c) Assume zero slope at point C.
Plot @ versus x showing values at A and B. What is the maximum slope in radians?
12. Repeat Problem 10 for the case where only the 1500-lb load 5in. from the left
support is applied.
13. Repeat Problem 11 except that the 1000-lb load at point A points upward and to
the left at 45 deg. Find co, at B and C due to both direct tension and bending.
14. A steel band 0.032-in. thick and 0.75-in. wide is bent around two pulleys, each of
radius R = 8in. The band is subject to a maximum direct tensile load of 200 lb.
(a) Find maximum tensile stress. Note that curvature d*w/dx? = 1/R. (b) Plot
stress across the thickness of the band where it bends to go around the pulley.
15. In Problem 14, find the optimum band thickness based on maximum stress,
that is, find thickness for which maximum stress will be least.

16. A 6-in. long machine member is simply supported at the ends. Stiffness EJ, = 2000
lb-in.? for the left half and EI, = 1000 lb-in.? for the right half. A 60-Ib vertical
load is applied 2 in. from the left end and a 40-lb vertical load is applied at the
center. Find shear, moment, curvature, slope, and deflection. Evaluate slope at
the ends and maximum deflection.
17. The rigidity of a 6-in. long machine member is EJ, = 10,000 lb-in.? from x = 0
(the left end) to x = 4. The rigidity is EJ, = S000 to the right of x= 4. The member
is simply supported at x = 1 and x = 3. A 10-lb overhung load is applied at x = 5.
Find shear, moment, curvature, slope, and deflection.
18. A 14-in. long circular shaft is simply supported at the ends. At x = 4 (measured
from the left end) there is a 3000-Ib vertical load and a 1000-lb horizontal load.
At x = 10, there is a 2000-lb vertical load and a 3000-lb horizontal load. Both
vertical loads are downward and both horizontal loads are outward. Sketch
shear and moment in the vertical and horizontal planes. Evaluate maximum
resultant moment at x = 4, 7, and 10.
a9: A steel bar has dimensions + x 5 x 12 in. It is simply supported at x = 1, 6, and
11,(measuring from the left end). A 100-lb vertical load is applied at x = 3.5 and
another at x = 8.5. (a) Find the reactions. (b) Find maximum bending moment.
20. A 5-in. long bar of polycarbonate is clamped at one end and has a 20 1b load
at the other perpendicular to its axis. (a) Find the required bar diameter based on
a safety factor of 1.8. Use average properties. (b) Find deflection at the loaded end.
21 - Repeat Problem 20 for a hollow polycarbonate bar with inner diameter = 80°% of
outer diameter.
22. (a) Find shear deflection and total deflection at the center of the I section shown
in Problem Figure 1. (b) Repeat for a point 5 in. from the left support.
23. Repeat Problem 22 for the case where only the 1500-lb load Sin. from the left
support is applied.
24. A solid rectangular cantilever beam has a width h and a depth 2h. A vertical
force is applied at the free end. The beam is made of steel with a shear strength of
one-half its tensile strength. (a) For what beam length is tensile failure most likely?
(b) For what length is failure in transverse shear likely?
25. A 4 x 23 channel section is oriented as in Figure 5-19 and subject to vertical
loads. Total depth = 4 in., total flange width = 2.5 in., and average flange and web
 27,
Problems

thickness = 0.5 in. The moment of inertia about the neutral axis is 8.8 in.* Find
the shear center.
26. (a) Draw shear and moment diagrams for the 1-in. bar in Problem Figure 26.
(b) Find bending and shear stress at points B and C on the 1-in. bar.

lin. dia.

Problem Figure 26

27. Find the angular rotation and vertical displacement at point B on the 1-in. steel
bar in Problem Figure 26.
28. Repeat Problem 26 with dimension AB = 8 in. and dimension BC = 4 in.
29. Repeat Problem 27 with dimension AB = 8 in. and dimension BC = 4 in.
30. An elliptical aluminum bar has a 1-in. major diameter and a ?-in. minor diameter.
The bar is 14 in. long and subject to a torque of 125 in.-Ib. (a) Find maximum shear
stress and its location. (b) Find angular deformation.
31. (a) Find the torsional rigidity of a 9-in. long open steel section similar to Figure
5-23b with an outside diameter of 2in. and 0.10-in. wall thickness. (b) Find
maximum shear stress if 100 in.-lb torque is applied.

32. Find the approximate torsional spring constant and maximum allowable torque
for a 60-in. long steel channel section with a 4-in. web and 2-in. legs, having 4-in.
section thickness throughout. Allow a working stress of 20,000 psi in shear and
use G = 11.5 x 10° psi. Find the diameter of a solid steel rod 60 in. long (a) with
the same stiffness;(b) with the same strength.
33. Find the torsional spring constant of a solid stepped steel shaft. Its diameter is
lin. for 8 in. of its length; 1.25in. for 5 in. of its length; 0.875 in. for 6 in. of its
length.
34. A 5-in. diameter steel bolt is tightened snug in a 10-in. long aluminum sleeve with
1 in.? cross section. Find the force in each part after a temperature rise of 100F°.
Use temperature coefficients of 6.5 x 107° for steel and 13 x 107° for aluminum.
35. Consider a solid unconstrained aluminum alloy bar with dimensions and co-
ordinates as in Figure 5-25. Find the thermal stress after a temperature change of

T = Tp v1)
2a
where a = 4 in. and Ty = 190F°.
 298
Chapter 5: Ana lysis of Stress and Displacement

36. Consider a composite bar similar to Figure 5-24d which is subject to a 210F°
temperature rise. Let layers 1 be 40°% carbon machine steel, 0.10-in. thick, layers 2
magnesium alloy, 0.20-in. thick, and layer 3 40°% carbon machine steel, 0.25-in.
thick. The bar is 1.5-in. wide and 8-in. long. Find elongation and stresses.
37. Find the deflection at the center of the stepped shaft in Problem Figure 37 where
I, = 1,. Use Castigliano’s theorem. ‘

Problem Figure 37

38. Find deflection at the load and at the centér of the stepped shaft in Problem
Figure 37 if J, = 3 and only one load is applied. Use Castigliano’s theorem.

39. Using Castigliano’s theorem, find deflection at point A in Problem Figure 39.

Problem Figure 39

40. Using Castigliano’s theorem, find deflection at point B in Problem Figure 39.
41. Let load P be applied vertically upward at point A in Problem Figure 39. Find
deflection at A if the section is constant throughout the member.
42. A 3-in. I beam (Problem Figure 42) is loaded in compression. A = 1.64 in.’:
I,, = 2.5in.*; I,, = 0.46 in.* There is a 72-in. unsupported length between ends
that may be considered pinned. Find the safe load based on a factor of safety of 3.
Use steel with S,,, = 42,000 psi.

1 ———|+|——— 1 Problem Figure 42

43. Repeat Problem 42 for an unsupported length of 100 in.

 299
References

44. A 3-in. steel I beam (Problem Figure 42) with 36-in. unsupported length may be
considered to have pinned ends. A = 1.64in.?; J,, = 2.5in.+; I,, = 0.46 in.*
Find the safe load based on a safety factor of 3. Use S,, = 42,000 psi.
45. Brace A in Problem Figure 45 is subjected to an axial compressive load F. The
angles do not twist, but the brace may rotate about its fasteners at both ends.
Find allowable load F for a factor of safety N.

Problem Figure 45

46. Find the second buckling mode of the pinned-clamped compression member
using Newton’s method. Does this value have any physical significance?
47. Find the second nonzero root of tan X = X using the secant method. Sketch a
flowchart for the numerical procedure.

REFERENCES
[1] S. Timoshenko and J. N. Goodier: Theory of Elasticity, 2nd ed. McGraw-Hill
Book Co., New York, 1951.
[2] S. Timoshenko: Strength of Materials, 3rd ed., 2 vols. D. Van Nostrand Co.,
New York, 1955.
[3] F. E. Miller and H. A. Doeringsfeld: Mechanics of Materials, 2nd ed. International
Textbook Co., Scranton, Pa., 1962.
[4] E. P. Popov: Mechanics of Materials. Prentice-Hall, Inc., Englewood Cliffs, N.J.,
1952.
[5] Hot Rolled Steel Shapes and Plates. U.S. Steel Corporation, Pittsburg, Pa., 1955.
[6] E. Jahnke and F. Emde: Tables of Functions, 4th ed. Dover Publications, Inc.,
1945.
[7] V. M. Faires: Design of Machine Elements, 4th ed. The Macmillan Company, New
York, 1965.
[8] C. D. Hodgman, et al.: Handbook of Chemistry and Physics, 37th ed. Chemical
Rubber Publishing Co., Cleveland, Ohio, 1955.
[9] O. W. Eshbach, et al.: Handbook of Engineering Fundamentals, 2nd ed. John Wiley
& Sons, Inc., New York, 1952.
[10] F. B. Seely and J. O. Smith: Advanced Mechanics of Materials, 2nd ed. John Wiley
& Sons, Inc., New York, 1952.
[11] D. D. McCracken: A Guide to FORTRAN IV Programming. John Wiley & Sons,
Inc., New York, 1965.
 300
Chapter 5: Analysis of Stress and Displacement

[12] B. A. Boley and J. H. Weiner: Theory of Thermal Stresses, John Wiley & Sons,
Inc., New York, 1960.
[13] R. W. Hamming: Calculus and the Computer Revolution. Houghton-Mifflin Co.,
Boston, 1968.
[14] H. M. Priest: Design Manual for High Strength Steels. U.S. Steel Corporation,
Pittsburgh, Pa., 1954.
[15] S. Timoshenko and J. M. Gere: Elastic Stability. The McGraw-Hill Book Co.,
New York, 1961.
Theories of Failure
Used in the Design of
Machine Elements

SYMBOLS

N = factor of safety Sc) = ultimate strength in

S, = endurance limit (modified compression, psi
for actual specimen), psi S,, = tensile yield strength, psi
S, = fatigue strength for n Sy pic) = Compressive yield
cycles, psi strength, psi
S;, = endurance limit (for 01,0 2,03 = principal stresses, psi
polished test specimen), O., = equivalent stress, psi
psi Om = Mean stress amplitude, psi
S, = ultimate strength in o, = range stress amplitude, psi
tension, psi Tmax = Maximum shear stress, psi

In many cases, a machine part fails when the material begins to yield
plastically. In a few cases, we may tolerate a small dimensional change and
permit a static load that exceeds the yield point. Actual fracture at the ultimate
strength of the material would then constitute failure. The criterion for
failure may be based on normal or shear stress in either case. Fatigure failure
is probably the most common mode of failure in machine elements. Other
modes of failure include excessive elastic deflection of some part, thereby
rendering the machine useless, or failure of a part by buckling. The actual
failure mechanism in a real machine element may be quite complicated ;
each failure theory is only an attempt to model that mechanism for a given
class of materials. In each case, a factor of safety is employed. The magnitude
of the safety factor depends on the probable accuracy of our assumptions,
the danger and cost of a failure, and other design conditions.
301
 302
Chapter 6: Theories of Failure Used in the Design of Machine Elements

SECTION 6-1
Stress at a Point and Combined Stress
Figure 6-1 shows the convention for designating stresses on an element
with faces perpendicular to the coordinate axes. Every machine element is,
of course, three-dimensional. It is unusual, however, to have stresses of
significant magnitude in all three directions. A two-dimensional stress
field adequately describes all but a few engineering problems.

dx ;

thy +y face dz

+y +z
Coordinate axes

(a)

OZ

Tzx

0y "yx

Te

(b)

Figure 6-1 Stress at a point. (a) An infinitesimal element in a machine

member. (b) Designation of normal and shear stresses (positive as shown;
not all stresses shown).

Combined stress in a two-dimensional stress field

Consider an element within a two-dimensional stress field such that the
only nonzero stresses are o,, 0,, and t,, (Figure 6-2). From equilibrium of
 303
Section 6-1: Stress at a Point and Combined Stress

Figure 6-2 Stress at a point in a two-dimensional stress

field.

forces, the normal stress on the skew face is given by

6d, = 6, COS” 6 + 21,, sin 6 cos 6 + o, sin? 8 (6-1)

and the shear stress by

— o,)sin 0 cos 0 + 7,,(cos? 6 — sin? 6) (6-2)

If the expression for normal stress is differentiated with respect to 0 and

the result set equal to zero, we obtain the values of @ (the principal directions)
corresponding to maximum and minimum normal stress in the xy plane.
These are the principal stresses
2 it//??
Oi
ar @}
6, and a, — <5 2+ | + 72 | (6-3)

The third principal direction is the z direction and the corresponding stress
0, = 03 was assumed to be zero. The principal directions (1, 2, and 3) are
mutually perpendicular. The planes perpendicular to the principal direction
axes are the principal planes.
Transformation of stresses is discussed in detail in books on mechanics ai
materials including Douglas [1] and references [1] to [4] of Chapter 5.
The expressions for normal and shear stress, equations (6-1) and (6-2), can
be written in terms of the double angle 20 and expressed graphically. This
plot of stress at a point is called the Mohr circle. For a two-dimensional stress
field, the Mohr circle is constructed as follows (see Figure 6-3).
 304
Chapter 6: Theories of Failure Used in the Design of Machine Elements

T axis

03 =0

Figure 6-3 State of stress at a point (Mohr circle

representation of stresses in the X Y plane for a
two-dimensional stress field).

. Plot o x? Tyy and 6,, —T,, On d,, T coordinates.

. Connect the two points to form the diameter of the Mohr circle. Draw
the circle.
. The line from the first point to the center of the circle represents the
x axis; the line from the second point to the center represents the y axis.

. to an angle 6 measured on the actual element as in Figure 6-2. Points

on the circle represent stresses in the xy plane.
. Identify principal stresses o, and o,.
. Identify the coordinate origin as a, (where o, = 0). See Figure 6-4.
Draw another circle with the line between o, and o, as its diameter;
draw another with the line between o, and oa, as its diameter. The
region between the circles represents all possible states of stress at a
point in the two-dimensional stress field. Maximum shear stress is
given by the radius of the largest circle, that is

lo, — ol

=) the largest ofi (6-4)

|n|2
 305
Section 6-1: Stress at a Point and Combined Stress

T axis

Region of all
possible states
of stress
(shaded)

oO, axis

Figure 6-4 Completed construction showing

maximum shear stress (case where o, and a, have
same sign; 03 = 0).

The three-dimensional case

In the general three-dimensional case, we have nonzero values of c,, T,.,,
and t,,. It can be shown that the three roots of the determinate
C—O aioe Gi
oan Gir 30 tee = 0 (6-5)
re ce o,—6
which we label o,, ¢,, and a; are the principal stresses in the general three-
dimensional case [Boresi, 2]. Shear stress is zero on the three corresponding
planes, the principal planes. Figure 6-5 shows a Mohr circle representation
for the general three-dimensional case, where circle diameters are distances
T{G7, 7303, and 5,05.
Maximum shear stress is given by the greatest circle radius. In general,
we have
lo, — 9)
2

lo. — a3}
S max = the largest of 5 (6-6)

lo, — a3]
Z
 306
Chapter 6: Theories of Failure Used in the Design of Machine Elements

T axis

Region of all possible

states of stress
(shaded)

Figure 6-5 State of stress at a point for general three-

dimensional stress field.

Principal stresses ¢, and co, will not, in general, be in the xy plane. Thus,
the construction procedure of Figures 6-3 and 6-4 may not be used in the
general three-dimensional case.

SECTION 6-2
The Maximum Normal Stress Theory
The maximum normal stress theory is based on failure in tension or
compression. It may be applied tomaterials that are relatively strong in
For omens of static loading, we designate the largest of the positive
principal stresses by o,,,,, and the magnitude of the most negative principal
stress by O(.max- For a design based on yielding, with tensile and compressive
yield points S,, and S ypc)? respectively, we have
S S
Shi = a and ihe (6-7)
for safety factor N. If we have a static load problem where yielding is of
no
consequence, or design for a material with no distinct yield point, ultimate
strengths in tension and compression, S,, and S,;,,, maybe used. The design
 307
Section 6-2: The Maximum Normal Stress Theory

criterion then becomes

ute) (6-8)

The maximum normal stress theory is

ine elements. Cast iron has no distinct yield point, and the ultimate
compressive strength is considerably greater than the ultimate tensile
strength. The stress-strain relationship is nonlinear. Bending stresses are
not given exactly by the equation
a Mz (6-9)
ne a -

and the ultimate strength based on

Mc
O x(max) a Sie (6-10)

will not equal the ultimate strength based on direct tension. For simplicity,
however, equations (6-9) and (6-10) are generally used.

Example 6-1: The Maximum Normal Stress Theory. The cast iron section
shown in Figure 6-6 has an ultimate strength in tension S, = 30,000 psi and
an ultimate strength in compression S,,., = 109,000 psi. A factor of safety
N = 2 will be used. Plot stress distribution across the section and find
allowable load P.

Solution: The solution will be based on stress on section A-A. The distance
to the neutral axis in bending is given by

ZA
Z=
yA

summed over the two parts of the T section, where z’ is used to represent
distance from the left edge of the T section to the center of a given part of the
T section. We find

, _ (1.75)(0.25)(1.125) + (1.2)(0.25)(0.125) _ 9 a5;

am (1.75)(0.25) + (1.2)(0.25) Pave
as shown in Figure 6-6.
The moment ofinertia about the neutral axis in bending is given by

I = X(Ip + Ad?)
summed over the two parts of the section. The term J is the moment of inertia
of a part of the T section about an axis parallel to the neutral axis in bending,
but through its own center of gravity. Distance d is measured from that axis
to the neutral axis in bending. Thus, we have

I = (0.25)(1.75)3/12 + (1.2)(0.25)3/12 + (1.75)(0.25)(1.125 — 0.72)

+ (1.2)(0.25)(0.72 — 0.125)? = 0.288 in.4
 308
Chapter 6: Theories of Failure Used in the Design of Machine Elements

qu ;

x P

(a) T section
machine
member

Figure 6-6 Design ofa cast iron section (n.a. is

the neutral axis in bending).
Section 3
AA 2:

Z
Section thickness 0.25”

6.85P
(tension)

15.96P
(compression) ec

distribution
across a
section

In this problem we have a combination of pure bending and direct compres-

sion on section A-A. Total section area is
A = (1.75)(0.25) + (1.2)(0.25) = 0.739 in.
The moment arm of the load P is 4 — z = 3.28 and moment on the section is
M = 3.28P. Distance from the neutral axis in bending to the extreme fiber is
Cc, = Z = 0.72 in. on the left (tension) side and c, = 2 — Z = 1.28 in. on the
right (compression) side. The maximum normal stress is simply the sum of
the direct stress and the bending stress. On the left side
—-P M
On(max) = o,(C,) = ae ata = = (usa 2

On the right side:

—-P M
Friainy = Fx(C2) = 72 15.96P
The stress distribution across the section is linear as shown in Figure 6-6b.
 309
Section 6-3: The Maximum Shear Theory

Equating maximum tensile and compressive stress to the working strengths

we obtain allowable load. On the tension side

ia) n(max)
OES,
—_ N

from which
P = 21901b
On the compression side

IOn(min| = a

from which

P = 3410 1b
Tension governs and the allowable load P = 2190 lb @

SECTION 6-3
The Maximum Shear Theory
The maximum shear theory is particularly useful for the design of ductile
steel machine members. It
conservative
is (safe) and easy to apply. It is
based on the assumption that failure occurs when the maximum shear stress
reaches a certain value. That value is the shear stress at the instant of failure
in the tensile test. We will designate yielding as failure. Referring to Figure

ee
stress
axis

Figure 6-7 Mohr circle representation of the

tension test at yielding.
OC. = Sup
Oy = 03 = 0 —
Normal stress axis

6-7, we see that the shear stress at yielding is half the yield point. Thus, the
maximum shear theory is given by

pypeee on (6-11)
 310
Chapter 6: Theories of Failure Used in the Design of Machine Elements

In ductile steels, shear failure sometimes occurs in the standard tension

test. If thé shear strength of a material exceeds S,,,/2, however, we still limit
the maximum shear stress to S,,/(2N) in order to avoid tensile failure when
applying the maximum shear theory. Thus, in a three-dimensional stress
field we use equation (6-6) to obtain

lo, — 92]
Se
Fai > the largest of4 |a, — a3| (6-12)

lo, — 93]
Consequently, for a two-dimensional stress field where o, = 0, we have

lo;
Sys
rie> the largest of; |c,| (6-13)

lo, — 9)
Principal stresses are given by equation (6-3) for the two-dimensional case
and equation (6-5) for the three-dimensional case. For the two-dimensional
case, if the principal stresses in the xy plane ( ite sign,
then maximum shear stress will lie in that plane, and we have the value of
maximum shear stress given by phn
a 2 1/2 ai
rae [+4 i 3, | Va (6-14)
If principal stresses 6, and o, have the same sign, then maximum shear
stress will not lie in the xy plane, and the value Computed in equation (6-14)
will be only of academic interest because we cannot specify the plane in
which a material is to fail. For o, and a, of\the same sign and a, = 0,
maximum shear stress is given by \

Tait On ee (6-15)
For the important special case in which

maximum shear stress lies\in the xy plane. Setting t,,,,= S,,/2N in equation
(6-14), we obtain )
/
|

\ ss
AG aateay (6-16)

Example 6-2: Design of an I-Section Machine Member. Let us design a

member as shown in Figure 6-8 to support a static load P = 1000 lb. A factor
of safety N = 3 will be used, based on the uncertainty of some values and the
danger in case of failure.
 Sid
Section 6-3: The Maximum Shear Theory

Flange thickness h
(both flanges)

Web
thickness t

—24,000

Figure 6-8 (a) Design of an I-section machine member. (b) Shear and moment
diagrams.

Solution: There are many possibilities, and we must make a number of

design decisions. Let us specify some dimensions and calculate web and flange
thickness. If the result is unreasonable, we review the original design decisions.
We will- specify a ductile steel with a yield point S,, = 60,000 psi and let
b = 2in. and d = 1.Sin.
Assuming the reactions to be equivalent to simple supports and setting the
sum of the moments about the left reaction equal to zero, we obtain
R, = 22001b. Then R, = 12001b as shown, and we may draw shear and
moment diagrams (Figure 6-8b).
The greatest contribution to the moment of inertia of the I section is made
by the flanges. Neglecting the moment of inertia of the web and the moment of
inertia of the flanges about their own central axes, we obtain an approximation
of the moment of inertia of the section
I = 2bhd? = 9h
for flange thickness h.
In problems of this type, we assume o, = 0, = T,, = Ty, = 0, noting that
bending occurs in the xz-plane. Maximum tensile stress occurs at the top
surface of the flange where moment is greatest (point A). The distance from
the neutral axis to point A is c = d + (h/2). The tensile stress at A is given by
Mc 4000
EE i rs aFo iw)
a I h
At point A, the shear stress t,. = 0 and we have the case similar to direct
tension. The maximum shear theory reduces to setting o,,4) equal to the
working strength S,,/N. We then obtain the required flange thickness:
h = 0.214 in.
At point B in the web where it joins the flange just to the left of reaction R;,
we have the worst combination of bending and shear. Again using the approxi-
mate value of J, and z = 1.5 — (h/2)
Mz
OxB) = air =~ 1SOF psi
 S12
Chapter 6: Theories of Failure Used in the Design of Machine Elements

Transverse shear stress may be roughly approximated by

eV Re
Tx2(B) ee Anon cad t

for web thickness t.

Using equation (6-14) with the above values for point B, we obtain the
required value of web thickness: t = 0.089 (say t = 0.1 in.). The results are
reasonable, but the reader may wish to compute more exact values of o,, and
T,, using the values of h and t that we have just found and the equations of
Chapter5. @

SECTION 6-4
Strain Energy and the Distortion Energy Theory
Strain energy
In the search for a rational criterion for the prediction offailure in machine
elements, strain energy has been given considerable attention. For the one-
dimensional case (Figure 6-9) the strain energy per unit volume is givén by
ee
Uy dimensional) —_ 70 x€5 (6-17)

O%

Figure 6-9 Strain energy for the one-

Oye, dimensional case.
2

Oe ex

In terms of the principal axes, 1, 2, and 3, equation (6-17) may be extended

to total strain energy per unit volume
U = (6, af O7&o ae 0363) (6-18)

because the principal planes are free of shear stress.

Using the stress-strain relationships
1
= rae — VO, — va3) (6-19)

and so on, we obtain the three-dimensional expression for strain energy in

terms of the principal stresses alone.
 S15
Section 6-4: Strain Energy and the Distortion Energy Theory

1
U = apt + 63 + 03 — 2W(a,0, + 6,03 + 0103)] (6-20)

, however, is considered
not a good criterion for failure
prediction
o, = 0, = 03, is unlikely to cause failure. In order to eliminate the contribu-
tion due to uniform stress, we define an average stress

Ow :
Ol) ain 5) ar We
(6-21)
6-21

Its contribution to strain energy is given by replacing each principal stress

in equation (6-20) by a,,. The resulting energy component is
3(1 — 2v)
Ware. aad IE o2, (6-22)
or, using the definition of a,,,
1 — 2v
U ay) = 6E

Distortion energy
The difference between the strain energy and the contribution due to
uniform stress will be identified as the distortion energy. It is given by
/
ip 1+ y
U,=U-U a(av) 3E a+ + 63 + 0% — 640, — 6203 — 6103) (6-24)

The failure criterion based on distortion energy is obtained by comparing

test

6, =6,=S,, 0, =0,=0

Replacing the yield point by the working strength, we set o; = S,,/N in

equation (6-24) to obtain the distortion energy for simple tension. Equating
the result to the right hand side of equation (6-24), we obtain the distortion
energy theory for the three-dimensional case.
2
a = 6, + 03 + 03 — 010, — 0703 — 40; (6-25)

or, in dimensions,
two where a, = 0;

< Sa\2
A (] = 0, + 63 — 040) (6-26)
 314
Chapter 6: Theories of Failure Used in the Design of Machine Elements

Combined bending and torsion or bending and

transverse shear
As an example, consider the special case where

The principal stresses are given by

Oy.: = 22 t Tmax (6-27)

where
Tmax = [(¢0,) ria le (6-28)

Substituting in equation (6-26) we obtain

atte dee Ve ar (6-29)

based on the distortion energy theory for this special case.
If we write the maximum
shear theory in similar form comparison,
for
we obtain

Syp
mo 2 4 412 3 1/2 (6-30)

The distortion energy theory has been shown to be quite accurate for
predicting failure in
For cases of fatigue loading intorsion and bending, investigators found that
the distortion energy theory
whereas the maximum shear theory predicted the approximate lower limit
of failure values [3].
For the case of combined bending and torsion (equations 6-29 and 6-30),
we see that the maximum shear theory is more conservative than the distor-
tion energy theory. This is generally the case. The difference in results is not
very great, however, and both theories are widely used in the design of
machine elements.

SECTION 6-5
Fatigue Failure and the Soderberg Criterion
In most engineering materials, a stress that fluctuates between two given
values, Omi, and o,,,,, 18 More likely to cause failure than a steady stress
equal to o,,,,. Since fatigue tests are expensive and time consuming, most
available data are based on reversed bending in a rotating member. For an
element on the surface of a rotating shaft subject to bending loads, the stress-
time history is given by Figure 6-10a. The test setup may be similar to the
one illustrated in Chapter 3, using a polished test specimen. The stress
 315
Section 6-5: Fatigue Failure and the Soderberg Criterion

(a)

120,000

100,000 =

80,000 L

§ a ——

40,000

20,000

0 i ais l alee aS
103 104 10 108 107 108
n, stress cycles

(b)
Figure 6-10 Fatigue. (a) Stress-time history for reversed bending. (b) Cycles to
failure at various stresses for a typical high strength alloy steel.

amplitude at which the. member fails after a given number of stress cycles
(n cycles) is designated as S,, the fatigue strength of the test specimen. As
n— oo, S, may approach a lower limit S;, the endurance limit of the test
specimen. Figure 6-10b shows results typical for a high strength alloy steel.
More extensive data are given in Chapter 3. An endurance limit has been
 316
Chapter 6: Theories of Failure Used in the Design of Machine Elements

observed for most steels. The surface imperfections of a typical machine

part affect the endurance limit and, in most cases, a value lower than Sj,
applies.
as given in the modified value of endurance limit is given by

VS) = OnG Gre es.

as given above. For tensile stress range (amplitude) o, and safety factor N,
a design for infinite life is described by

0, < (6-31)
Z|.
if mean stress is zero.
An alte:native fatigue testing device is shown in Figure 6-11. A flat
cantilever beam specimen may be tested in reversed bending, or the vise

Figure 6-11 A flexural fatigue testing machine. [Courtesy Fatigue Dynamics

Inc.]
 alg
Section 6-5: Fatigue Failure and the Soderberg Criterion

holding the specimen may be adjusted so that stress varies about some
nonzero mean value. Crank eccentricity may be changed to provide various
ratios of maximum to minimum stress.

(a)

COOLIPOPPLSSSSSLPLSASSSLASSLASSSSEOSSSSSSS
COLT SPPPSSSSSASSSLSLSSLALSLOLLLSSESS SALSA
$= IF, axis

Any point 0,,, 0, in the shaded region represents N

a safe state of stress.

(b)

(d)
Figure 6-12 The Soderberg criterion. (a) Stress-time
history for fluctuating stress with nonzero mean stress.
(b) The Soderberg plot. (c) Stress-time history for ¢,,;, = 0.
(d) Stress-time history for negative o,,in-
 318
Chapter 6: Theories of Failure Used in the Design of Machine Elements

The Soderberg criterion

If tensile stress at a point is given by a range stress a mean stress
as shown in Figure 6-12a, then both of these stresses contribute
to
. The Soderberg failure line (Figure 6-12b) is an approximate repre-
sentation of this effect. It is a line drawn between the yield point and the
endurance limit on mean stress—range stress coordinates. It was observed
that most failures of test specimens due to combinations of mean and
range loading could be represented as points on or above this line. If we
apply a safety factor N to the yield point and endurance limit, we obtain
the safe stress line, a line parallel to the failure line. A plotted point o,,,¢, on
or below the safe stress line represents safe loading. A combination of mean
and range loading may be checked graphically in this manner. The

When stress varies randomly between limits ¢,,;, and 74x, We simply define
> O max a Omin
/ Om =
Z

Ds)
3G, O max
= Omin

If we considerapoint on a body where istress concentration is present,

we i Then, the state
of stress is given by a point o,,, K ,o,, which should fall on or below the safe
stress line. When a ductile material is used; a coric
stress entration
factor is
to mean stress.
Suppose we load a material up to its safe limit with a combination of
mean and range stress. The state of stress could be represented by the point
0, K,o, on the safe stress line as in Figure 6-13. Noting that triangles AOB

=| State of stress Om, Keo,

x & fo}=

Range
axis
stress

8
Mean stress axis Syp
N
Figure 6-13 Soderberg representation of a material
loaded to its safe limit of mean and range stress.
 319
Section 6-5: Fatigue Failure and the Soderberg Criterion

and CDB are similar, we obtain the relationship

(S/N) —¢,,- S yP
Ko, hy
which may be rearranged in the form
S S
Ne =o, + Kyo. (6-32)

The right hand side of equation (6-32) may be consi

the static
dered equivalent ,
of the state ofstress. Thus, we define equivalent stress
G7)
Oh eS Cie te Ge (6-33)
S.

om Seq Svp Op axis

Figure 6-14 Equivalent stress.

Figure 6-14 is a graphic representation of equivalent stress. If we have

tentatively dimensioned a part, the “‘actual’”’ factor of safety at any point is
given by

N =—2 (6-34)
Gea
If that value of N is less than the overall design factor of safety that we have
selected, we must redesign the part.
In some problems, we are called upon to design a member to withstand a
isinmeaone orapserionsetoateanigey. A straightforward solution may be
obtained if the ratio of range stress to mean stress can be determined from
some known stress-load relationship. In such cases, it is convenient to
 320
Chapter 6: Theories of Failure Used in the Design of Machine Elements

rearrange equation (6-32) into the form

ps oe (6-35)
oO
Ke oe eee
Fay 1S,
If there is concentration
no stress in the part at the point in question, Kis,
of course, replaced by unity: Once mean stress o,, is found, the stress-load
relationship is used to obtain the required part dimension.

Example 6-3: The Soderberg Criterion Applied to Design of a Pressure

Vessel. A thin-wall cylindrical pressure vessel is subject to an internal pres-
sure p, which varies from 100 to 500 psi continuously. Mean radius R = 30 in.
Find the required thickness t of the cylindrical wall, based on a yield point of
70,000 psi, an endurance limit of 30,000 psi, and a safety factor of 2.
Solution: We will define mean pressure, p,, = (Dmax + Pmin)/2 = 300 psi and
range pressure, P, = (Pmax — Pmin)/2 = 200 psi. The Soderberg failure line and
the safe stress line are shown in Figure 6-15. Tangential, axial, and radial stress,

ksi
30

20

0 10 %m 20 30 40 50 60 70 ksi

Figure 6-15 The Soderberg criterion applied to design of a pressure vessel.

are respectively, o, = pR/t; 6, = pR/(2t); and o, = 0 (approximately). The

Soderberg plot will be based on maximum normal stress, o, = o,, with mean
value o,, and range amplitude o,. Since o,, is proportional to p, we know that
the ratio o,/o,, is equal to the ratio of range pressure to mean pressure,
P,/Pm = 2/3. In this problem, there is no stress concentration, that is, we set

We sketch the locus of points representing o,,, ¢, for any thickness t. The
locus is a line through the orizin with a slope of 2/3. Its intersection with the
safe stress line gives the state of stress o,,, 0, for the minimum safe value of
thickness t. The corresponding value of mean stress is
Gs SV NST DFO
as given by the construction on Figure 6-15 or as given by equation (6-35).
Using ¢,, = P»R/t, we obtain the minimum safe thickness for the pressure
vessel: t = 0.657 in.
 S21
Section 6-5: Fatigue Failure and the Soderberg Criterion

When there is a change in section, a hole, or other cause of stress concentra-

tion in a part, the fatigue stress concentration factor K is used in the analysis.
As noted above, for ductile materials it is usual to apply the stress concentration
factor when determining the range stress but not to apply the stress concen-
tration factor to the mean stress. If stress concentration was present in this
problem, we would use a line of slope K,o,/o,, through the origin to intercept
the safe stress line. Otherwise the procedure would be the same. @

~The
[ Soderberg criterion
5 isparticularly applicable
: toductile steel. However,
For a finite life (a given number of cycles) the
corresponding fatigue strength may be substituted for the endurance limit.
> If there is no apparent yield point for a given material, the yield strength based
/ onthe offset method may be substituted. If neither of the above are available,
the ultimate strength may be used in place ofthe yield point. If the ultimate
\ strength is used, it would be prudent to increase the factor of safety.
~— If we have nonzero principal stresses in two or more directions the
maximum shear theory or the distortion energy theory might also be con-
sidered. In the above example, the maximum shear theory gives identical
results. In a one-dimensional stress field, a negative (compressive) mean
stress is generally considered less damaging than a positive mean stress. A
reasonable design procedure would be as follows:
1. Insure that the magnitude of the maximum compressive stress does not
exceed the ultimate compressive strength or the yield point in com-
pression divided by the safety factor

Walaay ~ Suo/N or Soya

2. Insure that the range stress amplitude does not exceed the endurance
limit divided by the safety factor
Gnas SAIN.

The maximum shear theory applied to fatigue loading

In equation (6-6) t,,,, refers to the maximum shear stress for all planes in
an element. Time dependence was not considered. For repeated torsional
loading, we may compute the mean value of t,,,,, and label itt,,and then the

Applying a stress concentration factor for torsion K,,, a plotted point t,,,
K,,t, that falls on or below the safe stress line indicates a satisfactory design.
An analytical solution may be obtained in a manner similar to that used for
tensile loading.
For combined stress, we treat the fatigue effect first, and then use the
rmaximum shear criterion. Using equation (6-32) as a model, we simply
 oZZ
Chapter 6: Theories of Failure Used in the Design of Machine Elements

o account
for fatigue effects. For the special case in which

we have, from equation (6-16)

wh FS wy 2, S 2511/2
wa = = [om+ K, 20 + {ay + K, Sy | (6-36)

When equation (6-36) is applied to shaft design, we may simplify it further.

For a point on the surface ofa rotating shaft, a constant bending moment M
causes tensile stress o,. which is completely reversed with each rotation. The
effect of asteady torque T is a shear stress on the surface that does not vary
in time. Thus, for a rotating shaft transmitting power at asteady rate,we may
write a
| S S 2 1/2
> = (xrz ; a, | (6-37)
Tensile stress amplitude
_K Mc
Ox. = Ss i

and shear stress

i
Txy = ie
J]

on the shaft surface where K, is a shock factor based on the nature of the
applied load and K,, represents the stress concentration factor. There may be
one pair of values for moment and another for torsion. For a solid circular
shaft of diameter D, we have

the
maximum
shear theory
en
————— ~)

Ss. ils 2 1/2 |

> B=
\ yp —
|srs yp *
ear |
*2 \
ia
(6-38)
-z

where M* = K,K,M and Tek T|

In the case of a shaft with or other causes of stress
concentration, the worst combination of moment and torque may not be
obvious. It might be necessary to use equation (6-38) atseveral locations: For
 323
Section 6-5: Fatigue Failure and the Soderberg Criterion

a ductile material subject to steady load the stress concentration factor for
torsion may be taken as unity because it applies to the steady stress com-
ponent. The stress concentration factor for moment on a rotating member
must be applied at any change of diameter or other stress concentration
location. Shock factors and other design details for steady and varying torque
are treated in a later chapter.
Fatigue analysis should be considered whenever fluctuating load is present.
Springs, for example, frequently fail in fatigue. Section 2 of Chapter 14 treats
construction of the Soderberg diagram for spring design.

Distortion energy theory applied to fatigue loading

The effect of a time-varying load may be approximated by defining
equivalent values for each principal stress. Representing the component of
a, due to steady loading by o,,, and the component due to reversed load by
o,, we obtain
S,
GO eq = Oim SF Ko (6-39)

The other principal stresses are also treated in this way. When stress con-
centration exists in the presence of alternating bending and alternating
torsional loads, the appropriate stress concentration factors should be
applied directly to the range components of moment and torque. Using the
equivalent values of the principal stresses in equation (6-25), we obtain an
expression for the distortion energy theory for fatigue loading,
5
NE = Ge + 63 + 0% — 6,0, — 620, — F1F seq (6-40)

e two-dimensional equivalent (where o, = 0) is given by

Se
= = J(a? + 62 — 6162)ecq (6-41)
For the special
case where 6, = 0, = T,, = T,, = 0,we have from equation
(6-29),
Ss
oe = Jj (o- + 8). (6-42)

or

—P = [(o,,, + K ;o,,S gl SF ae Hoe at abe

xyr ye Sle 28)
For the s
the distortion energy hears yields
Seta= m23]
[See
SliiSs 2 a

+ re]
1/2

(6-44)
N mD e
 324
Chapter 6: Theories of Failure Used in the Design of Machine Elements

where the terms are defined in conjunction with equation (6-38). For identical
combined moment and torque loading, design
distortion
by the energy

Modified Goodman criterion for fatigue failure

The modified Goodman criterion resembles the Soderberg except that the
former is line is drawn from the endurance limit
S, on the range stress axis to the ultimate strength S, on the mean stress axis
(Figure 6-16). Another line is drawn from the yield point S,,, on the mean
stress axis, sloping upward and to the left at 45 deg. This pair of line segments

0, axis

Goodman failure line

Goodman safe
stress line

Any point 0,,, Ky 0, in the shaded region

represents a safe state of stress.

Figure 6-16 The modified Goodman criterion.

form the Goodman failure line as shown in the figure. Note that the 45 deg
line segment implies failure when maximum stress exceeds the yield point.
The Goodman safe stress line is constructed similarly, using the factor of
safety N. As with the Soderberg criterion, a point o,,, Ko, failing on or
below the safe stress line constitutes an acceptable state of stress. For
simplicity, the Soderberg criterion will be used in preference to the modified
Goodman criterion for most problems in this text.
 =)
Section 6-6: Cumulative Damage

SECTION 6-6

Cumulative Damage
Extensive fatigue data are not generally available for most materials.
Usually, the best and safest recourse is to base design stresses on the Soderberg
criterion, using the (modified) endurance limit S,. If we have a plot of S,
versus n for a given material (for example, Figure 6-17) we may allow higher

Ss Sp (corrected)
psi bn

120,000 } 60,000

100,000

80,000

Fatigue
strength
60,000

40,000

20,000 |10,000

0 103 104 10 106 107 108

n, stress cycles

Figure 6-17 Cumulative damage.

stresses for a finite number of cycles. In that case, S, replaces S}, in equation
(3-32) and design may be based on one of the equations in Section 5 of this
chapter.

Miner’s rule
is the simplest and probably the best criterion for general use. It is applied as
follows. Suppose the part is subject to n,, reversed stress cycles at a range
stress level S,,;; n,2 reversed stress cycles at a range stress level S,,., and so on.
Now, suppose we select a material described by Figure 6-17. Referring to that
figure, let the number of cycles to failure at stress level S,, be n, cycles; let
 326
Chapter 6: Theories of Failure Used in the Design of Machine Elements

the number ofcycles to failure at stress level S,,. be n, cycles, and so on. Then,
by ’

sil etd ie ae (6-45)

ny N2
constitutes failure.
Ofcourse, an actual machine part is unlikely to correspond exactly to the
fatigue test specimen. Thus, we may apply the appropriate correction factors
described in Chapter 3 to obtain the corrected fatigue strength of the part,
S,ycorrected): Furthermore, we may wish to use a factor
of safety N. Then,
values
n,,n3, and so
on, may be obtained from aplot OfSyicsnecteay/N versus
n. Using these values, inequality (6-45) becomes the criterion for safe design.

Example 6-4. Suppose we select a material corresponding to the S, versus n plot

in Figure 6-17. Furthermore, suppose the actual part size and surface finish
and the required reliability are determined and that a safety factor is selected.
Using equation (3-32) and the safety factor N, we may simply rescale the
ordinate as in Figure 6-17. For purposes of illustration, it is assumed that the
total effect of all these factors results in values Of Sycorrecteay/N which are half
the S, values. Let the load on the part produce reversed stresses of 55,000 psi
for n,, = 2000 cycles; 50,000 psi for n,.» = 5000 cycles and 40,000 psi for
n,3 = 40,000 cycles.

Solution: The limiting number of cycles corresponding, respectively, to the

above stress values aren, = 6000,n, = 16,700, andn,; = 130,000 cycles. Using
inequality (6-45) the result is

Ny N,2
= 23 50.940
ny Ng n3
and the part is safe. @

PROBLEMS
1. (a) Using Problem Figure 26 of Chapter 5, find the principal stresses and the
maximum shear stress at point B on the 1” dia. bar. Show a Mohr circle representa-
tion. (b) Repeat for point C.
- Repeat Problem | with dimension AB = 8 in. and dimension BC = 4 in.
3. Find the allowable load for the cast iron machine element in the Example 6-1 if
ASTM class 20 cast iron is used.

- (a) Find the maximum shear stress at point A of the I section in Problem Figure 1
of Chapter 5. (b) Find the overall factor of safety based ona yield point of 45,000 psi.
. Repeat Problem 4 for the case where only the 1500-lb load S in. from the left support
is applied.
. (a) Find bending and transverse shear stress at points A and B in Problem Figure 6.
(b) Find maximum normal stress and maximum shear stress at both points. (c) For
a yield point of 50,000 psi, find the factor of safety based on the maximum normal
stress theory and the maximum shear theory.
 54/,
Problems

60 in.
200 Ib 200 |b
eae
1.5 in.
B

Web and flange

4 thickness = 0.125 in
Problem Figure 6

. Based on a factor of safety of 2, find minimum web thickness and minimum flange
thickness for the member in Problem Figure 6.
. The I section in Example 6-2 was designed on the basis of approximations of
moment of inertia and shear stress. (a) Using these results, obtain more accurate
values of moment of inertia and shear stress. (b) Find the factor of safety based on
failure at A. (c) Find the factor of safety based on failure at B. Use the maximum
shear theory. Draw a Mohr circle representation of stress at B. (d) Is there a
difference in results if the maximum normal stress theory is used?
. A machine part utilizes a symmetric I section as shown in Problem Figure 9. Find
maximum normal stress and maximum shear stress at points B and C at section
A-A in terms of load F. Predict failure location based on the maximum normal
stress theory and the maximum shear theory.

Problem Figure 9

10. A 13-in. long rod is to be subjected to a constant tensile load of 8000 1b and a
constant torque of 700 lb-in. The rod is to be made of steel with a yield point of
100,000 psi. Design this rod, using the distortion energy theory and a safety factor
OleleD:
11. Design a nylon rod for the same application as the steel rod in Problem 10.
12. A thin wall cylindrical pressure vessel has an internal pressure p = 500 psi and
diameter 24 in. Find the required thickness for a material yield point of 60,000 psi
and a safety factor of 3. (a) Base design on maximum normal stress. (b) Base design
on maximum shear stress.
13. Repeat part (a) of Problem 12 if internal pressure varies from 0 to 500 psi and
the material has a corrected endurance limit of 33,000 psi. Use the Soderberg
criterion.
 328
Chapter 6: Theories of Failure Used in the Design of Machine Elements

“44. Repeat Problem 12 using the distortion energy theory if internal pressure varies
from 0 to 500 psi and the material has a corrected endurance limit of 33,000 psi.
15. A thin cylindrical pressure vessel of diameter D and thickness t is subject to internal
pressure p. (a) Find o;, 0, 03, Tmax in terms of p, D, and t. (b) Draw and label Mohr
circles. (c) For 20,000 psi allowable shear stress, internal pressure 100 psi, and
25-in. diameter, find the required value of thickness.

16. Based ona safety factor of 2, an endurance limit of 20,000 psi, a yield point of 50,000,
find the required diameter of a rotating shaft. Shaft length 10 in. between single
row ball bearings. The shaft has a vertical load of 1000 1b at the center and zero
torque.

17. A small leaf spring 0.25 in. wide by 3 in. long by t in. thick is subject to a load that
varies continuously from 0 to 3 lb (at the center). The leaf spring is supported at the
ends. S,, = 150,000 psi and S, = 70,000 corrected for size, surface, and so on.
(a) Draw the Soderberg safe stress line for a safety factor of 2.5. (b) Plot state of
stress line with slope o,/o,, and find safe limit of ,,. (c) Find required thickness f.
18. Repeat Problem 17 for a load that varies from 3 to 6 lb.
19. Force F varies from 0 to F,,,, continuously 50 times per minute on the part in
Problem Figure 19. Assume failure would occur due to bending stress at the fillet.
S,» = 100,000 and S, = 40,000 psi corrected for size, surface, and so on. (a) Show
the Soderberg safe stress line using a safety factor of 3. (b) Plot state of stress line
with slope K,a,/o,,. Find the safe limit of o,,. (c) Find Fyean- (d) Find F,,,, (greatest
allowable value).
(Note: Apply stress concentration factor to alternating stress only.)

Stress concentration
factor = 1.8

Problem Figure 19

20. Repeat Problem 19 for a load that varies from 4F,,,, upward to F,,,, downward.
21. Apply the distortion energy theory to the pressure vessel described in Example 6-3.
22. The rotating stepped shaft in Problem Figure 22 is mounted between single-row
ball bearings and loaded by two 10-in. pitch diameter pulleys. Use a yield point
S,, = 85,000 psi, an endurance limit $, = 50,000 psi, and a safety factor of N = 2.
Use a shock factor K, = 1.5 applied to torque and moment. Apply a stress con-
centration factor of Kyy = 1.3 at the step. Find the minimum safe values for
diameters D, and D,, using the maximum shear theory.
 329
References

Problem Figure 22

23. Repeat Problem 22 using the distortion energy theory.

24. A solid steel conveyor belt shaft is subject to a maximum moment of 20,000 in.-lb
and negligible torque. Find the required shaft diameter based on an endurance
limit of 40,000 psi (corrected) and a safety factor of 4. Ignore stress concentration.

25. Design a hollow shaft for the same application as Problem 24. Let the inside
diameter be 75% of the outside diameter.
26. A 30-in. diameter cylindrical pressure vessel has a 0.1-in. wall thickness. It is made of
steel with a corrected endurance limit of 36,000 psi and a yield point of 90,000 psi.
A safety factor of 1.5 is to be used. Pressure fluctuates through a 160-psi range.
(a) Is a mean pressure of 267 psi safe? (b) Find the maximum safe mean pressure.
27. A thin wall cylindrical pressure vessel is subject to an internal pressure p, which
varies from 100 to 500 psi continuously. Mean radius is 30 in. Design for a yield
point of 70,000 psi and an endurance limit of 30,000 psi, with a safety factor of 3.
Base results on the maximum normal stress theory.
28. Find the required thickness of the pressure vessel in Problem 27 based on the
distortion energy theory.
29. A solid circular bar is subject to a reversed bending moment of 1200 in.-lb for 2000
cycles; 1000 in.-lb for 100,000 cycles and 900 in.-lb for 10,000 cycles. Find the
required bar diameter based on the corrected properties (Sicorrecteay/N) given in
Figure 6-17.

REFERENCES
[1] R. A. Douglas: Introduction to Solid Mechanics. Wadsworth Publishing Co.,
Belmont, Calif., 1963.
[2] A. P. Boresi: Elasticity in Engineering Mechanics. Prentice-Hall, Inc., Englewood
Cliffs, N.J., 1965.
[3] O. J. Horger (Ed.): ASME Handbook of Metals Engineering Design, 2nd ed.
McGraw-Hill, New York, 1965.
 330
Chapter 6: Theories of Failure Used in the Design of Machine Elements

[4] G. Sines: Elasticity and Strength. Allyn and Bacon, Inc., Boston, 1969.
[5] P. C. Chou and N. J. Pagano: Elasticity. D. Van Nostrand Co., New York, 1967.
[6] J.O. Almen and P. H. Black: Residual Stresses and Fatigue in Metals. McGraw-Hill
Book Co., New York, 1963.
[7] P. G. Forrest, Fatigue of Metals. Pergamon Press, Elmsford, N.Y., 1962.
Shafts, Keys and Couplings
SYMBOLS

A = shear area of shaft key, in. M,, = mean applied bending moment,
D = diameter of solid shaft, in. Ib-in.
D; = internal diameter of a hollow M, = range of applied bending
shaft, in. moment, Ib-in.
= inside flange diameter, in. = factor of safety
Sp |= outside diameter of a hollow ll normal load on a key, lb
shaft, in. = number of spline teeth
D,, = outside flange diameter, in. = number of bolts
E = elastic modulus in tension, psi = number of splines
F = load on a shaft key, lb &3S= critical shaft speed, rpm
SS
2
F = load on a gear, worm, or pulley, = diametral pitch, in.
Ib r = shaft radius, in.
F = reaction at bearing supports, mn = mean spline radius, in.
Ib rpm = revolutions per minute
F, = bolt tension, lb S, = compressive strength, psi
h = depth of spline tooth, in. S. = corrected endurance limit, psi
hp = horsepower S.,; = corrected endurance limit for
I = area moment of inertia, in.* shear, psi
J = area polar moment of inertia, Ssyp =
yield strength in shear, psi
in.* Sy» =
yield strength in tension, psi
A II fatigue stress concentration T =
torque, Ib-in.
y
factor due to bending, tension, T,, =
mean applied torque, lb-in.
or compression T, =
range of applied torque, lb-in.
K, = fatigue stress concentration V =
shear load, Ib
factor due to torsion or shear w =
load per unit length, lb per in.
K,, = shock factor for bending, or ft
tension, or compression W = width of shaft key, in.
= shock factor for torsion W,, = weight of the mth rotating mass,
L = length of shaft key, in. Ib
L = length of spline tooth in contact, x = distance measured in the x
in. direction, in.
= effective spline length, in. y = deflection, in.
M = bending moment, lb-in. Ym = Static shaft deflection due to the
mth mass, in.

331
 332

Chapter 7: Shafts, Keys and Couplings

6 = resultant deflection, in. t = shearing stress, psi

0 = slope, deg Tmax = Maximum shearing stress, psi
@ = torsional deflection, deg Tms = mean shearing or torsional
u = coefficient of friction stress, psi
Om = mean bending stress, psi T,; = range of stress due to torsion,
o, = range of stress due to bending, psi
psi w = angular velocity, rad/sec
o,, = bending stress, psi

This chapter deals with the design of shafts, the use of keys to connect
other machine elements (for example, gears, pulleys, and so on) to shafts, and
the use of various couplings in connecting shafts.
Shafts are used in a variety of ways in all kinds of mechanical equipment.
Typical uses are power shafts, cam shafts, line shafts, and so on. As a result
of industrial application, particular definitions are associated with shafts
used for a definite purpose. These definitions are as follows:
Shaft. A rotating member used for the purpose oftransmitting power.
Axle. A stationary member used as a support for rotating elements such
as wheels, idler gears, and so on.
Spindle. A short shaft or axle (for example, head-stock spindle of a lathe).
Stub Shaft. (Also called head shaft.) A shaft that is integral with an
engine, motor, or prime mover, and is of such size, shape, and projection
as to permit its easy connection to other shafts.
Line Shaft. (Also called power transmission shaft.) A shaft that is directly
connected to a prime mover and is essentially used to transmit power to a
machine(s).
Jackshaft. (Also called countershaft.) A short shaft that connects a prime
mover with a line shaft or a machine.
Flexible Shaft. Permits the transmission of motion between two points
(for example, motor and machine) where the rotational axes are at an angle
with each other. The amount of power transmitted is of relatively low level.
Depending on the loading, shafts are subjected to constant bending and/or
torsional stress or a combination ofthese stresses caused by fluctuating loads.
To design a shaft subjected to a fixed loading, the reader should also refer to
the analytical methods discussed in Chapter 5, namely, Section 5-7 (for
torsion of round sections), Section 5-5 (for simply supported beams with
different loading), and so on.
In this chapter, we will consider a shaft as a machine element subjected to
fluctuating loads. The associated considerations of stress concentration and
endurance, therefore, will play a significant role in its design. A shaft designed
from.the viewpoint of these considerations will possess adequate strength.
However, of equal importance (sometimes of greater importance) is the
consideration of shaft rigidity or stiffness. A shaft having too large a lateral
deflection can cause excessive bearing wear or failure. A large lateral deflection
is also responsible for lowering the critical speed, which may cause the shaft
to vibrate violently if its revolutions per minute are at or near this speed.
 Bie
Section 7-1: Shaft Materials

Keyways are a source of stress raisers as are other geometrical discon-

tinuities, and care must be exercised in their selection and location in order
to minimize the resulting stress concentration.
Two shafts can be connected by shaft couplings of which there are many
varieties. The angle between two shafts, transmitted power, torsional and
lateral vibrations, shock loads, and so on, are all parameters that must be
considered in selecting a coupling to connect the shafts.

SECTION 7-1
Shaft Materials
Generally, shafting of 3-35 in. diameter is made of cold-drawn, carbon
steel round bars. Where toughness, shock resistance, and greater strength are
needed, alloy steel bars (for example, 1347, 3140, 4150, 4340, 5145, 8650) are
commercially available. These bars can be heat treated to produce the desired
properties. Where surface wear resistance is the dominant factor, a carburiz-
ing grade steel (for example, 1020, 1117, 2315, 4320, 4820, 8620, and so on)
can be used. However, to avoid increased cost, the designer should try to use
a plain carbon steel if it is at all possible. Other case hardening methods such
as nitriding, cyaniding, flame and induction hardening can also be used as a
means of producing a wear resistant surface.
Cold-drawn steel bars exhibit higher physical properties than hot-rolled
bars of similar steel. They have greater yield values and greater ultimate
and endurance strengths. However, the higher endurance values are some-
what offset by the residual surface tensile stresses caused by the cold drawing.
Since fatigue (or endurance) failures of shafts are tensile failures (torsional
failure is also possible), these residual surface stresses can contribute to
higher fatigue stresses. Also, cutting keyways, slots, and so on, relieves the
surface stress of the machined areas, causing the shaft to warp. The required
straightening adds to the cost, but can be beneficial if hammer peening is
used to straighten the shaft. As the reader may recall, peening and similar
processes produce a compressive surface stress that counteracts the effect of
fatigue stress. Where warping is a serious consideration, and cannot be
easily rectified by straightening, an oil-hardening alloy steel (for example
3140, 8640, and others) can be considered. Where closer tolerances on the
diameter are needed, warehouses can supply turned, ground, and polished
rounds or cold-drawn, ground, and polished rounds.
Shafts larger than 34 in. diameter are machined from hot-rolled carbon
steel. The machining must be deep enough to remove all of the decarburized
scale caused by hot rolling. Large diameter shafts, such as railroad axles,
press cranks, and so on (for example, a steel of 0.45% carbon) are usually
forged and then machined to the required size. Shafts of steel or nodular iron
are also made by casting (for example, automotive engine crankshafts).
 334
Chapter 7: Shafts, Keys and Couplings

SECTION 7-2

Bending Moments and Torque Acting on a Shaft

Gears, pulleys, flywheels, friction wheels, cams, and ratchets are mounted
on shafting in various combinations and locations. It is obvious from the
science of mechanics that, to determine the diameter ofthe shaft, it is necessary
to know the bending moment and torque distribution along the full length of
the shaft. With this information the designer can specify the required
diameters for different parts of the shaft.
The diameter ofthe shaft, or the diameter of each section, will depend upon
the combined stresses due to the bending moment and torsion. Depending
upon the particular problem, the exact location along a shaft where the
maximum stress will occur is sometimes obvious. Lacking such obvious
information, it is helpful to draw the shear and moment diagrams to locate
the points along the shaft where the moments are maximum. The following
example is an illustration of the above ideas.

Example 7-1. A 24-in. diameter pulley weighing 100 lb receives 30 hp at 360 rpm from
a shaft located at 45 deg below the pulley. An 18-in. pitch diameter gear C
weighing 50 1b delivers 40°% of the power horizontally to the right. Finally, a
12-in. pitch diameter gear E weighing 25 lb delivers the remaining power to
another gear downward to the left at an angle of 30 deg below the horizontal.
Both gears have 20 deg pressure angles. The system is to be used in a chemical
plant and corrosion can have a long term affect.
Determine the torques transmitted by the pulley and gears and the resultant
moment distribution.

Solution: Referring to Figure 7-1, the transmitted torques are computed as

follows:
Torque acting on the shaft between B and C
_ 63,000 hp _ 63,000(30) _
ii 5250 Ib-in.
y rpm * 360 os
The horsepower delivered by gear C is 40°, of the 30 hp. Thus, the torque
delivered by gear C is

Te 63,000(12)
e 360 2100 Ib-in.

The remaining 18 hp is delivered by gear E. Thus, the torque between gears C

and E is
63,000(18)
TiE = ————_
360 = 3150 Ib-in
-in.

The torque on the pulley is

D
(= De = 5250 lb-in.
or
 ae ae 335
Section 7-2: Bending Moments and Torque Acting on a Shaft

Das- cp

INA

ss
S,
We
Es
SSS3

Figure 7-1 Gear C is in mesh with gear G and gear E is in mesh with gear H.
The right side view shows the reactive forces of G on C and H on E, respectively.

For a properly mounted flat belt not subject to large overloads and having an
appropriate initial tension, we may assume a belt tension ratio!
JEG = 3
From the above two equations, we determine that
1056.25)
iis, = DNS) Ilo,
Thus, the total force causing bending at the pulley B is
JO = IP Fb IB = OOD ah AUS IS) = 87/5 Sl)
The tangential reactive forces acting on the gears C and E are computed as
follows (rc, rg are the pitch radii of gears C and E, respectively)
Te 2100 lis - SSO)
Fer = — = — = 233.331b: 9Fee = = = —— = 525.001b
rc 9 lp 6
The normal force tending to separate the gear teeth in contact is equal to the
product of the turning force F times the tangent of the tooth pressure angle
(see Chapter 10). Thus, the normal forces tending to separate gears C and G
and gears E and H are

Foy = Fez tan 20° = (233.33)(0.364) = 84.93 lb

Fey = (525)(0.364) = 191.101b
The forces acting on the system in Figure 7-1 are shown for clarity on Figure 7-2.
Using Figure 7-2, we take moments about the bearing A in the horizontal
plane.
(F, sin 45°)12 + Fe7(30) + Dy(40) + Fey sin 30°(50) — Fey sin 60°(50) = 0

' Actually this ratio depends upon the active arc ofbelt contact, the coefficient offriction between
the belt and pulley, the density of the belt material, and the surface speed of the pulley (see
Chapter 12).
 336
Chapter 7: Shafts, Keys and Couplings

Thus

Dy = 88.29 |b
Since D,, is positive, our choice ofits direction as shown in Figure 7-2 is correct.
Now we sum the forces in the horizontal plane.
Ay = F, sin 45° = Fer Tee Dy = Fen sin 30° == F 7 sin 60° — 0

Thus
Ay = 581.151b
Proceeding in the same manner, we take moments about bearing A in the
vertical plane.
(Fz cos 45°)12 + 100(12) + Fen(30) + 50(30) + Dy(40) — (Fey cos 30°)50

—(Frr cos 60°)50 + 25(50) = 0

Thus
Dy = 186.95 Ib
Since Dy is positive, our choice of its direction as shown in Figure 7-2 is
correct.
Summing the forces acting in the vertical plane.
Ay — Fg,cos 45° — 100 — Fey — 50 — Dy + Fey cos 30° + Fez cos 60°
—25=0
Thus,
Ay = 637.521b @

We can now proceed to draw (or sketch) the shear and bending moment
diagrams for the horizontal plane and for the vertical plane (Figure 7-3).

Figure 7-2 Schematic of force system for Example 7-1. The directions of the
forces at bearings A and D are arbitrary.
 557
Section 7-2: Bending Moments and Torque Acting on a Shaft

Recall from your mechanics course, if you will, that a maximum bending
moment occurs at that point along the shaft length where the shear diagram
crosses the zero axis.* This statement is verified by inspection of Figure 7-3
from which we can indeed see that the maximum bending moment does
occur at the point of zero shear in both the horizontal and vertical planes.
For the case shown in Figure 7-3 (that is, for Example 7-1), the maximum
moment in both planes is found at the pulley support point B. However, the
reader should note that this need not always be the case. Thus, the maximum
bending moment in both the horizontal and vertical planes need not (and in
most cases will not) be found at the same point along the length of the shaft.
Another possibility is the case where the shear diagram crosses the zero
shear line at more than one point. This would then indicate that there are one
or more ‘“‘maximum’”’ bending moments, namely, one at each point of zero
shear. In such cases, the designer may have to establish the resultant bend-
ing moment for the horizontal and vertical planes at each maximum. If the

Horizontal Plane Vertical Plane

(Fer) y— (Fen) y (Femly

+ Ferly
= 454.65-—95.55 = 165.49+262.50—25
Ay = 581.15 = 359.10 Ay = 637.52 = 402.99

12in. B 18 in. C10 in.

Fey = 618.63 Foz = 233.33 Dy = 88.29 Fey = 718.63 Foy = 134.93 Dy = 186.95

637.52
581.15

6973.75
in.-lb. in.-Ib.
7650.23 6190.32
in.
lb
-
6299.13
in.-lb.

Figure 7-3 Load, shear, and moment diagrams for Example 7-1 for the
horizontal and vertical planes.

2 This conclusion could be invalidated by a couple or moment acting in the longitudinal plane
of the shaft. In such cases, the maximum may be at a point other than that where the shear
diagram is zero.
 338
Chapter 7: Shafts, Keys and Couplings

shaft is one of constant diameter, it will then be designed for the ““maximum,”
maximum resultant bending moment and the value of the torque at that
shaft location.

If a stepped shaft is being considered (for example, Example 7- ,), then the

In the preceding discussion, the shaft diameter(s) and corresponding

stresses are determined by the combined stress equations.
For Example 7-1, the maximum bending moment in each plane occurs at
the pulley (that is, B). The bending moment at B is then the vector sum ofthe
moment in each plane. Thus
Mz = [(6973.75)* + (7650.23)7]*/* = 10,351.77 Ib-in.
and we also note again that the maximum torque is also at B,
Ty = 5250 Ib-in.

SECTION 7-3
Shaft Design Subjected to Fluctuating Loads Based
on the Maximum Shear Theory of Failure
In Chapter 6 we established that the maximum shearing stress of a
solid round bar subjected to steady combined static bending and torsion
(equation 6-14, if o, = 0) is

2
Tmax = x (a

where

32M 16T
Os, = D> d = ie (7-1)

For a hollow shaft

32M 16T
eievel (22)
°x * nD3[1 — (D,/D,)*] ~ nD3[1 — (D/D,)1
Thus, using the maximum shear failure theory and replacing o, and t, the
values from equations (7-1), we get

m OSS ek 16 oer
foal e Se DT (DED ee Te)
 339
Section 7-3: Shaft Design Subjected to Fluctuating Loads

a |= maximum shearing stress, psi (from Mohr’s circle)

S,,, = tensile yield of material, psi
II factor of safety
= outer diameter of shaft, in.
|= inner diameter of shaft, in.
<SS2 applied bending moment, Ib-in.
T = applied torque, Ib-in.
Equation (7-3) can also be used for a solid round shaft by making D; = 0.
An equation similar to equation (7-3) for fluctuating loads can be obtained
by substituting the Soderberg equation (6-33) for o, and its equivalent for t.
Thus we obtain
055,, a S 2

Tmax — Ne a 4 o Tins + 2tz,] (7-4)

As above, we can replace o,,, 0,, Tms, and t,, by equations (7-2) (using the
appropriate subscripts on a, t, M and T) to obtain equation (7-4) in terms of
bending moment and torque.

0.5S 16 S : S :
a
lena N w= nD>[(1 = ory M,,
mo"+ S
22M,]| r + |T,,
m + —“T,
S |

(7-5)
AS

where
Om = Mean stress due to bending + axial stress, psi
o, = range of stress due to bending + axial stress, psi
Tms = Mean stress due to torsion, psi
T,; = range of stress due to torsion, psi
M,, = mean applied bending moment, Ib-in.
M, = range of applied bending moment lb-iny =
T,, = mean applied torque, Ib-in.
T, = range of applied torque, Ib-in.
K, = fatigue stress concentration factor due to bending, tension or
compression (based on notch sensitivity)
K ,, = fatigue stress concentration factor due to torsion (based on notch
sensitivity)
= yield point strength of the material in shear, psi
= yield point strength of the material in tension, psi
1
Sa CrOsCr Cw Sn (see equation 3-32a)
oh
1
es
= CrCsCrCw ——Sns(see equation 3-32b)
fs
 340
Chapter 7: Shafts, Keys and Couplings

SECTION 7-4

Design of Shafts Subjected to Fluctuating Loads

Based on the Distortion Energy Theory of Failure
The stress equation based on the distortion energy theory of failure was
given in Chapter 6 (equation 6-26) as
Ss 2
Fa = 67 —6,0,+ 0%

As was demonstrated in Section 6-1, ifo, and a, are replaced by the relations
for principal stress, equation (6-3) where the normal stress is due to uniaxia!
loading only, we obtain equation (6-29)
S
Gee
As in the case of the maximum shear theory of failure, the Soderberg
equations (6-33) and its equivalent can be substituted for o, and t. The

r
resulting equation, with a factor of safety included is

S z Sey >
Om at ~, Bs 3 a 2, (7-6)
N Se S es
Equation (7-6) may also be expressed in terms of shaft diameter, applied
bending moment, and applied torque. Repeating what we did in the case of
the maximum shear theory of failure, we obtain
iS) 3 S oS S z
ve M, + 22M Lp SER TT
N nD3{I = mma SU é€ | 7 mts es
| oe
SECTION 7-5

Comparison of Maximum Shear and Distortion

Energy Theories of Failures as Applied to
Shaft Design
Because the maximum distortion failure theory is less conservative than
the maximum shear failure theory, the diameter calculated by using equation
(7-7) will be smaller than that calculated by equation (7-5).
However, on closer inspection, these two equations are remarkably alike.
Notice that if we divide both sides of equation (7-7) by 2, it is transformed
exactly into equation (7-5) except for the coefficient under the radical which
multiplies the applied torque components. For the maximum shear theory,
this coefficient is 1, and for the distortion energy theory it is 3. Thus, it is
apparent that both theories will lead to shaft diameters not too far apart from
each other. This would particularly be the result in those situations where
 34]
Section 7-6: Shaft Design Equations for Fluctuating and Shock Loads

the applied torque is much smaller in magnitude than the applied bending
moment.

SECTION 7-6

Shaft Design Equations for Fluctuating and

Shock Loads
Prior to 1954, the accepted basis for shaft design was the ASME code for
the “Design of transmission shafting, ’ B17c—1927, approved by The American
Standards Association. It was based on the maximum shear theory offailure,
and employed combined shock and fatigue factors as multiplying coefficients
for the applied bending moment and applied torque. This code was with-
drawn in 1954 and has not been replaced since the time of its withdrawal. It
is, therefore, not necessary to discuss this code.?
However, the previous discussion concerning the maximum shear failure
theory and the distortion energy failure theory (both of which incorporated
the Soderberg equation) resulted in equations (7-4) and (7-6) or (7-5) and
(7-7). These equations are rational but include only stresses due to fluctuating
loads and not shock loads. Although many shafts are not subjected to shock
loads, some applications do arise where such loading does take place (for
ane punch esse shafts for ball mills, and so on).
take nditions, it has been suggested [1] that multi-
plying meen: (that is, correction oo) be used in the shaft design
equation based on the distortion energy theory. However, there does not
appear to be a valid argument against using these same coefficients for the
shaft design equation based on the maximum shear theory of failure. Thus,
equations (7-4) and (7-5) (using the maximum shear theory) become,
respectively
MST
: O58 ue
Ps ec
SEP yo Kscoat Sa \? K Sapse ie 7-8
x Tmax N i4 Om et eae
Se Oy t aa Se Tsp ( )

x O.S375 16
(eae
max A N =
nD>[1 — (D,/D,)*] (7-9)
S 2 Se
x [Kol Maae eM, yr K,, tes ap er

3 The ASME code for the “Design of transmission shafting” can be found in any engineering
handbook (e.g., Mark’s, Kent’s, etc.).
 342
Chapter 7: Shafts, Keys and Couplings

Similarly, equations (using the energy distortion theory) (7-6) and (7-7)
become

Die ea os ge
N Sj KING
sb m ree
S r o oe ee
st as sm + Sop,
Ky sr (7-10)
e

5 32 vo 3 \2
NYP aD3[I ere Klee
A mts Pe) | beeeK .|ee
mE ee
ee. |
é es

(7-11)
The values for K,, and K,, are given in the following table.

Coefficient Value

1.0 for gradually applied or

steady loads

Keys Ke, 1.0-1.5 for minor shocks

1.5-2.0 for heavy shocks

SOURCE: H. A. Berchardt: A comprehensive method for

designing shafts to insure adequate fatigue life. Mach. Des.,
(April 25, 1963).

SECTION 7-7
Shaft Diameter for Example 7-1
If the shaft is to be machined from normalized C1040 steel, we obtain from
Table A-2 of Appendix A the following material properties
S, = 85,000 psi_—-S,,, = 50,000 psi

we use —— 7 Thus

S), = 0.58, = 0.5(85,000) = 42,500 psi

From Figure B-1 of Appendix B we determine that, for the profiled keyways
at B and C, K ,(bending) = 1.6 and K,,(torsion) = 1.3. For the sled-runner
keyway at E, we find: K bending) = 1.3 and K,,(torsion) = 1.3.
The shaft is to operate in a noncorrosive atmosphere, and is to have a
survival rate of 95%. Thus, from equation (3-23) and Table 3-2, we have
Cr = 1 — 0.08(DMF) = 1 — 0.08(1.64) = 0.869
From Figure B-3 of Appendix B we obtain for a machined surface
Cr = 0.77
 343
Section 7-7: Shaft Diameter for Example 7-1

Finally, since it is likely that the shaft diameter will be larger than $ in. and
perhaps smaller than 2 in., we will apply equation (3-24) and so
Cs = 0.85
We now can determine the corrected endurance limit by means of equation
(3-32a) (where Cy = 1 in this case)

Sa = OsesOS = 5g(0869)(0.77)(0.85)(42,500) = 15,108 psi

f
If the shaft is to be solid (that is, D; = 0), the load smooth, and the factor
of safety = 3.75, (see Section 1-3) we now determine the diameters for the
stepped shaft in Figure 7-1. The diameter between B and C is to be 25%
larger than that between A and B. The pulley and gears are assembled on the
shaft with a light press fit.
Consider the type of shaft load, we can conclude: (1) that since the load is
smooth, K,, = K,, = 1.0 (see the table Values for Shock Conditions);
(2) because the bending load is repetitive, M,, = 0*; and (3) since there are
no flywheels, clutches, and so on, the torque will be steady, thus T, = 0.
Applying this information to equation (7-11), we obtain
Shee Soon |Semen cee ee
~ WG se 90 Te
N a (: e e
dna ele)
In Section 7-2, both the maximum bending moment and maximum torque
were found to be at location B where the shaft diameter D,, joins the larger
shaft diameter Dz-. Possible failure can occur at the fillet joining the
diameters or at the keyway. The more critical of the two possibilities will be
the one with the larger product K ,-M.
The stress concentration at the pulley is due to the profiled keyway and
the light press fit. According to Peterson [2], for a press fit with an average
pressure, K, = 1.9 for shafts subjected to fluctuating bending loads. However,
due to inconclusive evidence and fretting corrosion, we shall assume
K, = 2.5. Thus, for the keyway and press fit combination we obtain®
K,M, = (2.5)(1.6)(10,351.77) = 41,407.08 Ib-in. (7-13)

where

M = max
Sr( rain) Ss
Vie max evi min =
2M r = M

f 2 2 2 :
Now, to obtain K ,; for the fillet, it is necessary to know both r/d and D/d (see
Figure B-4a in Appendix B) and q from Figure B-2 in Appendix B. Because

= Mee =F (=M)minl/2 — 0.
4M,
 344
Chapter 7: Shafts, Keys and Couplings

we do not have all the data to enable us to use these figures, we will have to
‘‘suess”’ at a K, and g. It is immediately apparent that unless we “guess” at
a K, > 4, which is the product of 2.5 and 1.6, the critical location for deter-
mining the shaft diameter d is going to be at the pulley where the keyway and
press fit are the overriding factors. Since a value of K , of such a high magni-
tude for a fillet is quite unlikely, we proceed with the design by using the value
of the bending moment at point B. Substituting M, = 10,351.77 in equation
(7-12)
Dp = 2.967 in.
Use Dap 00. Litus
Dec = 1.25(3.00) = 3.750 in.
We may now check the stress condition atthe fillet by assuming r = 7 in.
Thus, for r/d = 0.0625/3 = 0.0208 and D/d = 1.25, we obtain from Figure
B-4a in Appendix B K; ~ 2.28. From Figure B-2 in Appendix B we determine
that gq = 0.78 (that is, for r = 1/16 and annealed steel). And so, using
equation (3-29)
Kp=1+ qk,— )= 1 + 0.780.288 — 1) = 217
Returning now to equation (7-12) we can determine the factor of safety, N,
for the shaft diameter D,-. Note that in applying equation (7-12) the value
of M, is assumed to be the same as at section B. This decision is acceptable,
since the relatively wide hub of the pulley added rigidity to this section.
Furthermore, as can be’established from Figure 7-3, the bending moment
just to the right of section B is smaller than that at B and the calculation,
therefore, is conservative. Thus
Nos = yp
BC 32 J Syong Z i 3p

nDic Sart qa"

50,000
SS Sa
SR ee eS SS TD

7(3.750)? yf 115,108 ~ L227]

ee UUM LOSS ;
+ Ge290)
31595 2

For the purposes of completing the problem and determining the shaft
deflection and critical speed, we will assume that a series 213 ball bearing
(width 0.9055 in.) will be used to support the shaft at each end. To ensure a
satisfactory fit, those locations of the shaft on which the bearing will be
mounted are nominally 2.559-in. diameter. The fillet radius between the
3.00-in. diameter portion of the shaft and that portion which supports the
bearing is r = 0.06 in.

° Since D,,, > 2 in., one should repeat the calculation with a smaller C, (say C, = 0.75-0.80)
to obtain a corrected diameter D,,. However, to conserve space, this has not been done in the
example given above.
 345
Section 7-8: Shaft Deflection for Example 7-1

From Figure 7-3 we obtain

Mp = x/(3591.0)? + (4029.9)? = 5398 Ib-in.

Now we have r/d = 0.06/3 = 0.02, D/d = 3.000/2.559 = 1.17.
From Figure B-40 in Appendix B we obtain K, = 2.40, and from Figure
B-2 in Appendix B we again obtain q = 0.78. Thus

K, = 1+ q(K, — 1) = 1 + 0.78(2.40 — 1) = 2.09

However, the bearing fit on a revolving shaft is likely to have a press fit with
the shaft for which we can assume K, © 2.5. Using equation (7-12), we find
N for the shaft section between D aed E, where

1 .
5, = (0.869)(0.77)(0.85)5— (42,500) = 9700 psi

Nae
50,000 = 2.94
32 50,000 2
3
+ ey
7G! 50)
0)
oat am
Thus, the shaft diameter for the ball bearings is more than satisfactory. In
fact, if load, speed, and life requirements permit, it would be “in order” to
consider the use of a smaller size ball bearing.

SECTION 7-8
Shaft Deflection for Example 7-1
Having determined the shaft diameter required to withstand the severest
load conditions and the diameters for other parts of the shaft, the next step
is to calculate the shaft deflection.

or
enclosure) Also, thedeflection atgearlocations will
, increase thepressure angle, and
reduce the length of tooth contact. Beceeeeeartt deflection, then, impedes
proper functioning of gears.
In the course of finding the shaft deflection, the slope of its elastic curve
must be established. Knowing the slope will enable the designer to judge the
amount of skewing between gear teeth (an additional deviation from ideal
gearing theory). Furthermore, it will help him to determine the minimum
bearing clearance for sleeve bearings as well as whether self-aligning bearings
are required.
 346
Chapter 7: Shafts, Keys and Couplings

There are many methods’ for obtaining the deflection of a shaft or beam.
The simplest of these methods is by direct integration, particularly where a
shaft has a constant diameter throughout its entire length. However, most
shafts are “stepped” shafts. The use of direct integration, then, becomes
cumbersome since boundary conditions at each change of diameter have to
be satisfied.
The more popular methods for finding deflections are the area-moment
method, graphical integration method,® and the numerical integration
method.
We shall determine the deflection of the shaft in Example 7-1 by numerical
integration. This particular method is not only compatible with modern
nigh-speed desk calculators, but lends itself readily to computer solutions.
The bases for numerical integration are the differential equations below.
The sign convention for y and the distance along the shaft-axis are shown on
Figure 7-4.

Loadoad equati
equation GeeEl
a

Sh F iia ® any
ear equatioion Ra
aoe Ey
(7-14)
; 2 td EM
Moment equation qe

: d
Slope equation mele 0
dx

It is apparent that by considering finite distances along a shaft length (that

is, the x direction), the above equations can be integrated by numerical
summation. For example, y = L@Ax.
The format used here to find the deflection of the shaft in Example 7-1 is
based on one found in reference [3]. In accordance with this reference,
Tables 7-1 and 7-2 on pages 349-350, are determined by the following step by
step procedure:

7 One method that is quite simple and can be used for stepped shafts with various loads employs
the singularity function (also called Macauley brackets) to establish the shear, moment and
deflection diagrams. The technique, once understood, becomes algebraic and is also easily
adaptable for computer programming. An excellent explanation of this method can be found in
An Introduction to the Mechanics of Solids, by S. H. Crandall and N. C. Dahl, McGraw-Hill
Book Co., New York,

® Graphical methods of integration can be found in any good book on calculus. Specifically,
one may refer to Design of Machine Elements, 4th ed., by Virgil M. Faires, The Macmillan Co.,
New York, 1965, where a problem similar to the one used in the example in our present chapter
can be found. Also, it is suggested that the reader refer to Mechanical Vibrations, 2nd ed., by
Austin H. Church, John Wiley & Sons, Inc., New York, 1964.
 347
Section 7-8: Shaft Deflection for Example 7-1

2.559 dia.

0.45

ive}
= ren)
a8 aN
foe)
ice) io) 359.10
Lo. 233.33 ti
(oo)
xc
R=

Figure 7-4 The upper sketch shows the shaft diameters, distance between loads
and diameter changes, station numbers, and load magnitudes for the vertical plane.
The lower sketch shows the load magnitudes for the horizontal plane. All the other
data is the same as for the upper sketch and so is not repeated on the lower sketch.

Step 1. Divide the shaft into length intervals by marking a station number
at each force and change of section (see Figure 7-4). Where there are relatively
long sections of constant diameter, greater accuracy may be obtained by
also dividing these sections into smaller length intervals.”
Step 2. In column 1 of the computation sheet, enter the station numbers
on alternate lines.
Step 3. In column 2 of the computation sheet, enter the forces on the same
line as their station number. Upward forces are taken to be positive and
downward forces are taken to be negative. Note that those stations not
having any forces acting at that location are blank spaces in column 2.
Step 4. Calculate the vertical shear at each station by summing the forces
in column 2. Enter these shear values in column 3 one station below that
° This recommendation was not followed in the solution to Example 7-1 (i.e., dividing long lengths
of uniform cross section) in order to conserve space. However, avoidance of this recommendation
in no way alters the procedure.
 348
Chapter 7: Shafts, Keys and Couplings

for which it is calculated. As a check, the very last value for the shear should
be equal in magnitude to the last force in column 2, but opposite in sign.
Step 5. Enter on the same line as the station number the distance from
that station to the preceding one.
Step 6. Calculate the bending moment at each station by multiplying the
shear value in column 3 by the distance to the preceding station in column 4.
This product when added to the previous entry in column 5 is the bending
moment at that station. Note that unless station 1 is acted upon by an outside
couple or bending moment, the value at this station is zero.
Step 7. In column 6, on a line in between stations, enter the value of the
area moment of inertia (J) for that section of shaft length between the two
stations being considered.
Step 8. In column 7, enter on the same line as the area moment of inertia,
the product of the elastic modulus (E) and the area moment of inertia (J).
Step 9. In column 8, enter the results of dividing the values in column 5
by the values in column 7 (that is, M/EI). The procedure that is followed is
to divide those values in column 5 that precede by one line and follow by
one line the value of EJ in column 7. The first result is entered in column 8
on the line above the EI value, and the second result is placed on the same
line as the EI value.
Step 10. In column 9, enter the average value of M/EI (that is, column 8)
on the line in between stations. The average is obtained by summing the
M/EI value at a station with the M/EI value that follows the station and
dividing the total by 2.
Step 11. Calculate the slope at each station by entering the “running
total’ of the product of the average M/EI (that is, column 9) and the Ax
value (that is, column 4) found on the very next lower line from the average
M/EI value. The results are entered in column 10 on the same line as the
station numbers.
Step 12. In column 11, enter the arithmetic average of the slope from
station-to-station (that is, from column 10). These average slope values are
recorded on the lines in between the stations.
Step 13. Next, the deflection increment is obtained by multiplying the
value of the average slope (column 11) by the Ax value (column 4), entered
on the next lower line from the average slope value. The product is entered
in column 12 on the same line as the average slope.
Step 14. In column 12, total all the deflection increments between bearing
reactions. Change the sign of this sum and divide it by the distance between
the bearings. The result is the integration constant.
Step 15. Next, the integration constant is multiplied by the distance
between stations, Ax (column 4). This product is the integration constant
for each interval and is entered in column 13 on the same line as the deflection
increments.
Step 16. Add the deflection increment (column 12) and the integration
constant (column 13) and enter the sum as part of a “running total”? in
 349
Or s8urivaq u99Mj}0q DdULISIG
‘ul/"Ul E/0'7S9— = - = 40] x 2 ‘ejsuos uoNeIsaU]
STL T8097 — STL 18097 —
STL'18097 = (XV)o01 x OX = SHoddns BulIvIq UaaMjaqg SJUSUIAIOUI UOTD9HEp Jo WINS
O79 OILT 0 66-2OP ©)
075'1668 LV€9
Oer'ozs9— | OSOTISST | SOT'ISST so9st— = *y |)
0 OLL'I6E1 98°670b
61b 67 — 6SL'619 Eve LLel LV€9
STLZ9ET 80'L7IP ©
Spe9zE— Ie 611
IIO'LZ79-— | ELTCPOTT | 6ST9STT 970619 @)
LOOT PIS — 7208'6r6
980;p0ET— | SLP'P6LI 6£7'LO8 Te611
9L9'br8 8y'ZSE9 S)
666 TE9S — 67 167
ZOMETIO— | 8L7'LOPG LLESL9 ()
8LO'ZIS ZO'88bL
$L9°0009 — T€611
980'p0EI— | 9L7'LE8 8E9' 8h ©)
861'S8E p7'0S9L
698°E6SS — Te61T
LOOTESL— | IZPOECT OIT'€61 @)
701 88987
681 £67 — LV€9
61P £67 — 0£7'0 11S°0 ®
0 0
0
7ulql pul] qi-ul UL XV Ql’A qa ZED
“ul “ul ‘ul pei pes
»-OLX IT] ensauy jo |(xyv)AZ=W | suonris uoneis wonoray 55
“01 x 4 AOles D ‘(xV),0N x “6)|| Ol x 0 se NEL
10 jUaWOW }ugUIOUI usamjoq | Surpasaig 10 &8
uonsayaqd jueisuo) =i adojs (XV)01 W KS
adVloaAy 0L <= Sulpusg gouvisig | 1¥ 1e94S
uonei3aquy uaUIdI9UT
uonoayed TA
| uoneis g uwinjog Jo
0} aaneay adojs asvlaay
(p'Z eunBi4 0} 49491) |Z e]dwexy JO} speo] jed1W9A 0} aNp UO!I9a1}9q ~~L-Z PIGeL
 ‘ul/Ul Z78p'809— = ad = 40] X 2D ‘JueIsuoo uONeIseUT
CLT OEEDT —
ung
ZLZ'6EEPT = (XV)01 X OX = SHOddns BursVag uaamsaq sJUaUIEIOU! UOHS3YaP JO

er
0
OLO TESS

SOV @eOtONeVe
0787809 — OS8'S19F1 S8S 19bl £6t 8
90 T6SE
0

HAS T86 LES 6vL'LS

€£9 C6CI CTOTILE
CEU VLE Ss
COO TI8S— 9S OPPO! Ie7e60l 696 TP
678768 916679
9TS EV6y —
P96 9ITI — OTr6L9I SOL'6E8 PCT es
18S °98L TV PLe9
TLOSOPS —
SPL 8IS8— TI8BLL8 8SO'LTI 68L°CC
SES Lop 18°8689
909996 —
p96 9TTI — 99L' 818 €8¢ 60h TST'8S
Tew Ise 08°€L69
BES LOTS —
L96-LCOL — 9EL ELOT TSO9LT 6Ce OE

19
TS
LOSE =
(LIBS ELT— 9970 890°C

0
0

7 Urq] qi-ur “UL 'xy QA Wid Zo

uonosyeaqd

pel pex
ueisuod

X “ut

ge
x “ul

‘5-01 xX 1g7] ensouzjo |(xy)AZ=W | suonRig | woNIS uonseay

a MOiet
“901

01 x 46
“901

(xV)o01 x “YO 410 2g

yusWOW usamjeq | Burpedaig
0
4

adols (XV)01 3 USWOW

uonoayaq
=A
aoueysiq ye 1v9yS
suipucg
uoneszoiuy JUSUWSIOUT IBEIIAY
] uoneIs g uuINJOD Jo
0} aanejey edols IBRIIAY
Cisse ce ed Be e
(‘p-Z aanBi4 0} 49J91) 1-2 e}dwexg 10} speo] jejuOZOY OJ anp UO!9aIj9q = ?-L PIGeL

350
 BOM
Section 7-8: Shaft Deflection for Example 7-1

column 14. This “running total’”’ is recorded on the same line as the station
number and represents the deflection at that station. Because it is assumed
that the bearings are rigid,’° zeros are entered for the deflection at those
stations where there are bearings.
Where the shaft loads have to be separated into horizontal and vertical
components (as in our example problem), the resultant shaft deflection at
each station 1s obtained by calculating the square root of the sum of the
squares of the horizontal and vertical deflections at each station. Table 7-3
is a tabulation of these results for Example 7-1.

Table 7- 3 Resultant slopes (relative to station 1) and deflections for Example 7-1.

Station
number —>

Slope,
aX 10°, 693.402 | 1154.204 | 1303.555 | 1878.866 | 1917.926 | 2344.93]
rad

Deflection
yx 10°, 8253.036 | 7806.663 | 7132.641 | 453.540 12395.806
in.

The method of numerical integration, as outlined above, is readily applic-

able to shafts without any overhanging loads, to shafts with an overhanging
load at each end, and to shafts with only one overhanging load.
For shafts without overhang, all forces act between the bearing supports
and the table will not contain stations outside of the bearing supports.
Thus, if there is no shaft overhang, as in Example 7-1, Tables 7-1 and 7-2
will show recorded data for only eight stations.
On the other hand, shafts having an overhang at each bearing support
will show recorded data for stations outside of the bearing supports. Also,
it is necessary to change the algebraic sign for deflection increments and
integration constants for those stations to the left of the left bearing support.
This procedural change is required since the origin of the coordinate x-y
axis is placed at the left bearing support, reversing, thereby, the direction of
integration for those stations to the left of this support.
Of the results posted in Table 7-3, the following are of the greatest interest :
The vector sum of the average slope at the left bearing .is
0,, = 9.00000069 rad = 0 deg
The vector sum of the average slope at the right bearing (station 8) should
be considered from the left of station 8 and from the right of station 8.
The critical value is the larger of the two vector sums. Calculations show
that the vector sum to the right of the bearing at station 8 is more critical.
Therefore, we have
6,, = 0.0021314 rad x 0.122 deg
10 In the strictest sense, this assumption is not valid, except for light loads.
 OY
Chapter 7: Shafts, Keys and Couplings

The shaft deflections at the pulley (station 3), the 18-in. gear (station 6),
and the 12-in. gear (station 9) are respectively
d(pulley) + 0.007684 in.

6(18-in. gear) ~ 0.007133 in.

6(12-in. gear) ~ 9.012400 in.

The slopes at the 18-in. gear and the 12-in. gear respectively are

§(18-in. gear) = 0.0013036 rad ~ 0.0767 deg

0(12-in. gear) = 0.0023449 rad ~ 0.134 deg

The permissible misalignment tolerance for the Conrad type (also called
deep groove) ball bearings is +0.25 deg (see Chapter 9). Thus the bearings
selected are satisfactory because the slope at each bearing support is less
than the possible tolerance.
There are no clearly defined standards or restrictions concerning lateral
deflection of shafting. Lacking more specific information, the designer may
be guided by the following criteria [4]

1. For machinery shafting, the deflection should be no are than

0.001 in./ft of shaft length between bearing supports.
2. For shafts mounting good quality spur gears, the deflection at the
gear mesh should not exceed 0.005 in. (between gears) and the slope
should be limited to 0.0005 in./in. (that is, approximately 0.0286 deg).
3. For shafts mounting good quality bevel gears, the deflection at the
gear mesh should not exceed 0.003 in.

Our results indicate that the first restriction, as indicated above, has
been violated: 0.008253 in. > 0.001 x 3.3 = 0.0033 in. The second restric-
tion has also been violated: 0.007133in. and 0.01239 in. > 0.005 in.,
0.0767 deg and 0.134 deg > 0.0286 deg. Consequently, the shaft at both gear
locations must be stiffened.
As is often the case where gearing is involved, the shaft diameter is governed
by the dimensional restrictions imposed by the gearing. Thus, in Example 7-1
it is necessary to repeat the deflection calculations as performed in Tables 7-1
and 7-2. However, in repeating these calculations, the moment of inertia in
column 6 is expressed in terms of the diameter: J = 2D*/64. The required
diameter is then determined by equating the vector sum of the vertical and
horizontal deflections at the point of interest (that is, the 18-in. and 12-in.
gear) to the limiting restriction, namely, no larger than 0.005 in. In a like
manner, the required diameter for a limiting slope of 0.0005 in./in. can be
established. The larger of the two diameters would then be specified as the
diameter to be used for the design.
 353
Section 7-9: Shaft Design by Computer

SECTION 7-9

. Shaft Design by Computer

In order to demonstrate the use of the computer in coping with shaft
design problems, we used an RCA SPECTRA 70 machine to find a satis-
factory shaft diameter for Example 7-1 in preference to repeating the above
outlined numerical procedure. The results of this program are illustrated
in Figure 7-5. As can be seen from Figure 7-5, the diameter at station 9
(that is, the location of the 12-in. gear) has been increased from 2.559 in.
to 3.765 in. This new shaft size at the 12-in. diameter gear permits a shaft
slope of 0.0005 in./in. and a shaft deflection of 0.00264 in. Both of these
values satisfy the imposed restrictions. The diameters at the other shaft
stations have also increased in proportion to that required at station 9.
Undoubtedly, these other diameters can be reduced to 3.765 in., which
would result in a single diameter shaft. However, in so doing, the designer
must provide locating shoulders for the bearings and gears (for example,
snap rings). This final shaft design would also have to be checked for fatigue
stress, particularly at the geometric discontinuity caused by the snap ring
grooves. Verification of this design is left to the reader.
The other avenue ofinvestigation open to the designer in trying to reduce
the shaft diameter is to determine whether the overhang length of 10in.
(see Figure 7-4) can be shortened. Whether this is possible depends on how
this action will affect the location of other components of the machine of
which this shaft is only one part.
As stated, the computer program!’ used was written for an RCA
SPECTRA 70 machine, and compiled ona FAST FORTRAN COMPILER.
This special version of Fortran language allows free format input and output
statements. Therefore, some read and print statements will have to be
changed to conform with standard Fortran language. Other than this minor
adjustment, the program is completely compatible with Fortran.
The computer program is set up to calculate the shear, slope, and deflection
of a simply supported, stepped shaft with or without overhang. It also
offers the option of comparing the slope and deflection at a given shaft
station to respective limiting values cued into the computer by the designer.
Should these limiting values be exceeded (at the-station of interest), the
program will automatically request the computer to increase the shaft
diameter by a given percentage until the limiting values have been satisfied.
The program is capable of considering as many as 20 separate stations and
loads.
Input data to the program are entered in the natural sequence which is
the same as that used in the numerical procedure. To assist the reader in
understanding the program, a legend of symbols and description is presented
in Table 7-4.
'!' The computer program was developed by Dr. Jeffrey Schram of The Newark College of
Engineering. The authors are indebted to him for his kind assistance and contribution in
helping make this portion of the book more meaningful.
 dia.
in.
5.517
in.
3.765
dia. 4.414
dia
in. dia.
4.414
in. 3.765
dia.
in.

Station Q)

Figure 7-5 Shaft for Example 7-1 determined by computer analysis.

Table 7-4 Computer Legend

Symbol Description
NS I I I

OPT Option to check whether the slope and deflection

are within given limits (ON = 2, OFF = 1) -

NP Number of stations for which the user desires

to compare the slope and deflections with the
limiting values (must be at least equal to 1)

LP(I)” Station number at which the limiting values of

slope and deflection are to be compared with the
actual slope and deflection

DM(1)” Maximum deflection at LP(1)

SM()- Maximum slope at LP(I)

KSET KSET = | means that the shaft loading is only

in the vertical plane. KSET = 2 means that
the shaft loading is in both the vertical and
horizontal plane.

N Number of stations

NO Number of station up to the overhang

pt Diameter of the shaft at a given station

E(1)* Modulus ofelasticity at each station

X(1)" Distance between stations

F(I, 1) Forces in the vertical direction at each station

F(I, 2) Forces in the horizontal direction at each station

(This data is included only if KSET = 2)
a
} These quantities should appear in the following order:
Dil), E(), X(1). D(2), E(2), X(2)... DON). E(N). X(N)
¥ . . 5 .
These quantities should appear in the following order
LP(I1), DM(1), SM(1)

354
 BoD:
Section 7-9: Shaft Design by Computer

The actual program used is

L&G70 FORTRAN
DIMENSION D(19),E(19),X(19) V(20),DBM(20),BM(20),AI(20),Y(19
1),WK(19),WL(19),Z(19),DSL(20),SLO(20),U(19),DD(19),C(19),
DEFL(20),
2DEFX(20),SLX(20),UX(20)
$LP(20), SM(20), DM(20) ,F(20,2)
READ,OPT,NP
DORI Zia ig——P NP
2a READ, LP(L), DM(L), SM(L)
READ, KSET
READ, N,NO
MO =NO —1
M=N-—1
U(N) =0.0
DOF 19 1-M
READ,D(J),E(J),X(J)
DO 491 KZ =1,KSET
DXO) Bae SN
READ, F(J,KZ)
CONTINUE
DO 100 KZ =1,KSET
PRINT 50
FORMAT (‘1',16X,,SECTION’,9X,,FORCE’,12X,/DIAMETER’,9X,
MODULUS’,
$9X,’LENGTH’)
PRINT, 1,F(1,KZ)
DO 80 J=1,M
K=J +1
80 PRINT,K,F(K,KZ) ,D(J) , E(J) , X(J)
IND =O
V(1) =0.0
DO 1 J=2,/N
K=J-—1
V(J) =V(K) +F(K,KZ)
DBM(1) =0.0
DO2) —=2)N
K=J—-1
DBM(J) =V(J) *X(K)
BM(1) =0.0
DO!3 J =2)N
K=J-1
BM(J) =BM(K) +DBM(J)
IF (ABS(BM(J)) —1.) 31,31,3
BM(J) =0.
CONTINUE
DO4J=1,M
Al(J) =3.14159*D(J) **4/64
DO5J=1,M
Y(J) =E(J) *AI(J)
DO6J=1,M
WK(J) =BM(J)/Y(J)
DO7J=1,M
ad) 24
WL(J) =BM(K)/Y(J)
DO8J=1,M
Z(J) =.5*(WK(J) +WL(J))
DO9J=1,M
DSL(J) =X(J) *Z(J)
SLO(1) =0.0
DO 10 J =2,N
Kasie4
10 SLO(J) =SLO(K) +DSL(K)
DO 28 J =2,N
 356
Chapter 7: Shafts, Keys and Couplings

28 SLO(J) =SLO(J)/1000000.
DOM =m
K=J +1
11 U(J) =.5*(SLO(J) +SLO(K))
DOM2 1M
12 DD(J) =U(J) *X(J)
DDSUM =0.0
DO 13 J =1,MO
13 DDSUM =DDSUM +DD(J)
XSUM =0.0
DO 14J=1,MO
14 XSUM =XSUM +X(J)
CF = —DDSUM/XSUM
DOW S i= 17M
1185) CW) =xX@)+CF
DEFL(1) =0.0
DEFL(N) =0.0
DO 26 J =2,N
K=J—1
26 DEFL(J) =DEFL(K) +C(K) +DD(K)
PRINT 40
40 FORMAT (‘0’, 16X,,'SECTION’,9X,’FORCE’,15X,,
SHEAR’,
&9X,’ SLOPE ‘9X,’ AVERAGE SLOPE’, 9X,,DEFLECTION’ )
DOR 75s J—aEN
ADEF =ABS(DEFL(J))
IF (ADEF —.00001) 29,29,17
29 DEFL(J) =O.
17 PRINT, J, F(J,KZ), V(J),SLO(J), U(J) ,DEFL(J)
IF(KSET —1) 53,53,55
55 IF(KZ —1) 56,56,57
56 DO 58J=1,N
DEFX(J) =DEFL(J)
SLX(J) =SLO(J)
58 UX(J) =U(J)
100 CONTINUE
57 DO 71 J =1,N
DEFL(J) =SQRT(DEFL(J) *DEFL(J) + DEFX(J) *DEFX(J))
U(J) =SQRT(UX(J) *UX(J) + U(J) *U(J))
1 SLO(J) =SQRT(SLX(J) *SLX(J) +SLO(J) *SLO(J))
PRINT 83
83 FORMAT(‘0’,’ RESULTANT SLOPE AND DEFLECTION’)
PRINT 84
84 FORMAT(‘0’,16X,’'SLOPE’,8X,'AVERAGE SLOPE’,8X,’ DEFLECTION’)
DO 88 J =1,N
88 PRINT, SLO(J), U(J), DEFL(J)
53 IF(OPT —1.)99,99,160
160 DO 175 KP =1,NP
L =LP(KP)
IF(ABS(DEFL(L)) —-DM(KP)) 172,170,170
170 IND =1
172 IF(SM(KP) —ABS(SLO(L))) 174,175,175
174 IND =1
175 CONTINUE
IF (IND) 99,99,180
180 DO 185 J =1,M
PRINT 240, D(J), J
240 FORMAT(‘0’,10X,/DIAMETER = ‘, F10.4,,AT SECTION — ',12)
D(J) =D(J) *1.213
185 CONTINUE
GO TO 490
99 STOP
END
*RUN
 Bye
Section 7-10: Critical Speed of Shafts

SECTION 7-10
- Critical Speed of Shafts
The practicalities involved in manufacturing and assembly prevent the
center of mass of a symmetric, rotating system from coinciding with the
actual center of rotation. In addition, the static deflection caused by the
weights of the components being carried by the shaft cause a further shift
of mass center from the axis of rotation. Consequently, as the shaft increases
in rotational speed, the kinetic energy of the attached masses increases.
When the kinetic energy reaches the potential energy of the shaft caused
by the static deflection of the mounted masses, the shaft will begin to vibrate
violently. The rotational speed at which occurs is
called
. As in other vibratory
systems, higher critical speeds (at smaller amplitudes) can occur at some
multiple (that is, harmonic) of the fundamental frequency.
Equating the kinetic energy due to the rotation of the mounted shaft
masses to the potential energy of the deflected shaft results in an
. This equation is called he Raleigh
and its derivation may be found in any good textbook dealing
with mechanical vibrations.'? Thus, the lowest or fundamental critical
speed for a shaft simply supported at two points is

W, W. Way, tee Woy ale

ne = 18727] (Naa as) =| (7-15)
Wiryt + Wry3 + Wsy3 +--- + Wyn,
where W,, = the weight of the rotating mass in pounds and j,, = the
respective static shaft deflection due to W,, in inches.
For Example 7-1, it was suggested in Section 7.9 that the designer might
be likely to consider a shaft of uniform diameter, namely 3.765-in. diameter.
Considering the pulley weight of 100 1b, the 18-in. diameter gear weight of
50 1b and the 12-in. diameter gear weight of 25lb, we can determine the
static shaft deflection at each of these loads. This is simply accomplished
by means ofthe computer or some other convenient method (that is, numerical
or graphical). In this particular instance, with few loads and a uniform shaft
diameter, the area-moment method would be most direct—lacking a
computer. Nevertheless, the deflections at the 25-lb, 50-lb, and 100-lb loads,
respectively, are 0.538 x 107 ?in., 0.427 x 1073in., and 0.494 x 107? in.
Substituting in equation (7-15) we have

25(0.538 x 1073) + 50(0.427 x 107%) Sle

+ 100(0.494 x 1073)
n, = 187.7
25(0:538% 10% 2)> + 50(0:427_« 1052)7
+ 100(0.494 x 1073)?
n, = 8560 rpm

12 W. T. Thompson: Vibrations Theory and Applications, Prentice-Hall, Inc., Englewood Cliffs,

N.J., R. T. Anderson: Fundamentals of Mechanical Vibrations, The Macmillan Co., New York.
 358
Chapter 7: Shafts, Keys and Couplings

Recall that the shaft is to operate at 360 rpm—well below the critical
speed. Therefore, we do not anticipate any vibration problem resulting from
shaft rotation.
Actual tests have shown that the Rayleigh equation is slightly overly
optimistic since self-damping of the materials and the increased stiffness
attributable to gear hubs, and so on, tend to lower the calculated critical
speed. Consequently, a good “rule of thumb” in practice is to keep the

SECTION 7-11
Torsional Stiffness
Another important aspect to consider in shaft design 1s torsional stiffness.
Proper design requires that a shaft be able to transmit power uniformly
and with a steady motion. Shafts that permit an excessive angular displace-
ment may contribute to vibrations (both torsional and lateral), affect ‘gear
action, and cause premature bearing wear or failure. In addition, particular
applications such as machine tools require that the spindles be especially
rigid.
Although no standard torsional deflection has ever been established for
different shaft applications, it has become standard practice to limit the
torsional deflection for machinery shafting to 0.08 deg/ft of length [4] and
transmission shafting to 1 deg in a length of 20 times the shaft diameter [5].
Camshafts, particularly for internal combustion engines, should be less than
0.5 deg regardless of the shaft length [5].
The equation established for torsion (see Chapter 5) is

oS fe (7-16)
JIG ;
The polar moment ofinertia for a round hollow shaft is
T nD? De
J = 39!
—(D?e — Df)
;) = 39
—-|11 -Soa (7-17)
=

Substituting equation (7-17) into equation (7-16) and multiplying by

57.3 deg/rad, we arrive at
S84T L
6 (7-18)
D3 = DE/DAG
where @ is the torsional deflection in degrees for a shaft length L in inches.
As indicated earlier in this chapter, we can use this equation for a solid
shaft by making the internal diameter, D; identically zero.
To illustrate the application of equation (7-18), we will apply it to Example
7-1. Figure 7-6 shows a schematic representation of the torques acting on
the shaft and their locations. From Figure 7-6 we have
Tog = 5250b-in.
Tec = 3150 Ib-in.
 Boy
Section 7-12: Torsion of Shafts of Various Cross Sections

Bearing Bearing
5250 Ib-in. Out Out
12 in. 18 in. 10 in. eof10 in.

In 2100 Ib-in. 3150 Ib-in.

Figure 7-6 Torque distribution for shaft in Example 7-1.

Using the shaft diameter of 3.765 in. established in Section 7-9 and equation
(7-18) we can calculate the torsional deflection of the shaft from E, the
location of the 12-in. diameter gear to B, the location of the pulley.

584
Opp =Onot Oca = [(5250)(18) + (3150)(20)]
1D 10-6765)"

The result is well below the torsional stiffness conditions stipulated above.
Specifically for transmission shafting, we are allowed to 1 deg per 20 times
the shaft diameter. Thus, in this instance,

1 deg 38
ee
20(3.765)in. esin = id |S|
0x a SU
0.504 L
deg > 0.038 deg

Therefore, the shaft is sufficiently rigid. If the shaft were considered to be

machinery shafting (rather than transmission shafting), we would then have

38 in.
0.08 deg/ft x = 0.254 deg > 0.038 deg
12 in./ft

Again, we find that the shaft would be sufficiently stiff. It is of interest to

note that, had the designer decided to use a stepped shaft (Figure 7-5),
calculations would have proven it to be even stiffer than the straight shaft.

SECTION 7-12

Torsion of Shafts of Various Cross Sections

At various times, the designer must concern himself with shafts having
other than round cross sections. The actual stress analysis is rather complex
(see Chapter 5), with most of the results having been obtained by the
membrane analysis [6, 7]. Several shapes (for example, ellipse, square,
rectangle, triangle), however, have been treated rationally [6]. Table 7-5 is
presented here as an aid to the designer. Nevertheless, the reader is advised
to consult the many excellent references dealing with this topic for a complete
“in-depth” analysis [6, 8, 9].
Table 7-5 Formulas for torsional deflection and shearing stress of various
cross-sectional shapes

General formulas: 0 = ce T= - where 6 = angle of twist, radians; T = twisting moment, in.-lb; L = length, in.; 7 = unit shear stress, psi;
G = modulus ofrigidity, psi;K,in.*; and Q, in.? are functions ofthe cross section.

Shape Formula for K in @ = AED Formula for Shear Stress

KG

nd4 16T
7, max. at the boundary

= eit
'/35n(d4 —d)4) md4 —d, ) , max, at the outer
st boundary

assumed that r<r

Edt: 4 1.87¢
— 2i3nrt3,r = mean radius max. at both
dare 4? PP? edges away from
the ends

on
T= Tab? > Max. at the boundary
of the minor axis

ee Ta 3p,3
3b,? nee a A
SSS 2T ee
a ’ ae
K Tee a,{(1+q) 1] = ma, 6,2((1 +q)* — 1] outer boundary
of the minor axis

PE ; é
a , Max. at the midpoint
of each side

T= LOT max. at the midpoint

of each side

b3 bh 4 7

K= — [-8 — 3.367 ( mall t= Sa + LBA if max. at the, midpoint

a*b of each longer side

df Sse
re 5 Ta
ea ay ene near midpoint
= ila = HP =P ma MA ah i fs sd of f the longer side.
i

. aly + bt) — ty? — 2 Ce a near Midpoint

KN
a
213(a— t2)(b — 1) of the shorter side.
(Higher stresses occur at the inside corners.)

4.8T SAB,
| K = 0.140664 T= —— ,max. at the midpoint of
— b—»}
P 6? each side,

SOURCE: Colin Carmichael (ed.): Kent's Mechanical Engineers’ Handbook, Vol. IN, (2th ed., John Wiley and Sons, Inc.. New York, 1961.

360
 361
Section 7-13: Keys

SECTION 7-13
. Keys
Keys are used to prevent relative motion between a shaft and machine
elements such as gears, pulleys, sprockets, cams, levers, flywheels, impellers,
and so on. There are numerous kinds of keys (some of which have been
standardized) for various design requirements. The particular type of key
specified will depend upon the magnitude of the torque transmitted, type
of loading (that is, steady, varying, or oscillatory), fit required, limiting
shaft stress, and cost. Figure 7-7 shows the many types of keyed connections.
Of all the types shown in Figure 7-7, the keys most frequently used are
the square key, the tapered key, and the Woodruff key.
Table 7-6 Dimensions of square plain parallel stock keys (dimensions in inches, see
Figure 7-7a on next page)

Bottom ‘ Bottom Bottom Wi Bottom

Width of Key- Width of Key- of Key- see of Key-
Shaft and seat to and seat to seat to ane seat to
Diam- Thick- Opposite Thick- | Qpposite - Opposite - Thick- Opposite
eter ex of side of ness of Side of side of ness of | cide of
ae Shaft, : , Shaft, ; Shaft,

z i wi . 5 5 3.309
8 16 8 8
2 a it 3 ae 5 3373
16 8 2 8 “16 8

A8 316 216 28 a42 28 3.437

u 3 i 3 op 8 3.690
16 16 8 8 =e} 8
3 Bi ;u 3 3 3 3.881
4 16 16 8 4 8
13 23. 13 3 51 3 3.944
16 16 4 8 “8 4
7 3 )43 J 315 3 jl 4.042
8 16 16 2 “16 4 4

115 | 4 '8
7 1
:
3 2
3
fe)
:
|re4282
1 11 Ws
IG 15 3h
33 34 U
14 2
4.296

it1 af
1 2 3
ue a7ul 3; izil | 4.550
itI t1 ali
oe 1i a3 i& ail.
ie 1;1 | 4.740
3
] iz 41 1
23 15 3
37% I8 1
55 4W 4.803

a1 1
ri a.
256 il
5) i
35 78 3
SZ i
Is 4.900

1 =5 5 ul
aa 15 a
38 a8 15
6 il
15 5.091

3
iS 5
fe 25SE
73 B)
8 3
34 if
8 6 |il
5 SSS)

SOURCE: ASME: ANSI Standard B17.1-1967.

*Tolerance on W is - 0.0020 in. for shafts 7 to it in., inclusive; -0.0025 in. for shafts i to 3h in., inclusive; -0.0030 in. for shafts 3g to 6 in., inclusive.
 eanbig
g-Z snoueasadky
Jo yeys ‘shay

362
VZ
5

s
Y

‘ul
sad

at
jade)
7

BsaqpAay
ion
(4)

QZ
ASIKCY

fy
(4)
sayiea4
Aay payiasu!
Ul ay) "qny
 363
Section 7-13: Keys

Table 7-7 Dimensions of square and flat taper stock keys (dimensions in
inches, [see Figure 7-7d])

Stock Length, L

Shaft Diameter Advanc-

(inclusive) On Mini- Maxi- ing by
Height Incre-
ments of

+0.0020
5/g - 7g +0.0020 3/g
Wii 6-11 /4 +0.0020 I/5
1 5/16-13/g + 0.0020 S/gt
1 7/16-13/4 +0.0020 3/4
113/) 6-214 +0,0025 I
2 9/1 6-23/4 +0.0025 I1/q
pi leer Ps +0.0025 11/5
3 3/g -33/4 +0.0030 13/4
3 7/g -41/5 + 0.0030 2
4 3/4 -51/5 +0.0030 21/5
5 + 0.0030
SOURCE: ASME: ANSI Standard B17.1—1967.
*This height of the key is measured at the distance W, equal to the width of the key, from the large end.
14 I/y in. length instead of 4 3/g in.

Because the actual stress distribution for keyed connections is not

completely understood, a factor of safety of 1.5 should be used when the
torque is steady. For minor shock loads, a factor of safety of 2.5 should
be used, and one up to 4.5 should be used for high shock (especially if the
loads are reversible). To prevent the hub from rocking on the shaft when
using a straight key (that is, no taper) and to assure a good grip, the hub
length should be at least 25 % larger than the shaft diameter. The minimum
key length should also be at least 25% larger than the shaft diameter. The
hub is usually kept in place on a straight key by means of set screws. When
set screws are not used, a light press fit between the hub and shaft should
be specified [4].
Dimensions for various types of keys have been standardized'? and can
be found in engineering handbooks. Data for sizing only the square, tapered,
and Woodruff keys are listed in Tables 7-6, 7-7, and 7-8, respectively. The
letters at the top of these tables correspond to Figures 7-7a, b, and g.

'3 ANSI Standard B17.1-1967.

 364
Chapter 7: Shafts, Keys and Couplings

Table 7-8 Abbreviated list of Woodruff key dimen-

sions (see Figure 7-7g on page 362)

1 Distance
Nomina ,
Suggested : of Key, Below | Shearing
ee igi 2 ae in. Center” Area,
Oo. es
in. in.
At X Bin.

204 3 ; 0.030
305 4 2 0.052
4o5| : 2 0.072
Ws)be
Meh, ye 23
208 6m 6 4

gio| Sees | + x14]Po 0547 |0537]

14-13 2gq | 0.296
312} 14-13
Arh gealea|1x14]S19. 0641
O }06031 | 464 | 0356
1012| 143-24 |S x15 |0641 |0.631]
IS) py Jk Hhle gn adl gq | 0.438
wil

Te ||3x13eee| oat
1212 | 14-24 0.517
SOURCE: ASME: AWSI Standard B17,2—1967.
*This dimension is given to help make drawings and layouts,
+The key extends into the hub a distance of A/2.

Standard square key

A standard square key is perhaps the most common type of key where the
dimension W is equal to one quarter of the shaft diameter (see Table 7-6).
As shown in Figure 7-7, it is fitted so that one half of the thickness is in the
shaft and the other half is in the hub. In the hub, the keyway is cut by a
key-seating machine or by broaching. The keyway in the shaft is cut by
milling as shown in Figure 7-8. A plain milling cutter having a face width W
is used to cut the “sled runner” keyway. An end mill with a diameter equal
to the width W ofthe key is used to cut the profiled keyway. Where possible,
the sled runner type should be used since the stress concentration for this
kind of keyway is less than that for a profiled keyway. However, near a
change of shaft diameter, the “sled runner’? keyway may extend into the
shaft shoulder and thus compound the stress concentration (that is, the
fillet radius and the keyway). In such cases, it may be advisable to use a
profiled keyway which would not extend to the fillet radius.
 365
Section 7-13: Keys

GAS

Sled runner keyway Profiled keyway

(a) (b)

Figure 7-8 Method of cutting keyways in shafts.

If standard square or flat plain parallel keystock is used, the fit in the
shaft can be “‘line-to-line”’ or slightly loose with deliberate clearance between
the top of the key and the slot in the hub. As a result, the key must be held
in place with a set screw through the hub and pressed on the top of the key.
For very wide keys, the key can be held in place with countersunk flat head
or cap screws if the shaft is not weakened.
The possible loose fit of a key presents a complex stress analysis problem
that is not completely understood. Figure 7-9 shows an exaggerated illustra-
tion of how the position of the key is rotated, producing a severe and complex
stress distribution on the sides and edges of the key as well as the edges of
the keyway.
When it is important to insure a tight fit for the key in the shaft, the
designer cannot use standard keystock. Instead, he must specify a tolerance
on the key that will provide a slight interference fit. However, for purposes

Va Figure 7-9 A loose fit in the keyway.

 366
Chapter 7: Shafts, Keys and Couplings

Figure 7-10 A tight fit in the keyway.

of calculation, it is assumed that a stock key is fitted as shown in Figure 7-10,

implying, thereby, that the reactive forces shown are uniformly distributed
along the full length of the key. The procedure followed in sizing a key is
illustrated by Example 7-2.

Example 7-2. A gray cast iron gear with a 3-in. long hub transmits 1200 lb-ft of torque
to a steel shaft. The gear material is ASTM 40, the 2-in. diameter shaft is made
of AISI 1045 cold-drawn steel, and the key is to be cut from AISI 1020 cold-
drawn square stock. If the transmitted torque produces intermittent minor
shocks, find (1) the minimum length of key and (2) the length of key, if the key
and shaft are to be of equal strength and of the same material.
Solution: From Table A-1 in Appendix A for ATSM 40, we obtain
S,¢ = 140,000 psi. From Table A-2 in Appendix A for AISI C1020 and C1045,
we have, respectively, S,,, = 66,000 psi and S,,, = 85,000 psi. For intermittent
shock we assume the factor of safety N = 2.0.
Because the precise position of the force F (see Figure 7-10) is not exactly
known, the reactive force on the key is assumed to act at the outside shaft
diameter.
Thus, the transmitted torque is
ts) (7-19)

If it is assumed that the shear stress on the key acts in that plane of the key which
is tangent at the shaft diameter, we have
je Ei
Ss, == (7-20)
A WL
where A is the shear area across the key.
Substituting equation (7-20) into equation (7-19) for F, we obtain

if
_ S,WLD (7-21)'*
»
where S, is the design shear stress and L is the length of key.

'* Careful inspection of equations (7-21) and (7-23) will show that, if the maximum shear theory
of failure is used (i.€., S,,,, = 0.5S,,), then a square key is equally strong in shear and compression.
The reader can verify this statement by equating the torques of each equation.
 367
Section 7-13: Keys

Since for a square or flat key one half of the thickness lies within the shaft
and the other half lies within the hub of the gear (pulley, cam, and so on), there
is a compressive or bearing stress on the side ofthe key, the side of the keyway,
and the side of the hub. The compressive stress is expressed as

S,=—= (7-22)
where A is the bear ng (that is, ) compression) area. Substituting equation
(7-22) into equation (7-19) for F, we obtain
_ S.WLD
it (f-23)°
ean
where S, is the design bearing (compression) stress.
Part /. Using equation (3-6), S,,,syp = 0.58S,,, yp? and with N = 2, we have for the
key

ce a eosin

Referring to Table 7-6, we find that for a 2-in. diameter shaft, a 3-in. square
key is required. Thus, using equation (7-21), the length of the key based on
shear is
OT: 2(1200)(12)
= Ilia
ie S,WD 19,150(0.5)(2)
Using equation (7-23) and a design stress of S, = S,,/N = 66,000/2 = 33,000
psi, the length of the key based on compression (that is, bearing) is

_ 4T _ 4(1200)(12)
IL, = Is
~ S.WD — 33,000(0.5)(2)
It is obvious that further calculation for the length of key based on the
bearing stress of the keyway in the shaft or bearing in the keyway slot of the
cast iron hub is not necessary because in both cases the value of S, is greater
than that for the key. As a result, the length of key required for these two cases
will be smaller than either of the above calculations. From these calculations,
we conclude that a 4-in. square key, 1.75in. long is required. However, in
accordance with the statement made in the earlier part of this section, the
minimum key length must be at least 25°% larger than the shaft diameter or
2.5 in. long in this application. To insure that the key does not rock in the axial
plane of the shaft when the set screw in the hub is brought against its surface
and tightened, we shall finally specify that the length of the key be 3.00 in. With
a key of this length, the factor of safety in bearing (compression) becomes
4T — 4(1200)(12)
= 19,200 psi
= De 3(0.5)(2)
Thus

66,000
= —— = 3.45
19,200 :

'S See footnote concerning equation (7-21).

 368
Chapter 7: Shafts, Keys and Couplings

Part 2. From tests performed at the University of Illinois it was determined

that the torsional strength of a solid round shaft was reduced by a keyway in
accordance with the empirical formula for relative strength given in equation
(7-24).
e = 1.0 — 0.2w — 1.1h (7-24)
where e¢ is the relative strength, w is the ratio of the width of the keyway to the
diameter of the shaft, and his the ratio of the depth of the keyway to the shaft
diameter. In this case we have
0.5 0.25
e=10-0.2 5] — 1.1 ea = 0.8125

Therefore the torque that the shaft can transmit is

S,nD?
i 081252" ==) S125 (7-25)
16
Equating equation (7-25) to equation (7-21) (that is, the strength of the key in
shear), we have
S,xD* _S,WLD
0.8125 if
or
nD? m2)?
TU 15 en
0.8125 SW 0.8125 (0.5) = Sit

If we now replace S, in equation (7-25) by 0.58S,,,, we can equate this torque to

equation (7-23) (that is, the strength of the key in bearing). Thus

0.8125
0.58S,,2D? _ S.WLD
16 4

Since S,, = S, for a ductile material

n(2)?
L = (0.8125)(0
(®. 8) 5g)
gy==0:
0.471 = 2.96
2.961 in.
4(0.5)
use L = 3.00in. @

Standard flat key

The standard
flat key (Figure 7-7b) is used where the hub of the gear,
pulley, and so on, is
thin. The depth of the keyseat in the shaft is standard,
whereas the depth in the hub is shallow. The extra thin flat key (Figure 7-7c)
has a shallow keyseat in both the shaft and the hub. It is intended for use
where the shaft is hollow and the hub is thin.
Tapered keys depend on the frictional contact between the hub and the
shaft to transmit the torque. The most common type of tapered key is
 369
Section 7-13: Keys

(a) (b) (c)

Figure 7-11 Shaft‘and hub assembly using a tapered key. (a) The assembly.
(b) Shaft loading. (c) Key loading.

either the Gib-Head type or the Plain type (see Figure 7-7d). This key
(tapered) is used to transmit medium level torques and is ‘“‘locked”’ in place
radially and axially by the wedging action of the key between the hub and
the shaft. Sometimes, tapered keys are also designed to have bearing pressure
on the sides when fitted in the keyway. As shown in Figure (7-7d), the top
surface of the key has a standard taper of } in./ft of length, which complies
with a mating taper in the hub. The lower part of the key surface is flat and
is sunk into the shaft. The Gib-Head type permits the key to be readily
removed by the prying action of a tool (for example, a screwdriver) between
the hub and the Gib head. However, it is advisable when specifying such a
key that the Gib head be so located as not to project beyond the end of a
shaft for obvious reasons of safety. If possible, a plain flat taper stock key
should be used.
As stated above, the transmitted torque between the shaft and the hub
depends upon the frictional force. This torque can be determined by balancing
the forces on the free body diagram of the shaft (refer to Figure 7-11). From
Figure 7-11b we obtain

T =3u,ND +4y,ND
But

N = LS,
Thus

T =+(u, + u,)DWLS, (7-26)

Using the values of 1, and py, stated below, this equation becomes
T = 0.135DWLS, (7-27)
 370
Chapter 7: Shafts, Keys and Couplings

where

T = transmitted torque, Ib-in.

1, = coefficient of friction between shaft and hub. A reasonable value to
assume is 0.25 [5].
> = coefficient of friction between the key and the shaft keyseat. Good
practice dictates that the key be greased and so 1, can be taken as
0.10 [5].
S. = compressive stress in key, psi.
From Figure 7-l11c we can obtain the force F required to drive the key
into the assembly by summing the horizontal forces
F=2p,N + Ntana = WLS(2u, + tan a) (7-28)

Using yw, = 0.10 and tana = 0.0104 (that is, value of {in./ft of taper),
equation (7-28) becomes

F = 0.2104WLS. (7-29)
Where light, steady, nonoscillating loads are to be transmitted, and a
keyseat would weaken the shaft excessively, a flat key (Figure 7-7e) can be
used. The key can be flat or tapered on one side. When one side is flat, a
set screw through the hub is necessary to keep the key in place. On the other
hand, when the surface of the key that contacts the slot in the hub is tapered,
a wedging action takes place at assembly, and equations (7-26) and (7-27)
can be used to establish the key size and its torque capacity.
As can be observed in Figure 7-7e, the flat key requires that a flat surface
be milled on the shaft to seat the key.

Woodruff key
The Woodruff key (Figure 7-7g) is a light duty key which, because of its
deeper keyseat, weakens the shaft but has the advantage of being able to
align itself readily with the hub, as it is free to rotate within the semicircular
keyseat. It is widely used in the automotive and machine tool industries,
and is usually SSN eicannecmn AY Thistype
of key has the advantage of (1) not tipping or rolling over because of the
deep keyseat, (2) readily adjusting itself to the mating hub due to its ability
to rotate freely within the keyseat, and (3) being well adapted for use in
fitting tapered shafts to hubs.

. A special milling cutter is used

to cut the keyway in the shaft, allowing for clearance in both the diameter
and width.
Dimensions for standard Woodruff keys are given in Table 7-8. From
this table, it is possible to calculate the torque carrying capacity of a
Woodruff key.
 Sul
Section 7-13: Keys

Example 7-3. A gear that transmits 4000 Ib-in. of torque is to be keyed to a 1}-in.
diameter shaft by a j by 13(that is, Key No.809) Woodruff key. What is the factor
of safety of the key for both shear and bearing? The key material is AISI 1030.
Solution: From Table A-2 in Appendix A for cold-drawn 1030 steel
S,» = 76,000 psi. Thus, to obtain the allowable shearing stress we use equation
(3-6)
S.ypSyp = 0-588, = 0.58(76,000) = 44,100 psi
Now, the torque capacity of the key in shear is
D 1.
1D ed ee (4,100)0262)[15°] = 7220 Ib-in.
Syp 5)

To obtain the torque carrying capacity of the key in bearing, we need to know
the bearing area. From the second footnote in Table 7-8, we know that the
portion of the key that projects into the hub is equal to one half of the A
dimension in Table 7-8. With this information, we can establish the bearing
area A,. Thus, the torque carrying capacity of the key in bearing is
D 0.25 125
T = S,,A,~ = 76,000 {— x 1.125] |——] = 6670 lb-in.
TN ape: 2 2
Finally,
N(shear) = 7220/4000 = 1.81
N(bearing) = 6670/4000 = 1.67. @

For light duty or medium operation, it is also possible to use a taper

pin'® as a key, as shown in Figure 7-7f, but the taper is 4 in,/ft (that is,
Morse standard paper pins) rather than ; in./ft as indicated in the figure.

ayy

Figure 7-12 A grooved pin used for holding a hub to a shaft. These pins do
not require the hole to be reamed. The assembly is held in place by elastic
deformation of the edges of the three sectors, and the effective diameters of these
edges are enlarged by the three axial grooves.

16 For sizes and data on taper pins, see ANSI Standard B4.20-1958.
 372
Chapter 7: Shafts, Keys and Couplings

Figure 7-13 A roll pin as used to fix a hub to a shaft. The assembly is
held together by the spring action of the pin.

The tapered hole is reamed at assembly so that half the pin diameter is in
the hub and the other half of the diameter is in the shaft.
Quite often, straight dowel pins, grooved pins (see Sigure 7-12) or roll pins
(see Figure 7-13) are used to fix a hub to a shaft. These methods of fixing a
hub to a shaft are limited to light duty service.

Heavy duty keys

Heavy duty keys are the Nordberg key (Figure 7-7f), the Kennedy key,
the Lewis key, and the Barth key, of which the latter three are not shown in
Figure 7-7.
The Nordberg key (Nordberg Manufacturing Company) is a pin type key
that is available both in straight and tapered shapes. However, for heavy
duty service, the tapered shape is used. The key is driven into a reamed
hole at the end of the shaft. As with the Morse taper pin, half of the pin
diameter is located in the shaft, and half is located in the hub. As indicated
in Figure 7-7f, the taper is > in./ft.
A feather key (Figure 7-7h) is used when it is necessary to permit a hub to
have axial movement along the shaft and to prevent any rotation between the
shaft and the hub. The key is either screwed to the shaft with a running fit in
the hub or is held in the hub with a running fit in the shaft. Generally, the
former design is preferred. When two keys 180 deg apart are used, analysis
indicates [5] that the axial friction force required to move the hub along the
shaft is one half of that required when one key is used. As a guide in sizing a
feather key, the bearing pressure on its side should not exceed 1000 psi [12].
 B73
Section 7-14: Splines

SECTION 7-14

- Splines
The keys discussed in Section 7.13 tend to weaken a shaft because of the
keyway. By cutting splines, which are integral with the shaft, the effect is
the same as having added multiple keys without having weakened the shaft
to the same degree. There are two forms of splines: (1) straight splines and
(2) involute splines.

Straight splines
Straight splines are the older of the two forms (adopted by the SAE in
1914) and are generally being replaced by the stronger involute forms.
Nevertheless, many machine tools, automotive equipment, and so on,
employ this form, and some designers continue to use it because of its
relative simplicity. Figure 7-14 shows the standard SAE 4, 6, 10, and 16
spline fittings. Table 7-9 establishes the maximum dimensions for these
splines based on the nominal shaft diameter D.
To obtain a particular class of fit, the shaft dimensions are varied and
reflect the type of material used, its heat treatment, and the machining
method employed. The splined fittings (for which the dimensions in Table
7-9 and the SAE Handbook apply) are soft broached. The machine tool
industry varies the aforementioned method of establishing fits by varying
the spline fitting dimensions rather than the splined shaft dimensions. This

Table 7-9 Formulas for Dimensions of Splines

No. Standard All Permanent Not Slide Slide

of Sizes, Fits Fit Under Load Under Load

PRA
Re Sram
Nominal

by yin. from 3 0.241D | 0.075D | 0.850D |0.125D | 0.075D

: 1
6 in, to 13; Ah 2a 0.250D | 0.050D 0.075D | 0.850D | 0.100D | 0.800D

10 Same as above,
Le
plus by > in.
from 3 to 6 in. 0.156D | 0.045D | 0.910D | 0.070D | 0.860D |0.095D | 0.810D

16 by tin. from 2
to 6 in. 0.098D | 0.045D |0.910D | 0.070D | 0.860D | 0.095D | 0.810D

SOURCE: SAE Handbook. Society of Automotive Engineers, New York, 1948.

For D and d, tolerance allowed of —0.001 in. for shafts 2 to | - in., inclusive; —0.002 for shafts 2 to 3 in., inclusive; —0,003 in. for

shafts fy to 6 in., inclusive, for 4-, 6-, and 10-spline fittings; tolerance of —0.003 in. allowed for all sizes of 16-spline fittings.

For W, tolerance allowed of —0.002 in. for shafts 2 to is in., inclusive; —0,003 in. for shafts 2 to 6 in., inclusive, for 4-, 6-, and

10-spline fittings; tolerance of —0.003 allowed for all sizes of 16-spline fittings.
 3/4
Chapter 7: Shafts, Keys and Couplings

4 spline 6 spline 10 spline 16 spline

Figure 7-14 SAE standard straight spline fittings.

approach is preferred because a uniform grinding operation can be performed

on the shaft. Thus, should a splined shaft have a free fit and a press fit with
two different spline fittings, only one shaft grinding operation is needed—the
required fits being obtained by varying the dimensions of the fittings. The
machine tool industry requires a wider range of fits than those listed in
Table 7-9 and has, therefore, designated the following fits: free fit, sliding
fit, push fit, light drive fit, and press fit. Specifications concerning these
fits may be obtained from reference [13].
According to the SAE, the equation for the theoretical
torque capacity of
straight
sided splines with sliding (based on 1000-psi spline pressure) is
T = 1000nr,,hL (7-30)
where

T = theoretical torque capacity, lb/in.

n = number of splines
lm = (D + d)/4 in. (see Figure 7-14)
h = depth of spline, in. (see Figure 7-14)
L = length of spline contact, in.

involute splines
— —— are becoming the predominant spline form because they

These splines have the gen orm of internal and external involute gear
teeth (see Chapter 10) with a 30 deg pressure angle'’ and one half the depth
ofastandard gear tooth. The external spline can be formed either by hobbing
or by a gear shaper. Internal splines are formed by broaching or by a gear
shaper. To control tolerances, the minimum effective space width and the
minimum major diameter ofthe internal spline are held to basic dimensions.
The external spline, however, is varied to obtain the desired fit. The advan-
tages of involute splines are (1) maximum strength at the base of the tooth,
(2) accuracy of spacing and the equalizing of the bearing pressure stresses
among the teeth, and (3) the elimination of the need for grinding due to the
smooth cutting action of a gear hob. The nomenclature associated with
involute splines is shown in Figure 7-15.
'” Pressure angles of 45, 20, 25, and 14} deg are also sometimes used. However both the SAE
and ANSI specify only the 30 deg pressure angle in their standards.
 375
Section 7-14: Splines

Splines are made having a flat root or a fillet root (see Figure 7-15) and
are available in 14 diametral pitches designated as fractions, namely 2,
3, B> 10> D> Te de> da, 4S, 20, 24, 32, 4 and $2. Within each fractional
designation, splines can be cut having 6 to 50 teeth. The numerator of these
fractions is the diametrical pitch P (see Chapter 10), and the denominator
(always double the numerator) controls the depth of the tooth.
There are two types of spline fits'® (see Figure 7-16):

ene ! tooth Major diameter

Major diameter chamfer dimen.
et radius
fillet radiu GLE a Major diameter
Major diameter we AY @ ;
UPAR] Us S height
ile height ‘
We
ret nt. CAILNY:
4%

y Ge A A, YG
$ ZY KS \\
Minor diameter
Minor swne\ $0 wD Pg A\< YL fillet radius
chamfer height Ze IN
a ee Major SS
diameter NN S
Minor diameter Minor Minor Minor diameter
chamfer angle dha: Pitch diameter fillet height
Te diameter TIF
diameter diameter

a = addendum d, = minor diameter

b = dedendum-flat root N, = number of teeth
b, = dedendum-fillet root p= circular pitch
c = clearance P= diametral pitch
d = pitch diameter t = circular tooth thickness
dp = base circle diameter t, = circular space width
do = major diameter ¢ = pressure angle
TIF = diameter at junction of involute form with fillet
Dimensional relations;
d=N,/P,p =n /P, t= 1/2P, a = 0.500/P

= Diameter Pitch = 5through = s through 96

48

Major dia. (internal) | d,; = (NV, + 1.8)=P | do; = (N, + 1.8) =P

Minor dia. (external) | d,. = (NV, —1.8)="P | d-, = (N, — 2.0) =P
Dedendum (internal) |6, = 0.900 =P b, = 0.900 =P
Dedendum (external) |6, = 0.900 + P b, = 1.000 = P

Figure 7-15 Nomenclature and some dimensional relations for the involute
spline profile. Note that the vertical line separates the internal spline form from the
external spline form. The dotted curve is the outline of the optional fillet root form.
(From H. L. Horton (ed.): Machinery’s Handbook, 15th ed. The Industrial Press,
New York, 1957.]

18 Prior to 1959, the minor diameter was also used for tolerance control, but this basis of control
has since been deleted from the standards.
 376
Chapter 7: Shafts, Keys and Couplings

Figure 7-16 (a) Major diameter fitted spline.

(b) Side fitted spline.

1. Majer diameter fitted splines where the fit is controlled by variation of

the major diameter of the external splines.
2. Side fitted splines where the fit is controlled by varying the tooth
thickness. It is customary to use this type of fit for fillet root type
splines. However, it also can be used for the flat root type spline.
The two types offits are divided into two classes:
Class 1. A loose fit where there is clearance at all points.
Class 2. A close fit, which is close on either the major diameter or the
sides of the tooth.
According to reference [14], “for carrying of any load or to obtain stability,
there is seldom need to have a spline longer than its pitch diameter.’ The
spline length (L), required to transmit a torque (T) is obtained as follows.
We found in Chapter 5 that the torque carrying capacity of a shaft (based
on shearing stress) was

wes mD?(1 as D?/D?.)Ss

fe (7=31)22
= 16
where the terms are those defined in Figure 7-15 except for D;, which is
the internal diameter of a hollow shaft. If we assume that all the teeth are
in uniform contact,”° the torque carrying capacity of the shaft (based on the
compressive strength of the teeth) is

ip_S.AND
:
_ S,LhND
; . :
(7-32)
where

A = area of contact of one tooth = Lh

h = height of tooth contact = 0.8/P = 0.8D/N, (from reference [14])
Replacing h in equation (7-32), we get
T = 0.4S,LD? (733)
*° Again we note that this equation can also be used for a solid shaft, namely, D, = 0.
*° This assumption is not strictly true, since manufacturing accuracy plays a decisive role in
determining spacing errors. Furthermore, shaft alignment also affects tooth contact and tooth
load distribution.
 /,
Section 7-14: Splines

The shear area at the pitch diameter ofthe splines is DL/2 and so the shearing
stress 1S

S, = (7-34)

Solving equation (7-34) for T and equating it to equation (7-33), we find

that S,/S,= 0.5093. Therefore, we conclude that the splines are more
bens stressed in1 shear than in SE (that is, bearing). Thus, shearing

If we nowape hat ‘heaera of the pling teeth in shear should be

the same as the shaft strength in shear, we can equate the shearing stress of
equation (7-34) with that of equation (7-31) to obtain

[bee
De (Wes D7) -35
4D? Ve)

For a solid shaft, equation (7-35) becomes

us oe

However, according to the SAE, “‘actual practice has shown that due to
inaccuracies in spacing and tooth form, the equivalent of about 25% of
teeth are in contact, so that a good approximate formula for a splined shaft
is’

Di an D?/D?.)
L,= D? (7-37)

where L, = effective spline length, and D; = 0 for a solid shaft.

Involute serrations are involute splines that have a 45 deg pressure angle
and are used mainly for close fits not subjected to sliding. However, it is
possible to use other classes of fits. The nomenclature used for splines
applies gels to involute serrations. Involute pitches listed in the SAE
Handbook are 48, 48, 24, 32, 49, ¢8 8%, 85, and 428, and complete tables
for involute serration specifications are provided.
Involute serrations have teeth that are shallower than involute splines
and also have broader bases, less depth of contact, and quite frequently
provide manufacturing advantages. Compared to involute splines, involute
serrations have greater contact pressures, sliding resistance, and radial
forces for the same loads. By using finer pitches, a wider range of index
positions is possible by virtue of the greater number of teeth. For more
detailed information, the reader is again referred to the SAE Handbook.
 378
Chapter 7: Shafts, Keys and Couplings

SECTION 7-15
Couplings
Couplings are used to connect two shafts. For example, a coupling is
used to join the shaft of an electric motor to the line shaft of a machine or
a hydraulic turbine to an electric generator or, for practical reasons, is
used to sectionalize a long shaft, and so on. Couplings used for such typical
applications are called permanent couplings because their connections would
only be broken for repairs and/or general maintenance. Those applications
which require the shafts to periodically disengage are called clutches, and
are discussed in Chapter 13.
Permanent couplings are classified into two groups (1) rigid couplings
and (2)flexiblecouplings.
Rigid couplings
Rigid couplings are used for shafts having good collinear alignment and
must be installed with care and forethought. Connecting misaligned shafts
with a rigid coupling can lead to bearing failure, fatigued shafts, worn
flanges, or broken flange bolts. When used on line shafting, support bearings
should be located near the coupling and checked for both static and dynamic
balance. Although simple in design, rigid couplings are generally restricted
to relatively low speed applications where good shaft alignment or shaft
flexibility can be expected.
There are three important types of rigid couplings. These are the compres-
sion, ribbed, and flange type couplings (see Figure 7-17). Figure 7-17a shows
a compression coupling (also called the Seller’s coupling) that has two split
cones keyed to the shafts. By tightening the bolts, the cones are forced to
‘squeeze’ the shaft by the wedging action of the double conical shell.
These couplings are used on shafts from 1;4—8;% in. diameter. The inspection
hole in the conical shell permits an observer to see when the cones are drawn
together.
Figure 7-17b shows a flanged compression coupling in which the split
double cone does not move axially but is squeezed against the shaft by the
wedging action of flanges drawn together by the bolts. This type of coupling
transmits torque only by the frictional force between the shaft and the split
double cone eliminating, thereby, the need for a key and keyway. The
flange dimensions D, and L are approximately the same as those for flanged
couplings (see Figure 7-17b).
Compression couplings are available for shafts up to 3-in. diameters, and
can be made to order for nonstandard or oversized shafts.
Figure 7-17c is made in two axial parts clamped together by means of
bolts as shown. A square key is used-to maintain shaft alignment as well as
to locate one half of the coupling. Torque is transmitted mainly by friction
(due to the clamping action of the bolts) and partially by the key. This
 -9U09 SIO][IS ©) ‘Surjdnoo pis jo sodky,
‘Burdno
ZiL-Z eanBig
posure uorssaidwi0o (2) o paqary
!
eae asta “Sul[dn
(q) oo
== ‘Suljdn
(P) oo asuey_{ ‘surdnoo Asain0D] yaq-urT
| KK|K
[uoneiodiog
ZZZLEIL lg
= Peso[u9Ad ‘UOISIAIOWA
G
CA
TE LL
LE E ZZA
2
\

g eae a
=
go1dn hz ut ‘jayawelp
is
"a = dvt G10} —p<
9L p ase
s}jOg “pasn JO4
; ,
yur “ul
a>Fz sles ‘yayawelp
= 8
sede,
aie
s1joqg “pasn
(P)
(q)

(9)

=
ma

\ZZ
aes
uN
Yili
SZ
EW

379
(P)
(a
 380
Chapter 7: Shafts, Keys and Couplings

type of coupling is intended for heavy duty service. Standard units are
available for shafts up to 7-in. diameters and can transmit the torque based
on AISI C1018 shaft strength.
The most widely used of the rigid couplings is the flange coupling (Figure
7-17d), which is capable of transmitting large torques. These couplings,
although larger in diameter than the two types previously described, are
shorter and about 20-25 % lighter. Although flange couplings have not been
standardized,?! they are commercially available to fit shafts up to 8-in.
diameters, and are rated at a torque capacity equivalent to AISI C1018
shafting.
As can be seen in Figure 7-17d, flange alignment (called register) is obtained
in one of two ways. One way is to permit one shaft to act as a pilot and enter
the mating flange to about 3 in. The second way (not shown in Figure 7-17d)
is to machine a shallow projection (that is, rabbet) on one flange face and fit
it to a female recess cut in the face of the other flange. Alignment is obtained
by carefully machining the diameter of male projection and female recess.
Best results from this kind of coupling are obtained when the flanges are
shrunk on the shaft and then finish machined with the shafts straightened
afterwards. For greater reliability and strength (for example, marine and
automotive drive shafts), flanges are often forged integrally with the shafting.
When flanges are force fitted onto shafts, square keys are used, but tapered
keys are used when flanges are placed on the shafts with a slip fit.
Torque is assumed to be transmitted in one of two ways, (1) by means of
carefully fitted (that is, ground) bolts in holes that are line reamed at assembly
or (2) by the frictional force produced between the two flange faces when
bolted together. The first method, requiring selective assembly, needs a dowel
pin and hole (or some other marker) on each half of the flange. This will
guarantee the realignment of the same bolt holes if the flanges have to be
disconnected and then reassembled. With virtually no clearance between
the bolt holes and bolts, the bolts are subjected to shearing stress.?? By
assuming that the torque is transmitted by virtue ofthe friction force between
the flange faces, there is no need for ground bolts and line reamed bolt holes.
In fact, enough clearance can be allowed between bolts and holes to assure
that the flange can be assembled in any angular orientation of the bolt
holes. Therefore, the bolts are only subjected to a tensile stress. Transmittal
of torque by friction is not recommended where loads are unsteady or
reversible.
The flange surface that seats the bolt heads and nuts is machined or spot
faced perpendicularly to the bolt hole axis, assuring that the bolts are not
*! An exception is ANSI B49.1-1969 “Shaft couplings, integrally forged flange type for hydro-
electric units.” These flanges are standard for shafts ranging from 34—48-in. diameters.
*? In the strictest sense, this is not completely true. It is more likely that the bolts would be
subjected to the combined stress of bending and shear. The bending component would be
caused by any relative displacement of the flange halves up to the minimal clearance between a
bolt and hole.
 38]
Section 7-15: Couplings

deliberately subjected to bending when tightened. As can be seen from

Figure 7-17d, the heads of the bolts and nuts are covered by projecting
flanges as a safety measure to protect workmen. Because of these projections,
flange couplings are also known as “‘safety flange couplings.”
As “a rule of thumb” guide, the length of hub, diameter of hub, and
flange diameter can be based on the relations in Figure 7-17d. It is also
common practice to use a number of bolts in accordance with empirical
equation (7-38)
P= 05Dee 3 (7-38)
where n is the number of bolts and D is the shaft diameter. Dimensions of
safety flange couplings for shaft diameters from 1-12 in. are listed in
Machinery’s Handbook [15]. Before specifying such flanges, the designer
would be well advised to check the coupling strength.
If a flange coupling is to be designed (or specified) using ground and
fitted flange bolts, the designer should check the strength of the following
areas:
1. Direct shear failure of the bolts. The bolts are assumed to carry their
fractional part of the load which, when divided by the cross-sectional
area, should result in a shear stress below that of the material strength
divided by the factor of safety. Due account should also be taken of
any service factors based on industry experience.?* A particular
approach is to make the bolts as strong in direct shear as is the shaft
in torsion. Note that, if in the designer’s judgement “‘excessive’’ bolt
clearance would introduce a bending stress, then the bolts must be
designed on the basis of combined stress (that is, transverse shear and
bending).
2. Bearing ofthe bolts. Bearing of the projected area of the bolt in contact
with the side ofthe flange hole. For steel, the allowable stress in crushing
can be assumed as twice the yield value in tension.
3. Shearing of the flange at the hub. The shear area resisting failure is the
hub circumference times the web thickness of the flange.
4. Shearing or crushing of the key. The calculations for these possible
modes offailure are described for a square or tapered key in Section 7.13
of this chapter.
In contrast to fitted bolts, a flange coupling designed on the basis of
friction-torque capacity requires a somewhat different analysis than that
described above. To assure no slip, we must have
LEN b

where T, is the frictional torque and T is the applied torque. If Fo is the

bolt tension, p the coefficient of friction between the flange faces, D, the

?3 See manufacturers’ catalogs.

 382
Chapter 7: Shafts, Keys and Couplings

outside diameter of the flange friction face, D; the inside diameter of the
flange friction face, and n the number of bolts, T; is

‘Na!
T, = Fyr = pF n(? |527 (7-39)

where F, is the frictional force and r is the mean radius at which F, is applied.
If we Aon determine the torque-capacity of the shaft, we may derive an
equation determining the bolt load F, required. Thus for a yield in shear
of 0.5S,,, for steel (that is, using the maximum shear theory) and a factor
of safety N, we have
0.5S, TD SM b
S — yP — 2 =

é N J nD3(1 — D}/D*)
or 7

i——
nS, ,D3(1 — D#/D4)
64N
Thus, relation (7-39) becomes

DD; ~ TS ypDo( — D}/D*)

F
enue > 64N
Solving for Fo,
mS,,D?(1 — D?/D*)
yee
(7-40)
o* 16un(D, + D)N
where D; = 0 for a solid shaft. From relation (7-40), we can calculate the
minimum bolt force required to transmit a torque by the friction force
between the flange faces.
In a similar fashion, we can obtain the bolt force required to transmit a
torque by friction based on the shearing of a key by means of equations
(7-21) and (7-39) or that which is based on the bearing of a key by means of
equations (7-23) and (7-39). The resulting relations are respectively

S,,WLD
Fy > —2—__ 7-41
o* pn(D, + D,)N oD)
S.WLD,
lg = (7-42)
un(D, + D)N
The required bolt area is determined by the largest Fy found by equations
(7-40), (7-41), and (7-42) and then dividing this load by the tensile yield
strength of the bolt material specified or selected.
Flexible couplings
Flexible couplings are used to connect shafts subject to one or more
kinds of misalignment (see Figure 7-18) and to reduce the effect of shock
and impact loads that could be transferred between shafts. Flexible couplings
 383
Section 7-15: Couplings

are broadly classified into two groups, (1) couplings with kinematic flexi-
bility that employ rigid parts, and (2) couplings that have resilient parts.
Flexible couplings employing rigid parts are designed for specific shaft
misalignments and so do not possess constraints in the misaligned directions.
Such couplings transmit torque without backlash or angular play (that is,
no torsional flexibility) other than that due to manufacturing tolerances and
wear. As a consequence of this type of construction, couplings with kine-
matic flexibility (that is, lacking torsional flexibility) are incapable of dampen-
ing the transmittal of shock and impact loads. Figures 7-19, 7-20, 7-21, 7-22,
and 7-23 are common types of flexible couplings having rigid members.
Flexible couplings that contain resilient components are capable of
accommodating shaft misalignments as well as shock and impact loads.
Because these types of couplings possess torsional flexibility, they also act
as ‘“‘detuning” devices, thereby altering the vibration properties of the
connected system. Figures 7-24, 7-25, and 7-26 are photographs of different
types of flexible couplings that contain metallic resilient components.
Another large group of flexible couplings contain nonmetallic flexible
members. In this group, there are two types of designs, (1) where the non-
metallic flexible material is subjected to compression by being held between
projections from the coupling flanges and, (2) where the nonmetallic flexible

Alignment

Parallel - Offset Misalignment

Bese P Figure 7-18 Types of shaft

{ misalignment.

Symmetrical Angular Misalignment

A>B

pelle

cape
Non-Symmetrical Angular Misalignment
 Figure 7-19 Double slider coupling
(Oldham’s coupling). [Courtesy Link-
Belt Enclosed Drive Division, FMC
Corporation. ]

Figure 7-20 Slider block coupling.

[Courtesy Zurn Industries Inc., :
Mechanical Drives Division.]

Figure 7-21 Double roller chain

coupling. [Courtesy Link-Belt Chain
Division, FMC Corporation. ]

Figure 7-22 Gear coupling. [Courtesy Zurn Industries, Inc., Mechanical Drives
Division. }

384
 385
Section 7-15: Couplings

Figure 7-23 Flexible disc

coupling. [Courtesy of Thomas
Coupling Division, Rex Chainbelt
Corp.]

material is fastened (or bonded) to the coupling flanges and thereby subjected
to shear. The latter type of design results in a coupling having a “‘softer”’
torsional spring constant than a compression type design. This important
difference is attributable to the fact that the nonmetallic flexible member
(usually an elastomer) is not confined in the “‘shear type” design, whereas

Figure 7-24 Franke pin coupling.

[Courtesy Waldron Coupling Division,
Midland-Ross Corp.]
 356
Chapter 7: Shafts, Keys and Couplings

Figure 7-25 Laminated metal radial spoke coupling. [Courtesy Brown

Engineering Company. |

the nonmetallic material is squeezed between the flange projections in the

“compression type’’ design.
Figures 7-27, 7-28, and 7-29 are illustrations of the compression type,
nonmetallic flexible member couplings. Figures 7-30, 7-31, 7-32, and 7-33
are illustrations of the shear type, nonmetallic flexible member couplings.

Figure 7-26 Steelflex coupling.

[Courtesy of Folk Corporation,
Subsidiary Sunstrand Corporation. ]
 387
Section 7-15: Couplings

Figure 7-27 Ajax rubber-cushioned sleeve bearing coupling. [Courtesy Ajax

Flexible Coupling Co. Inc.]

It is obvious from the many flexible couplings shown, that there is a

coupling that will satisfy almost any design requirement. The designer will
find that each manufacturer’s catalog contains the specialized information
required to size a proper unit for any particular coupling. Although the
designer would normally not be confronted with the detailed design of a
coupling, he should, on such occasions, work closely with the manufacturer.

Figure 7-28 Sure-Flex coupling.

[Courtesy of Gerbing Manufacturing Co.]
 Figure 7-29 Rubber insert coupling.
[Courtesy of Metal Products Division,
Koppers Company, Inc.]

Figure 7-30 Bonded rubber disc couplings. [Courtesy Lord Kinematics, Lord
Corporation, Erie, Pa.]

Figure 7-31 Para-flex coupling. [Courtesy Dodge Manufacturing Division,

Reliance Electric Company. ]

388
 Figure 7-32 Airflex coupling.
[Courtesy the Falk Corporation,
subsidiary Sunstrand Corporation. ]

Figure 7-33 Morflex coupling. [Courtesy Morse Chain,

Division of Borg-Warner Corp.]

389
 390
Chapter 7: Shafts, Keys and Couplings

SECTION 7-16

Universal Joints
A universal joint (also known as Hooke’s joint or Cardan’s joint) is a
kinematic linkage used to connect two shafts that have a permanent angular
misalignment.?* As illustrated in Figure 7-34, a universal joint consists of
a driving yoke 1, which can be connected to an input drive shaft, a cross
link C, and a driven yoke 2, which can be connected to an output shaft.
Note that both yokes are free to pivot about pins AB and MN of the cross
link C.
e
velocity ratio, (ie., if the input shaft has a constant angular velocity «, the
output shaft will have a non-constant angular velocity w,). We can demon-
strate this fact by referring to Figure 7-34 where illustration (b) is the top
view of illustration (a) and illustration (c) is a view of illustration (b) after
it has been rotated by 90 deg. From Figure 7-34b we find that the velocity
of point M is v,, = w,rcos 6, perpendicular to pin MN which is in the
plane of yoke 2. Thus, we obtain the angular velocity of yoke 2 as

Ge 0 jf) G1 COS OF — On icCOse

We now can write the velocity ratio as
Oy, On 1
= = 7-43
WM, @,cosd cosd |

Now looking at Figure 7-34c, we see that a rotation of 90 deg has caused
pins AB and MN (perpendicular to the page) to assume the positions shown.
From Figure 7-34c we obtain v, = rm, which 1s represented by the velocity

Figure 7-34 Basic components of Hooke’s joint. When | has turned 0 = 90 deg
(and 180, 270, 360 deg), the driven member 2 has turned ¢ = 90 deg (and 180, 270,
360 deg); but at other values of 0, 0 and ¢ are not the same.

** The universal joint can be used in place of couplings where the only misalignment between
shafts is angular. It cannot be used to compensate for offset shaft alignment or axial play, although
the latter condition can be accommodated by use of telescoping shafts.
 ov,
Section 7-16: Universal Joints

vector @. Since the pin joint at N cannot have two separate velocities, velocity
vector € must be a component of velocity vector v,,. Hence, we have
rO,
Vv. =

Ey ECOS.O
and obtain

Vien rO, Wy
Oy) = =

The velocity ratio of this 90 deg rotated position then is

2
@, @,/cosd
SEE 5 (7-44)
We can readily see by comparing equations (7-44) and (7-45) that the speed
of the driven shaft 2 varies from a minimum of w, cos 6 to a maximum of
@,/cos 6 for a 90 deg rotation. From 90 deg to 180 deg, the driven shaft
speed becomes a minimum. This cycle repeats itself every 180 deg. Now,
from spherical trigonometry, it can be shown?> that
tan @ = tan @cos 6 (7-45)
Next, we differentiate equation (7-45) with respect to time t for some fixed
angle 6 between shafts.

do
sec? of = S607 a cos 6

Or

w, cos d/cos? 8 ~—_@, cos 6/cos* 6

oO, = =
1+tan?@ 1+ tan? @cos?6
- @, cos 6/cos? 0 - , Cos 6
~~ cos? 6 sin? @ (cos? 0 + sin? 0 cos? 6)
—— + cos” 6] —;
sin’ 6 cos* 8

But sin? 6 = 1 — cos? 0. We finally obtain a relation for the output shaft
velocity w, as a function ofthe input velocity w, the angle between the shafts
6 and the angular displacement of the input shaft 6.
@, COs 0
> (7-46)
1 — sin? 6 sin? 0

Differentiating , with respect to t produces the equation for the acceleration

5 For example see Mechanics of Machinery by C. W. Ham and E. J. Crane, McGraw-Hill Book
Company, New York, 1948; or Kinematics of Machines by R. T. Hinkle, Prentice-Hall, Inc.,
Englewood Cliffs, N.J., 1964.
 392
Chapter 7: Shafts, Keys and Couplings

= io)

Angles between driving

-_----—
and driven shaft , deg.

——_——

shaft,
driven
of
deg.
log
and OOD
Lead Soe
ft
Oo

10 [ee esed ee od Be es eed eT

0 10 20 30 40 50 60 70 80 90 100110 120130140150160170
180

Rotation of driving shaft,deg.

Figure 7-35 Plot of how the speed of the output shaft

leads and lags the speed of the driver shaft 6 = 20, 30, and
40 deg.

of shaft 2 (note that m, and 6 are constant).

dw, _ w{ sin* 6 cos sin 20
(7-47)
dt (1 — sin? 6 sin? 6)
0, =

We can readily ascertain from equation (7-46) that w, varies with the
angular position @ of shaft 1 (for a constant w, and 6). We also observe from
this same equation that the magnitude of the variation of @, is larger, with
larger angles 6 between shafts. This variation is vividly illustrated by Figure
7-35, which is a plot of equation 7-46 showing how the driven shaft speeds up
and slows down for each revolution of the driver shaft 1.
This phenomenon can further be verified by equation (7-47), which shows
that the angular acceleration of the driven shaft increases with increasing
angles 6. As a result of these output velocity variations, severe vibrations may
be introduced into the driven machinery or equipment.

Ithough well constructed “joints”

can accommodate angular shaft misalignments as large as 40 deg. Also,
depending on the quality of the universal joint, the designer should plan to
use these connections for low or moderate shaft speeds and power.
By using a double universal joint (Figure 7-36b) or two single universal
joints connected by a short shaft (Figure 7-37) it is possible to connect two
shafts having considerable angular misalignment. This arrangement of
universal joints can also provide a constant velocity ratio between input and
output shafts. To achieve this desirable condition when both the driver and
driven shaft are in the same plane, it is necessary for the yokes of the input
and output shafts to be in the same plane and for the angles between the
driver shaft, the driven shaft, and the connecting shaft to be equal. Figures
7-38 and 7-39 illustrate two arrangements for achieving this constant speed
ratio arrangement.
Figure 7-36 Universal joints. (a) Single
joint and (b) double joint. [Courtesy
Dana Corporation, Spicer Universal
Joint Division.]

(b)

Figure 7-37 Two universal joints joined

by a short shaft. [Courtesy Lovejoy Inc.]

393
 394
Chapter 7: Shafts, Keys and Couplings

Figure 7-38 Two arrangements of a double universal.

joint system for achieving a constant velocity ratio.

Sometimes, design circumstances require that the double universal shaft

assembly connect input and output shafts that cannot be in the same plane.
In such cases, it is still possible to obtain an arrangement that will result in a
constant velocity ratio, but it is imperative that (1) the yoke at each end of
the connecting shaft be so oriented that the planes containing them be at the
same angle y formed by the planes of the driving and driven shafts (that is,
the connecting shaft yokes are perpendicular to planes A and B as shown in
Figure 7-39), (2) that the angles formed at joints of the connecting shaft with
the input shaft and the output shaft be equal (that is, 8, = 6, in Figure 7-39).
Because universal joints do not provide for axial play, provision for this
requirement can be made in the connecting shaft or at the connection between
the driver or driven shaft.

Plane of shafts
2 and 3

Y oke right angle

to plane B

Y oke right angle

to planeA

Figure 7-39 Correct angular arrangement for

achieving a constant velocity ratio when three shafts are
not in the same plane. Note: For this arrangement to
function as desired, 8, must be equal to B,.
 39D
Section 7-17: Flexible Shafting

Several types of universal joints have been developed by various manu-

facturers because of their wide application (for example, in automotive,
agricultural, aircraft, industrial, and instrument applications) and because
of the need for one that has a constant velocity ratio as well as for one that is
able to operate satisfactorily under high speeds.

SECTION 7-17
Flexible Shafting
Where power is to be transmitted between two points located in such a
manner or configuration that a connection by shafting is not simple or
feasible, and combinations ofbelts, chains, and/or gears would be complicated
or costly, the designer might consider the use of a flexible shaft. There are two
principal kinds of flexible shafting, (1) the shaft designed to transmit power
in one rotational direction and (2) the shaft designed primarily to transmit
motion for remote control applications that can be worked rotationally in
both directions. Typical power transmission applications are business
machines, food handling machines, multiple spindle drills, portable tools,
tachometers, indicator mechanisms, and speedometers. Typical remote
control applications are power operated automobile seats, automobile
windows and convertible tops, aircraft controls, medical and dental
equipment, and so on.
Flexible shafts are constructed of 1 to 12 layers of wire which are helically
wound on a mandrel. Each layer is wound in an opposite direction (that 1s,
the same as a right hand or left hand screw), and the top layer determines
whether the cable is either a “right lay”’ or a “‘left lay”’ cable. The shaft is then
encased in a sheath and end fittings are attached. Wires used to fabricate
flexible shafts can be wrought steel, stainless steel, phosphor bronze, Monel,
and Inconel. For high temperature application, stainless steel is used. Monel
and phosphor bronze are used respectively for applications requiring
corrosion resistance and nonmagnetic properties.
Shaft diameters ranging from a 0.043-in. to 2-in. diameter are available
and are rated in accordance with the dynamic torque capacity in the winding
direction corresponding to the radius of curvature. The larger the shaft
diameter, the more pronounced is the reduction of the torque capacity with
the smaller radii. For example, a 0.050-in. diameter shaft which has a torque
capacity of 0.26 lb-in. for a 10-in. radius of curvature decreases to 0.16 Ib-in.
for a 4-in. radius of curvature—a reduction of 38.5%. On the other hand, a
0.187-in. diameter shaft has a torque capacity of 11.0 lb-in. for a 10-in. radius
of curvature and a 4.0 lb-in. torque capacity for a 4-in. radius of curvature—a
63.6 % reduction of torque capacity.
Flexible shafts for power transmission should have a left-hand lay for
clockwise rotation and a right-hand lay for counterclockwise rotation.
These directions are given as viewed from the power source. Figure 7-40
shows a typical shaft construction and assembly.
 396
Chapter 7: Shafts, Keys and Couplings

Casing fitting
Thrust washer,
Coupling nut Flexible shaft

SSSSNS
SSS Pete od eee Sisse
>>> >)

—— SMT im af Ye
eae qegie
===
la 0 a ooo ace na ae

Housing of Flexible _ fitting

casing
driven member Intergral-formed Coupling nut
drive square
driving member
Integral-formed
drive square Spindle of
driving member

Figure 7-40 Typical flexible shaft construction showing connections. [Courtesy

S. S. White Industrial Products, Pennwalt Corp.]

As with couplings, manufacturers’ catalogs?° describe the proper procedure

and considerations required to select an appropriate flexible shaft.

PROBLEMS
1. A solid transmission shaft made of AISI Type 316 cold-drawn stainless steel is used
to transmit 95 hp to a series of chemical mixing vats at 100 rpm. Since the atmo-
sphere may be corrosive or at best unknown, assume a factor of safety of 4 (see
Section 1.3). What is the shaft diameter?
2. Assume you wish to replace the solid shaft of Problem 1 with a hollow shaft having
an inside diameter equal to the diameter of the solid shaft. What is the wall thickness
of the hollow shaft? What is the per cent difference in weight per foot of shaft length
between the hollow and solid shafts?
3. It has been decided because of atmospheric corrosion to replace a AISI 1030
cold-drawn steel shaft with an aluminum alloy 2017-T4 shaft. What diameter
aluminum shaft must be used? If the aluminum shaft is to have the same angular
stiffness as the steel shaft, what is the weight ratio of the shafts?
4. A horizontal power transmission shaft is supported by two bearings 40 in. apart.
The shaft overhangs the right bearing by 12 in. and supports a 30-in. pitch diameter
straight-toothed spur gear. On the left side there is a shaft overhang of 10 in.

*6 For example, F. W. Stewart Corp., S. S. White Ind. Products, Elliot Mfg. Co., Stow Mfg. Co.
 397
Problems

supporting a 50-in. diameter flat belt pulley. The respective weights of the gear and
pulley are 200 and 600 lb. The gear is driven by a pinion in a rotational direction
such that the tangential turning force on the gear tooth acts upwards (ignore the
radial force tending to separate the gears in this problem). The pulley that delivers
the power vertically downward has a belt tension ratio (that is, tight side to loose
side) of 2.5. The shaft must transmit 25 hp at 150 rpm. Assume 100% transmission
efficiency (that is, no belt slips and no losses at the gears). (a) Draw the loading,
shear, and moment diagrams. (b) Determine the shaft diameter by the maximum
shear failure theory. The material is AISI 1040 annealed steel and both the gear and
pulley are held on the shaft by a sled-runner key. (c) What would the shaft diameter
be if it were subjected to a definite minor shock caused by a single revolution clutch
located on the pinion shaft (that is, before the gear input)?

. Schematic Problem Figure 5 shows a straight tooth, involute gear mounted midway
between ball bearing supports with an overhanging flat belt pulley. Power input
to the pulley is 15 hp at 1000 rpm under steady load. The pulley weighs 80 Ib and
the gear weighs 401lb. Determine the shaft diameter (that is, configuration) if
lengths AB and CD are to be the same diameter and length BC is 50 % larger. The
radius of the fillet joining sections AB and BC is 0.0625 in. The radius of the fillet
joining sections CD and BC (that is, where the bearing is shouldered) is 0.010 in.
The gear is keyed to the shaft by a square profiled key and the pulley by a sled-
runner key. The shaft material is AISI 1020 cold-drawn steel. Assume zero pulley
slip. Base your design on the maximum shear theory of failure. See equations
(10-12) and (10-13) for gear forces.

pitch dia. 9°4F

essure angle
8 in.
eee ;

Problem Figure 5

. Problem Figure 6 shows a fixed gear train in which the shaft is machined from
AISI 2330 cold-drawn steel. Power input to the system is at gear A, and the output
is at gear D. Gear C is mounted on the outside diameter ofa single revolution clutch
that transfers power to gear D. This power changes from 20 hp to 120 hp and back
to 20 hp for each revolution of the shaft. Gear B is keyed to the shaft by a sled-
runner key and the clutch by a profiled key. If the shaft is not stepped and a safety
factor of 1.5 is assumed, determine the shaft diameter. In this problem, the radial
forces tending to separate the gears are to be ignored and only the tangential
turning force components are considered. Assume also that minor shock is contrib-
uted by the engaging clutch for each revolution of the shaft. Ignore the weights of
the gears.
. Solve Problem 6 taking into account the forces tending to separate the gears. The
gears have a 20 deg pressure angle.
 398
Chapter 7: Shafts, Keys and Couplings

g in. pitch 31:

Problem Figure 6

8. Problem Figure 8 shows a schematic for a shaft of a punch press which mounts
a flywheel, a flat belt pulley, and a gear D. The pulley weighing 120 lb is driven by a
belt located at 45 deg with the horizontal, as shown. The input to the pulley is 30 hp
and it rotates at 300 rpm. A straight toothed involute spur gear located at D and
having a 20 deg pressure angle drives a gear F. This gear, in turn, drives a crank and
connecting rod that operates the punch. The weight of the gear is 30 lb. As in all
mechanical presses, the stored energy of a flywheel is used during the punching
operation. In this case, a flywheel weighing 1000 Ib and having a radius of gyration
of 20 in. is located at A. During the punching part of the cycle, it is assumed that
the power delivered to gear D is doubled—namely, 30 hp is obtained from the
flywheel and 30 hp is obtained from the belt. The timing of the cycle is such that
the location on the shaft at which the moment is a maximum experiences an
alternating tensile and compressive stress. The flywheel is locked to the shaft by
a sled-runner keyway, whereas the pulley and gear are both held by a profiled
keyway. Using a safety factor of 2.5 and AISI 3140 cold-drawn alloy steel, determine
the shaft diameter by the distortion energy failure theory accounting for the impact
load due to punching.

eas

Connecting rod
to punch

10 in. pitch dia-

419.4
10;ne==

Problem Figure 8
 399
Problems

9. The worm drive shown in Problem Figure 9 supplies 60 hp steadily at 1600 rpm
and is driven by a 10-in. pitch diameter 20 deg pressure angle gear. The pitch
diameter of the worm is 3 in. and it is desired to make the worm an integral part of
the shaft if possible. The figure shows the forces acting on the worm and input gear
where the thrust load is taken by the bearing at B. The shaft is to be made of
AISI 1040 hot-rolled steel. The fatigue stress concentration factor for the worm
threads can be taken as K, = 1.5 for both bending and shear. Determine the follow-
ing: (a) the minimum root diameter of the worm thread by the energy distortion
theory of failure (assume a safety factor of 2.0); (b) the shaft diameter D at the change
of section where the shaft joins the worm ; (c) the safety of the shaft diameter found
in part (b) at the gear A.

fz
i“ gied-runne \ ! NY r

fa as Problem Figure 9
6 in:

6 in-
ene
10. Problem Figure 10 shows a shaft that is a component of a parts feeder for a pro-
duction machine. The shaft is supported by a split sleeve bearing at B and a regular
sleeve bearing at D and mounts two 144-deg pressure angle straight tooth involute
spur gears at A and C. Gear C is keyed to the shaft by a profiled keyway, and gear A
is held by a sled-runner keyway. Input to gear A is 10 hp at 300 rpm. Output is at
gear C. The shaft material is made of AISI 1340 quenched in oil and tempered at
1200°F. (a) Determine the factor of safety N at the }-in. fillet radius. (b) Determine
the factor of safety of that portion of the shaft which holds gear C. Use the maximum
distortion theory of failure in determining both (a) and (b).

2.000 in. pitch dia. 1.500 in. dia.

(Typical)

1 2 .
—8!in, fillet radius

bearing
Center line of

Center line
of bearing

Problem Figure 10
 400
Chapter 7: Shafts, Keys and Couplings

11. The worm shown in Problem Figure 11 is made of AISI 8740 alloy steel quenched
in oil and tempered at 1000°F. An input torque of 60,000 in./Ib is applied to the
right end of the shaft. The pitch diameter of the worm is 6.923 in. and is the diameter
at which it is assumed the forces R, E, and T are acting. The root diameter is 5.701 in.
The force opposing rotation T (shown as coming out of the paper in a perpendicular
direction) is 6200 Ib. R is 1950 lb and is the force that tends to cause the worm and
worm gear to separate. E is 6600 lb and is the reactive thrust force acting on the
worm. The fatigue stress concentration factor at the root of the worm thread can
be taken as 1.5 for shear and bending. Using the distortion energy theory of failure,
determine the factor of safety N at the critical points of the shaft.

4169 in. 1.65

in. 1.615 iin.

Center line of the worm

Probiem Figure 11

12. The input to pulley A in Problem Figure 12 varies from 20 to 30 hp and back to
20 hp for each revolution at a speed of 200 rpm. Pulley A weighs 150 lb and pulley
C weighs 300 Ib. Both are keyed to the shaft by a profiled key. The shaft is machined
from AISI 1020 hot-rolled steel. Determine the following: (a) The shaft diameter
required for a factor of safety of 2.5. (b) The slope of the shaft at bearing D. (c) The
maximum deflection of the shaft between bearings B and D. (d) The shaft deflection
at pulley A. (e) The lowest critical speed of the shaft.

Problem Figure 12
 401
Problems

13. Power is delivered to the pulley and taken out at the gear. The forces acting on each
of the elements are shown on the Problem Figure 13. The pulley weighs 350 Ib and
the gear weighs 150 lb. If the shaft material is AISI 1030 normalized steel, determine
the required shaft diameter by the maximum shear theory of failure. The load is
steady and both the gear and pulley are fixed to the shaft by a profiled key.
500 Ib

26 in. dia. pulley

1500 Ib
\ lo
56

Problem Figure 13

14. In Problem Figure 14, a 20-in. diameter flat belt pulley receives 25 hp at 500 rpm.
This power is transferred to a straight toothed 20-deg pressure angle spur gear with
a pitch diameter of 10 in. The pulley weighs 250 lb and the gear 150 Ib. If the loads
are steady, determine the shaft diameter required based on the maximum shear
theory of failure. The shaft is turned from AISI 1030 hot-rolled steel. Use a factor of
safety of 2. Both the gear and pulley are held on the shaft by a profiled key.

30°, 7,

Problem Figure 14

15. The steel shaft shown in Problem Figure 15 is simply supported by bearings at
reactions R, and Rr. Determine the slope at each bearing and the maximum
deflection. Use the numerical method of Section 7-8. Check your answer by using
the computer program of Section 7-9.
 402
Chapter 7: Shafts, Keys and Couplings

1500 Ib
2500 Ib

Problem Figure 15

. For the shaft diameter found in Problem 5, determine the following: (a) The slope
at the bearing supports. (b) The maximum shaft deflection. (¢) The lowest critical
speed. Is the shaft in danger of vibrating excessively? Solve parts (a) and (b) by the
computer using the program found in Section 7-9.
. Solve parts (b), (c) and (d) of Problem 12 using the numerical method described in
Section 7-8.
. Solve parts (b), (c), and (d) of Problem 12 using the area moment method.
. Solve parts (b) and (c) of Problem 12 by graphical integration.’
. Determine the slope at bearing D and the maximum shaft deflection for
Problem 10.
. Determine the following for Problem 13: (a) The slope of the shaft at bearings A
and D. (b) The maximum shaft deflection. (c) The lowest critical speed of the system.
. Determine the following for Problem 14: (a) The slope of the shaft at bearing 4.
(b) The maximum shaft deflection. (c) The lowest critical speed of the system. For
parts (b) and (c), use the area moment method.
. Verify the statement that ‘‘a square key is equally strong in shear as it is in com-
pression.”
. If the shaft and key are made of the same material, determine the length of square
key required as a function of the shaft diameter.
. A plain standard square key is used to hold a 25-in. long hub on a 1;5-in. diameter
shaft. If the key and shaft are of the same material, having an allowank stress of
8000 psi, what is the factor of safety against shear failure of the key if the torque
being transmitted is 3500 in.-Ib? The length of key is 24 in.
26. A gear is keyed to a 3-in. diameter shaft, and delivers 80 hp at 200rpm to the
punch of a mechanical press. The gear materiai is class 60,000 cast carbon steel, the
shaft material is AISI 1040 hot-rolled steel, and the key material is AISI 1020
cold-drawn steel. Because of the shock loading, a safety factor of 4 is suggested.
Determine the length of a standard square key. What is the torque carrying capacity
of the shaft due to the keyway?

*” For example, see (1) Design of Machine Elements by V. M. Faires, The Macmillan Company,
New York, 1965.(2) Design ofMachine Elements by A. Vallance and V. L. Doughtie, The McGraw-
Hill Book Company, New York, 1951. (3) Mechanical Engineering Design by J. E. Shigley, The
McGraw-Hill Book Company, New York, 1972.
 403
Problems

27. An electric motor for a chemical blender supplies 60 hp at 1750 rpm to a 1}-in.
diameter shaft made of AISI 1045 cold-drawn steel. A cast steel gear with a 23-in.
long hub is keyed to the shaft by a square key. The gear material is class 85,000
carbon steel. Select a standard square key and determine its length if the key
material is AISI 1020 cold-drawn steel. Assume a factor of safety of 2.5 (mild shock)
since the viscosity of the material being mixed varies. By what percentage has the
torque capacity of the shaft been reduced due to the keyway?

28. A cast iron pulley is keyed by a number 812 Woodruff key to a 17-in. diameter
shaft made of AISI 1040 cold-drawn steel. If the shaft rotates steadily at 200 rpm,
what horsepower can the key transmit if it is made of (a) AISI 1020 cold drawn
steel, and (b) AISI 1060 quenched in oil and tempered at 1200°F.

29. A foot pedal lever of akick press is keyed to a 2-in. diameter shaft made of AISI 1040
cold-drawn steel by a radial taper pin with a 0.5-in. mean diameter. The pin is made
of AISI 1060 steel, oil quenched and tempered at 600°F. The lever, which is 22 in.
long, is subjected to a reversing load. For a factor of safety of 2 based on the
endurance strength, determine: (a) The safe lever load for the shaft. (b) The safe
lever load for the pin (direct shear only). (c) The safe lever load for the combination
of the shaft and pin.

30. In some instances, a key is so designed that it will fail if a limiting horsepower,
torque, or stress is exceeded. This is deliberately intended to protect equipment
“downstream” from the key in case there is a jam or overload. Thus, consider a
2-in. diameter shaft made of AISI 1040 cold-drawn steel subjected to torsion only.
If it is desired that the key shear at 75% of the shaft strength (based on ultimate
strength), how long would a standard square key have to be? The key material is
AISI 1020 cold-drawn steel.

31. A straight sided splined shaft of an automotive rear axle has 16 splines. The shaft
diameter D is 2.00 in. For a “‘not slide under load” fit, determine the torque carrying
capacity of this shaft and horsepower at 3000 rpm. The splines are 1.75 in. long and
are cut from AISI 1040 cold drawn steel.

32. A solid shaft for an automobile transmission has a 10-tooth involute spline with a
diametral pitch of 48. The shaft is made of AISI 4340 steel quenched in oil and
tempered at 1000°F. Based on yield, what is the factor of safety if the shaft transmits
150 hp at 3000 rpm?

33. A rigid flange coupling connects two 5-in. diameter AISI 1040 cold-drawn shafts.
The two halves of the coupling are held together by six evenly spaced 1-in. diameter
bolts located on a 12-in. bolt circle. Each flange half is keyed to the shaft by a
standard AISI 1020 cold-drawn square key. The flange web thickness is 15 in. The
flange material is class 60,000 annealed cast carbon steel and the bolts are made of
AISI 1030 cold-drawn steel. Because the coupling can be expected to experience
moderate shock, a factor of safety of 3 with respect to the yield strength should be
applied to the torque capacity of the shaft (see Section 1-3). (a) Assuming that the
bolts fit the bolt holes with zero clearance and are “finger tight” when assembled,
determine the factor of safety for each possible mode of failure. (b) If the bolts fit
the holes with clearance, determine the factor of safety based on yield strength for the
bolt tension force (that is, minimum tightening force) required to prevent slip in
the coupling. Also determine the factor of safety for the bolt tension force based on
(c) shearing of the key, and (d) compression of the key. For the flange configuration
see Figure 7-17d.
 404
Chapter 7: Shafts, Keys and Couplings

REFERENCES
[1] H. A. Berchardt: A comprehensive method for designing shafts to insure adequate
fatigue life. Mach. Des., (April 25, 1963).
[2] R. E. Peterson: Stress Concentration Design Factors. John Wiley & Sons, Inc.,
New York, 1962.
[3] R. Bruce Hopkins: Calculating deflections in stepped shafts and nonuniform
beams. Mach. Des., 33(14): 159-164 (1961).
[4] Virgil F. Faires: Design of Machine Elements, 4th ed. The Macmillan Co., New
York, 1965.
[5] M. J. Siegel, V. L. Maleev, and J. B. Hartman: Mechanical Design of Machine
Elements, 4th ed. International Textbook Co., Scranton, Pa., 1968.
[6] S. P. Timoshenko and J. N. Goodier: Theory of Elasticity. McGraw-Hill Book Co.,
New York, 1951.
[7] Sir Geoffrey Taylor: The use of soap films in solving torsion problems. Scientific
Papers of Sir Geoffrey Taylor, Vol. 1. Cambridge University Press, London, 1958,
ppal-23)
[8] I. S. Sokolnikoff: Mathematical Theory of Elasticity. McGraw-Hill Book Co.; New
York, 1956.
[9] C. T. Wang: Applied Elasticity. McGraw-Hill Book Co., New York, 1953.
[10] Colin Carmichael (ed.): Kent’s Mechanical Engineers’ Handbook, Vol. I, 12th ed.
John Wiley & Sons, New York, 1961.
[11] R. R. Moore: Bulletin No. 42, Engineering Experiment Station, University of
Illinois, Urbana, IIl.,
[12] Harold A. Rothbart (ed.): Mechanical Design and Systems Handbook, Sect. 27.
McGraw-Hill Book Co., New York, 1964.
[13] J. B. Armitage: Straight splined shaft. Mech. Eng., 70: 738-742 (Sept. 1948).
[14] SAE Handbook. Society of Automotive Engineers, Inc., New York, 1954.
[15] H. L. Horton (ed.): Machinery’s Handbook, 15th ed. The Industrial Press, New
York, 1957.
Journal Bearings
and Lubrication
SYMBOLS

A = cross-sectional area S’ = Sommerfeld Number

a I= eccentricity T=
torque
b = width of bearing T; = frictional torque
C = heat transfer coefficient, —eMIPELAlinc-mst
Btu/[(hr)(ft”)
(°F)] maximum velocity in x-direction
C, = specific heat = intermediate velocity in
c = radial clearance x-direction
D = diameter of bearing = viscosity index
Fe—MOTCe load, lb
Fp) = frictional horsepower = absolute viscosity (usually in
f = coefficient of friction centipoise)
H = heat transfer, Btu/hr mathematical proportionality
= film depth symbol
h, = minimum film thickness empirical pressure—viscosity
L = length of bearing constant
n = speed, rpm = eccentricity ratio
n’ = speed, rps = partial bearing angle
P = load per projected area of absolute viscosity (usually in
journal reyns)
p = pressure kinematic viscosity
Q = flow rate = mass density or specific gravity
R = maximum radius shearing stress
r = radius
S = viscosity, Saybolt Universal
Seconds

This chapter will present a discussion of the fundamentals of lubrication,

with particular emphasis upon the design of journal bearings. The analytical
approach as well as the use of design charts will be discussed.

405
 406
Chapter 8: Journal Bearings and Lubrication

SECTION 8-1
Introduction
A major problem faced by engineers through the years has been how to
prevent the loss of useful energy due to friction. It has been estimated that of
all the energy produced throughout the world,as much as one third to one half
is wasted because of friction. Friction may be defined as the opposing force
that is developed when two surfaces move relative to each other. Lubrication
has then been defined as the art of reducing the friction developed between
two surfaces moving relative to each other. This is usually done by placing a
substance between the moving surfaces. Although the material that is called
a lubricant is usually in the liquid state, solids and gases are also used as
lubricants.
Plain bearings, as distinguished from the rolling element bearings to be
discussed in the next chapter, are usually classified as plain journal or sleeve,
thrust, spherical, pivot, or shoe-type thrust. Another method of classification
is to designate the bearing according to the type of lubrication used. .
VQTOQa U } U

(liquid or gas) to separate the moving surfaces completely. When this ideal
condition is attained, we speak of the lubrication as being thick film. Other
commonly used terms for this type of lubrication are fluid film, stable, or
perfect. In order for hydrodynamic lubrication to exist, relative motion must
exist between the two surfaces and a pressure must be developed. The pressure
development is accomplished by the wedging action that results when the
two surfaces are not parallel. Figure 8-1 illustrates the situation just described.
If the fluid film gets thinner and is no longer able to separate the moving
surfaces, partial metal-to-metal contact can occur; this type of lubrication is

Moving surface

Stationary surface

Figure 8-1 The conditions necessary for hydrodynamic lubrication to exist are
shown. Pressure is developed in the lubricant due to the wedging action between
the two surfaces having relative motion.
 407
Section 8-2: Types of Journal Bearings

called boundary lubrication. As more metal-to-metal contact takes place the

friction developed becomes greater, and, unless the bearing was designed to
operate under this thin film type of lubrication, failure will occur. One ofthe
design criteria which is very often used is the so-called minimum film thick-
ness. In this method, the designer makes certain that the film thickness does
not become less than some minimum safe value.
Boundary lubrication is usually not planned by the designer. It depends on
such factors as surface finish, wear-in, and surface chemical reactions. Low
revolution per minute, heavily loaded bearings, misaligned bearings, and
improperly lubricated bearings are usually more prone to operate under
boundary lubrication. If a bearing operates under boundary lubrication but
does not fail, it is possible that the wearing-in that takes place will then allow
the bearing to operate with hydrodynamic lubrication.
Boundary lubrication presents yet another problem to the designer: it
cannot be analyzed by mathematical methods but must be dealt with on the
basis of test data.

SECTION 8-2
Types of Journal Bearings
it at su ial « OWI journal
bearing. As can be seen from Figure 8-2, a journal bearing consists of two
main parts, a shaft called the journal and the hollow cylinder that supports
the shaft known as the bearing. In most applications, the journal rotates
while the bearing is stationary. However, there are applications where the
journal is stationary and the bearing rotates, and even some where both the
journal and bearing rotate. i
As will be seen in Section 8-10, the pressure distribution around a journal
bearing varies greatly. As a result, journal bearings can be made with the
full bearing thickness around the whole circumference (a full journal bearing)

Figure 8-2 An example of a journal bearing.

 408
Chapter 8: Journal Bearings and Lubrication

or with the full bearing thickness around a portion of the circumference and
an oil cap around the remainder of the circumference (partial bearing).
Figure 8-3A shows a full journal bearing, whereas a partial journal bearing
is depicted in Figure 8-3B. A partial bearing having zero clearance 1iS known
as a fitted bearing.

Full Journal Bearing

Journal Bearing

LO}
a

iin
spouses
Where c = radial clearance = ie Gf
r; = radius of journal
rp = radius of bearing
L = bearing length
Partial Journal Bearing

Oil cap

Figure 8-3 (a) Full journal bearing.

(b) Partial journal bearing.

(b)
 409
Section 8-3: Newton's Law of Viscous Flow (Viscosity)

SECTION 8-3

Newton’s Law of Viscous Flow (Viscosity)

One of the fundamental problems involved in lubrication theory is the
effect of the internal resistance of the fluid being used as a lubricant. The
expression for this internal resistance, called viscosity, can be obtained by
applying Newton’s law of viscous flow to the system shown in Figure 8-4.
In the figure, the lower plate is stationary, while the upper plate is moving to
the right with a velocity U. The two parallel plates are separated by a lubricant
film of thickness h.
When two plates, having relative motion, are separated by a lubricant, in
this case an oil film, a flow of oil takes place. In most lubrication problems,
conditions are such that the flow that occurs is laminar. By laminar flow we
mean that the fluid is in layers or laminae, which are maintained as the flow
progresses. When this condition is not met, we speak of the flow as being
turbulent.
Under conditions of laminar flow, we may assume that perfect adhesion is
achievable. In other words, the lamina immediately adjacent to the moving
plate moves with the same velocity as the plate, whereas that immediately
adjacent to the stationary plate has zero velocity. The intermediate laminae
move with velocities that vary linearly from 0 to U. Because the laminae have

A = cross-sectional area

LLLLLLLA
Figure 8-4 Two plates having a relative velocity, which are separated by an oil
film and under laminar flow, have a linear velocity gradient, as shown.
 410
Chapter 8: Journal Bearings and Lubrication

different velocities, each layer must slide upon the adjacent layer. A force, F,
is required to produce this sliding. The resistance the fluid offers to this force,
is known as the shear stress, which in turn causes fluid friction.
Newton’s law of viscous flow states that at any point in the fluid film, the
shearing stress is proportional to the velocity gradient.

du
t(stress) oc —
dy
or

d
c= or (8-1)
where su is a proportionality constant called either the coefficient of absolute
viscosity, absolute viscosity, or more usually simply viscosity, du is the
change in velocity, and dy is the change in film thickness.
Because the velocity varies linearly across the film,
du _ U
dyh
The shearing stress may be written as
IP
t=—
A

where A is the cross-sectional area being sheared. Therefore

Fall
er 24
or

AU
F =p (8-2)
Thus the shearing or frictional force required increases with velocity and
area. It will be recalled that solid friction is relatively independent of speed and
area but dependent on load.
Viscosity, a fluid’s resistance to flow, can be expressed dimensionally by
solving equation (8-2) for pw.

mde (force) (length)

m
AU (length?) (length/time)
In the English system

(1b) (in.) _ Ib-sec _

(in.)(in/sec) in?)
 41]
Section 8-4: Hagan-Poiseuille Law (Flow Through Capillary Tube)

The reyn, named in honor of the English physicist Reynolds is the viscosity
unit used in the English system. In the n the standard unit is the
poise, named after the French physician Poiseuille.
(force)(time) _ dyne-sec
length? = = =cm Z
Tee
A smaller unit, the Seales aaamae is frequently used.
The simple c e yns é ses (cp) is
=> lereyne= 0:9 10 Cp

SECTION 8-4
Hagan-Poiseuille Law (Flow Through Capillary Tube)
Because many methods that are commonly used to measure viscosity make
use ofthe flow through a capillary tube, it is important to discuss the Hagan-
Poiseuille law for laminar flow through a capillary tube. The following
assumptions will be made and refer to the tube shown in Figure 8-5. The
pressure drops gradually along the tube in going from left to right, the flow
is steady and laminar, the fluid is incompressible, the fluid completely fills
the interior of the tube, the capillary diameter is small enough so that the
pressure drop across the tube can be ignored, and the fluid does not accelerate
as it flows through the tube.
A cylindrical element offluid of radius r and length dx is shown in Figure
8-5 with the pressure (p) and shearing stress (t) drawn in. Because the element
is in static equilibrium, the summation of forces along the x axis must equal
zero.
prr? — (p — dp)ar? — 12ardx = 0
dp = 2 dx (8-3)
Newton’s law of viscous flow may be written
du
Ly
(8-4)
End View

Figure 8-5 The free body diagram ofthe forces acting in the x direction, on an
element dx, of the lubricant flowing through a capillary tube.
 412
Chapter 8: Journal Bearings and Lubrication

the minus sign is used since the velocity u decreases as the radius increases.
Substituting equation (8-4) into equation (8-3) and solving for du,we get

dp = ———2d
i: r dr
and
r dp
du = ———d
i 2u dx
Integrating with respect to r gives

Bee iedil |d
2" dx
li dpr
= — C
ieee
where C is a constant of integration. Using the boundary condition that ‘the
velocity = 0, whenr = R we get

Therefore

(8-5)

Since equation (8-5) is the equation of a parabola, the velocity distribution

across a capillary tube is a paraboloid.
The flow rate (Q) can be determined by integrating the product of the
velocity at some radius r by the cross-sectional area of the differential element
at that radius. A cylindrical shell must be used because it has uniform velocity
over its cross-sectional area, whereas the cylindrical element used previously
will not have uniform velocity. Figure 8-6 shows the element we will consider.
dQ =udA

Q
SUpatags
ps ip Ree ieee
cn)
vt; nmr dr

dp Rear eta
irae 4}o
mR* d
g=-—-= (8-6)
 413
Section 8-4: Hagan-Poiseuille Law (Flow Through Capillary Tube)

End View

LLL LLL LLL LLL

Figure 8-6 The cylindrical shell free body diagram used to derive the
expression for the flow rate through a capillary tube.

Equation (8-6) could be used as the basis for determining values for the
absolute viscosity pL.
Another interesting flow equation results, when an inclined, open ended
capillary tube is considered. Because the tube shown in Figure 8-7 is open to
the atmosphere at both ends, the pressure difference between the two ends is
due only to the vertical height, h. Thus,
dp _ pgh
dx iL

where p is the mass density, g is the acceleration ofgravity, and L is the length
of the tube. Substituting into equation (8-6) we get

or

Figure 8-7 Sketch of an inclined, open ended

capillary tube.
 414
Chapter 8: Journal Bearings and Lubrication

Here ji/p defined as the kinematic viscosity, is represented by the greek letter
v, and is expressed in stokes. Obviously the conversion between absolute and
kinematic viscosity involves only the mass density.

=3 (8-7)
where v = kinematic viscosity in stokes, » = absolute viscosity in poises,
and p = mass density.

SECTION 8-5
Viscosity Measuring Devices
The Saybolt Universal viscometer is perhaps the most widely used
viscosity measuring instrument. It utilizes the equation for flow through a
capillary tube to measure the kinematic viscosity of lubricating oils. The
instrument itself consists of a standard 60-cm? receiving flask and a reservoir
surrounded by a constant temperature bath. The oil sample to be tested is
placed in the reservoir and, when the bath and the reservoir reach the desired
test temperature, a stopper at the bottom is removed. The time required for
60-cm* of the oil to flow through a capillary tube is recorded. This time,
measured in seconds, has become a standard unit of viscosity known as
, or simply SUS.
A rather simple formula may be used to
18
Z =), (mos = Se (8-8)
where Z = absolute viscosity in centipoise at test temperature t,S = Saybolt
Universal Seconds, and p, = specific gravity at test temperature, f.
The specific gravity at the test temperature may be obtained from equation
(8-9)
> Pr = Poo — 9.00035(t — 60) (8-9)
where P59 = specific gravity at 60°F (standard) and ¢ = test temperature,
4 ae

Example 8-1. Determine the absolute and kinematic viscosity in reyns and centistokes
(cs) of an oil that had an SUS rating of 120 at 100°F (the specific gravity at
60°F can be taken as 0.89).

Solution:

Pr = P60 — 0.0003 5(t a 60)

= 0.89 — 0.00035(100 — 60)

= 0.89 — 0.014
= 0.876
Ze p{o2as x =
 415
Section 8-6: Temperature Effect on Viscosity

180

= 0.876(24.9)
= 23 GD

w= 21.8 x (1.45 x 1077)

= 31.6 x 1077 reyns
OES
=— =—— = 24,
a ao cibee ok
v can also be obtained directly from
180 180
v = 0.228 — ie 0.22 x 120 — ‘0 e 249cs @

SECTION 8-6
Temperature Effect on Viscosity
Unfortunately the viscosity of lubricants varies with changes in tempera-
ture. In fact, gases such as air have an increased viscosity with increased
temperature, whereas the viscosity of liquids decreases with increased
temperature. Clearly, the designer would prefer lubricants that have as small
a viscosity change as possible with respect to temperature change.
A means of rating the response ofan oil’s viscosity variation to temperature
change was worked out in 1929 by Dean and Davis [1]. They introduced the
concept of a viscosity index. Equation (8-10) is used for calculating the
viscosity index.
L—U
VI= x 100 (8-10)
Joi Jal
where VI = viscosity index in per cent, L = viscosity of a standard 0% VI
oilat 100°F, H = viscosity of astandard 100% VI oil at 100°F, U = viscosity
of the unknown VI oil at 100°F.
The procedure for determining the viscosity index ofan unknown oil is as
follows:
1. Measure the viscosity of asample of the unknown oil at 100 and 210°F.
The viscosity at 100°F is used as U.
2. From the standard oils having a VI of 100%, select one that has the
same viscosity at 210°F as the unknown oil. The viscosity of this oil at
100°F is then used as H.
3. From the standard oils having a VI of 0%, select one that has the same
viscosity at 210°F as the unknown oil. The viscosity ofthis standard oil
at 100°F is then used as L.
Substitution of these values into equation (8-10) will give the value of the
viscosity index for the unknown oil. This procedure can be represented
graphically as shown in Figure 8-8. Table 8-1 is a listing of L and L — H
 416
Chapter 8: Journal Bearings and Lubrication

values to be used in equation (8-10).

Equations for determining L and H when the
viscosity at 210°F is above those given in the table as well as procedures for
determining viscosity indexes over 100% can be found in the ASTM hand-
book.

Table 8-1 Values of L and D for Calculating Viscosity Index from Saybolt
Universal Viscosity.

Saybolt Universal Saybolt Universal

Viscosity at 210°F. Viscosity at 210°F.
SUS SUS

Table 8-1. (continued)

Saybolt Universal Saybolt Universal

Viscosity at 210°F. Viscosity at 210°F.
SUS (Cf)
SUS

14854
185 15373
190 15901
195 16436
200 16981
205 17536
210 18098
215 18669
220 19249
225 19838
20435

SOURCE: Extracted from ASTM D567, Standard Method for Calculating Viscosity Index.
 417
Section 8-6: Temperature Effect on Viscosity

SUS
Viscosity,

Temperature, °F

Figure 8-8 Illustration of graphic procedure that can be used to

determine the viscosity index.

Example 8-2. Determine the viscosity index of an unknown oil having a viscosity of
89 SUS at 210°F and 1400 SUS at 100°F.

Solution: From Table 8-1, for 89SUS, interpolation gives L = 2065,

L — H = 1099. Substituting into equation (8-10) we get
L-—U 2065
— 1400

665
=——
x 100=60% @
 418
Chapter 8: Journal Bearings and Lubrication

SECTION 8-7
Petroff’s Law

Under this condition, the journal is

concentric in the bearing. Obviously, these conditions of operation could not
be applied to an actual bearing. However,
1. The designer
usually uses Petroff’s equation to obtain a first approximation of the
frictional losses and from then on uses more exact methods (to be discussed
in Section 8-10) to obtain frictional values. Two concentric cylinders repre-
senting a journal rotating in a bearing are shown in Figure 8-9.
From Newton’s law

and the frictional force is

so that the frictional torque is

U
T, = Fr = pA—r
h

Since U = rw, w = 2nn/60, and A = 2xrL

2mn
T,7 == H2ar
p2arLr—
rope

_ a plein (8-11
emaich
where

T, = frictional torque, in.-lb

p = absolute viscosity, reyns
L = length of bearing, in.
r = journal radius, in.
n = journal speed, rpm
h = radial clearance or film thickness, in.

The frictional horsepower can be obtained from equation (8-12).

3 T,n
F Fy== (8-12)
63,000
 419

©)
Section 8-8: Hydrostatic Lubrication

| Figure 8-9 The two concentric cylinders shown

represent a journal rotating in a bearing with zero
eccentricity.

“ae |

SECTION 8-8
Hydrostatic Lubrication
Although most of the applications involving journal bearings are of the
hydrodynamic type, to be discussed in Section 8-9 of this chapter, there are
enough instances involving hydrostatic lubrication to warrant a brief
discussion. Some of the more common examples are thrust bearings, oil lifts
needed during starting of heavily loaded bearings, bearings used in low speed
applications, and many others too numerous to mention.
In order to illustrate this type of lubrication in a little more detail, consider
the vertical shaft thrust bearing shown in Figure 8-10. The sketch is merely
to be used to demonstrate the principle upon which this form oflubrication
operates because an actual thrust bearing of this type would be of a much
more complicated construction. The vertical shaft, rotating at nrpm is to
support a vertical load W. Oil at an inlet pressure py, is pumped into the
 420
Chapter 8: Journal Bearings and Lubrication

LZ.
Oil
( outlet

To reservoir |To reservoir

Figure 8-10 Simplified sketch of a vertical shaft thrust bearing.

recess of radius R;. The oil is forced outward under continuously decreasing
pressure, then upward along the sides of the shaft, and then finally returns
through a system of piping to the reservoir.
Another type of hydrostatic problem concerns hydrostatic lifts. This
problem arises when journal bearings are starting, stopping, or operating at
low speeds. The problem can usually be resolved by supplying oil under
pressure to the bearing.
The final type of hydrostatic lubrication we will consider is the so-called
squeeze film problem. For example, when a machine is temporarily shut
down, the load carrying capacity should become zero because the relative
motion needed to develop the pressure to support the load is no longer
present. However, it was observed that, contrary to expectations, a load
carrying capacity did indeed exist. The explanation was that the squeezing
out of the lubricant as the surfaces approached each other developed a
 42]
Section 8-9: Hydrodynamic Lubrication

pressure which was capable of supporting a load for a certain interval of time.
The designer is interested in knowing how much time is required for the film
to reach a predetermined minimum value. Clearly, this information can be
used to determine whether the film will break down before the machine is
started up again. If it is determined that the film would have broken down
(resulting in metal-to-metal contact), an external supply of pressure may be
needed to lift the surface before the machine is started.
A much more complete discussion of hydrostatic lubrication, including the
derivation of equations needed to compute such items as load carrying
capacity, flow, and supply pressure required, can be found in reference [2].

SECTION 8-9
Hydrodynamic Lubrication
The hydrodynamic theory of lubrication has been in existence since the
1880s. Beauchamp Tower’s experimental work on railroad journal bearings
induced Osborne Reynolds to perform the mathematical analysis that led to
the formulation of the differential equation bearing Reynolds’ name. A
simple definition of what is meant by i i hat

We will now proceed to derive the two-dimensional (one-dimensional

flow) Reynolds equation. There are several assumptions that are necessary
to the derivation.

1. The lubricant is Newtonian; in other words, it obeys Newton’s law of

viscous flow.
2. The flow is laminar.
eS). The lubricant is incompressible.

4. The pressure across the depth and the width of the lubricant film is
constant.
5. The viscosity of the lubricant is constant throughout the film
6. The inertia forces due to the acceleration of the lubricant are small
enough to be considered negligible.
7. The lubricant film is thin enough so that any effect of bearing curvature
may be ignored.
8. The bearing is assumed to have an infinite width. In other words, there
is no end leakage (flow along width of bearing).
9. Perfect adhesion exists between the lubricant and the bearing surfaces.
This means that the layer of lubricant immediately adjacent to a bearing
surface moves with the same velocity as the surface.
Let us consider the situation represented in Figure 8-11. The moving surface
supports a load W, and has a velocity in the x direction of U. We will be
 422
Chapter 8: Journal Bearings and Lubrication

Moving
surface =

Stationary
surface

Figure 8-11 A lubricant separates the stationary plate from the plate moving
with a velocity U. The load supported is W.

interested in drawing a free body diagram of the forces acting in the x

direction on the element of sides dx, dy, and dz as shown. The free body
diagram is shown in Figure 8-12.
Summing the forces in the x direction, we get

0 4)
ee Oe ea. dy dz — pdy dz — t, dx dz + |t,+ "x dy dx dz
Ox oy
i) )
= oP ax dy dz + “*dx dy dz =0
aye Oy
or
Op 0t,~,
ay ae dxdyadz=

a OT
(+ ey) ax ae

(cot 82 x) dy dz pdy dz
- oo dy

dx

ee

T, UX dz

Figure 8-12 A free body drawing of the forces acting on the elemental
volume shown in Figure 8-11.
 423
Section 8-9: Hydrodynamic Lubrication

Therefore
Op Ot,,
ax ay
From Newton’s law

" Ou
ae meay

Tee Ga?
ey Moy?
Therefore,

ap _
Ox May?
or

eu 1 ap
dy? ox
It is now possible to obtain an equation for the velocity, u, by integrating
the above equation twice.

CieeeOD es
dy pox”

1 dp y’
Py Paper were (8-13)
The two boundary conditions we will use to evaluate the constants of
integration are (1) u = U when y = O, (2) u = 0 when y = h where U is the
velocity of the moving plate.
Applying (1) to equation (8-13) we get c, = U and applying (2) to equation
(8-13) we get
PA
0= 1 dph
+ es Tee

or
es PU
eerie ox ) oh
Therefore
1dpy? 1 @p U
Berean oe 6 Re
1) dee y
ee pean
a al Fel) ie hy) + Ue
| 4 :
(8-14)
 424
Chapter 8: Journal Bearings and Lubrication

We now make use of the assumption of incompressibility to write an

equation stating that the quantity offluid entering the elemental cube shown
in Figure 8-12, must equal the volume of fluid leaving the cube. In the
following equations, u = velocity along x axis, v = velocity along y axis, and
w = velocity along z axis.
a)
udydz+vudxdz+wdxdy = [+ oedexdydz + (>
+ = as|dx dz
. y
Ow
- [»+ was] dx dy
0z

aOe Oy aDgel a een

or

Ou ar Ov fsow =
Ox "Oy "Gz

But since we assumed no end leakage, ¢w/éz = 0. Therefore

ao
dy
_ ou6x
We now differentiate equation (8-14) to get
6) il @
gs
dx Ox |Qu exZy— hy + u(1— 3)
Therefore

Ov 0 |1 op
= Deedeede ee
oy i Ox 2p Ox h
Upon integrating with respect to y, with y going from 0 to h, we get

ao
* Ov
o Oy
"@ | op Wy —h)
6 OX | Oe meee
a= (oe
"aU
eels
y
9 OX h

Uy}
Se ute eto ae i ney
elk Ox Ox 2uU 0. Ox oh 0
Because we assume that perfect adhesion exists between lubricant and
bearing, v = 0, when y = 0, and y = h. Therefore

0=— Gone tlh: a ans Bs 27) umesh

ox E 2u |3 | | Fuln 4 ane
6 [h*> ap\ _U oh
€x \12u dx} 2 ex
 425
Section 8-10: Design Charts

and

"5
Calencp oh
a = 6uU— el
Ox a Ox CL)
Equation (8-15) is the one-dimensional flow Reynolds’ equation. If end
leakage is not ignored, an analysis similar to the above will yield the following
form of Reynolds’ equation shown as equation (8-16).

te fie) - eave (8-16)

Ox Ox Oz Oz Ox

This equation has been solved by numerical methods.

SECTION 8-10
Design Charts
Raimondi and Boyd were able to apply digital computer techniques
toward the solution of the fundamental lubrication equations, and present the
results in the form of design charts and tables [4]. The solutions obtained are
mathematical and thus give the theoretical results that the designer can
expect. The charts are plotted with the bearing characteristic number, also
known as the Sommerfeld number, S’, as the abscissa.
2 !

cae (“ ~ (8-17)

S° = bearing characteristic number, dimensionless

r, = radius ofjournal, in.
radial clearance, in.
II absolute viscosity, reyns
relative speed between journal and bearing, rps. In some cases n is
SAS
SS
expressed in rpm, with the result that S’ is in sec/min, instead of
being dimensionless
PP =Aoad pert projected area of journal, psi
Figure 8-13 is a plot of temperature versus viscosity for various SAE oils,
and Figure 8-14 is a diagram of the pressure distribution around a journal
bearing and also shows some of the nomenclature commonly used.
Before proceeding to the presentation of the design charts, some of the
problems of presentation must be discussed. Boyd and Raimondi first
published a paper [6] in which they considered bearings having no end
leakage ;in other words, bearings having L/D (length of bearing to diameter
of bearing) ratios equal to infinity. Then in reference [4], part I, they extended
the work to bearings having L/D ratios of 1, and in part II, L/D ratios of5and $
were discussed. Inall their papers to this point, they assumed that the lubricant
 Temperature, “C
0 20 40 60 80 100 120 140

viscosity,
Absolute
ureyns
Absolute
viscosity,
centipoise

Ib
10~®
10-6
sec/in2)
(ureyn
=
reyns

Temperature, °F

Figure 8-13 Viscosity-temperature chart for determining viscosity of typical

SAE numbered oils at various temperatures. [From A. A. Raymondi and J. Boyd:
A solution for the finite journal bearing and its application to analysis and design,
Parts I, II, II. Trans. American Society of Lubrication Engineers, 1(1): 159-209
(1958).]

426
 Figure 8-14 The pressure variation
around a journal bearing is shown. [From
A. A. Raymondi and J. Boyd: A solution for
the finite journal bearing and its
application to analysis and design, Parts I,
Film pressure Il, Il. Trans. American Society of Lubrication
distribution Engineers, 1(1): 159-209 (1958).]

Pmax

(dimensionless)

ho ccm

Locus of journal center for various

values of bearing characteristic No. S

variable,
thickness
film
Minimum
—

0 ‘O1e § 202 04.06 .08.10 20 40.60 801.0 2.0 4.0 6.0 8.0 10
ae me)Tall
Bearing characteristic no.,S =

Figure 8-15 thickness variable versus bearing characteristic

number (L/D = 1). [From A. A. Raymondi and J. Boyd: A solution for the finite
journal bearing and its application to analysis and design, Parts I, II, III. Trans:
American Society of Lubrication Engineers, 1(1): 159-209 (1958).]

427
 =F
Cc

variable,
friction
of
Coefficient

0 .01 02 .04 .06 .08.10 .20 40 .60 .801.0 2.0 4.0 6.0 8.010
iz. Z '
Bearing characteristic no. S’ -(2)ape
c

Figure 8-16 For determining Ca From A. A. Raymondi and

J. Boyd: A solution for the finite journal bearing and its application to analysis and
design, Parts I, II, III. Trans. American Society of Lubrication Engineers, 1(1):
159-209 (1958).]

Flow
variable,

0 .01 .02 04 .06 .08 .10 .20 40 .60 .801.0 2.0 4.0 6.0 8.010

Fi: 2
Bearing characteristic no., S’= G Jess

Figure 8-17 For determining [From A. A. Raymondi and

J. Boyd: A solution for the finite journa! bearing and its application to analysis and
design, Parts I, II, III. Trans. American Society of Lubrication Engineers,
1(1): 159-209 (1958).] :

428
 429
Section 8-10: Design Charts

film was continuous throughout the bearing circumference and that the film
pressure was high enough to prevent rupture. Although this assumption
was valid for lightly loaded or high speed bearing applications, many bearings
operate under conditions that do cause film rupture. For this reason, part II],
was devoted to the results obtained when film rupture was considered. The
designer can therefore use the results of parts I or IJ to ascertain whether
subatmosphere pressure exists (with consequent film rupture). If rupture does
occur, the results of part III are to be used.
A major difficulty encountered in presenting this material in a text of this
kind is the question of space limitation. The solution finally arrived at was
to present a series of charts from part III for L/D ratios equal to 1. Figures
8-15 through 8-19 present design charts that are sufficient for classroom
purposes. Those desiring more complete information, can find it in the
references cited. The best way to explain the use of the charts is to do an
example.
Example 8-3. Analyze a 180-deg partial journal bearing operating under the following
conditions: n = 3600 rpm; n’ = 60rps; W= 1600lb; pp = 2 x 1077 reyn;
c = 0.002 in.; r; = 2in.; L = 4 in.
Solution:

W 1600
in 2r,L (2)(2)(4) — ou Es
2 \?2(2 x 1077)(60)
oe cana Cyatt100 Fig
L/D = 4/4 =1
6 = 180 deg

where D = 2r,, and where @ is the partial bearing angle shown in Figure 8-3.
Part 1: To determine the minimum film thickness, refer to Figure 8-15.

ho
z = 0.39
Therefore
hy = 0.39 x 0.002 = 0.00078 in.

Since (iaatamatemetamiamemelieetet hereore

al Lope {0.39
Cc

= 0.61
If this value for minimum film thickness is less than the value the designer has
chosen as a minimum, the design must be changed.
Another way in which the chart can be used is to decide on a minimum film
thickness and read the corresponding Sommerfeld number. Bearing dimensions
or viscosity can then be adjusted to conform to this Sommerfeld number.
 430
Chapter 8: Journal Bearings and Lubrication

1.0

0.9

0.8

er Q

Flow
ratio,

0 01 02 04 .06 .08.10 .20 40 .60 .801.0 2.0 4.0 6.0 8.016

2
Bearing characteristic no.,S 1_{\"
= 3 un’
P

Figure 8-18 For determining aie [From A. A. Raymondi and J. Boyd:

A solution for the finite journal bearing and its application to analysis and design,
Parts I, II, III. Trans. American Society of Lubrication Engineers, 1(1): 159-209
(1958).]

Part 2: The frictional power loss will next be determined. This is the power
needed to overcome the fluid friction in the bearing.
From Figure 8-16 for S’ = 0.12

i z = 0.0022

Then frictional torque, T; is

T,; = f Wr; = 0.0022 x 1600 x 2 = 7.04in.-Ib
and

_ Tn _ 7.04 x 3600
Fi, = 0.4hp
P 63,000 ~—-63,000
Part 3: Another variable of some interest to the designer is the flow, Q. The
chart for determining the flow requirement is given in Figure 8-17. If the
lubricant is supplied to the bearing from some external source, say an oil
reservoir, the rate at which the lubricant must be supplied can be determined
from the figure.
It should be pointed out that the failure to supply a flow Q will not necessarily
cause a bearing failure but will change the performance characteristics of the
bearing. In other words, if the bearing is to have the performance characteristics
indicated by the charts, the amount of oil supplied to the bearing must equal Q.
 43]
Section 8-10: Design Charts

It must be emphasized that, when bearings are lubricated by external means

under pressure, special methods of solution must be used.
From Figure 8-17, for S’ = 0.12

Therefore

OB Ee a00025 60 a4

== 0 in.7 sec

P
JyCoAt

Temperature
variable,
rise

0 .01 .02 .04 .06 .08.10 .20 40 .60 .801.0 2.0 4.0 6.0 8.010

Bearing characteristic no.,S'= (4) un

Figure 8-19 For determining film temperature rise. [From A. A. Raymondi and
J. Boyd: A solution for the finite journal bearing and its application to analysis and
design, Parts I, II, III. Trans. American Society of Lubrication Engineers, 1(1):
159-209 (1958).]
 432
Chapter 8: Journal Bearings and Lubrication

Part 4: Another consideration with regard to flow is the amount of end

leakage. This end flow will cause a decrease in pressure along the bearing. The
quantity of oil flowing can be obtained by using Figure 8-18 and the ratio
0/0. For S’= 0,12

Qs = 0.58
Q
Q, = 0.58 x Q = 0.58 x 3.07 = 1.78 in.?/sec

Part 5 : The temperature rise that occurs in the oil film due to fluid friction
is of obvious interest to the designer. It is almost impossible to get a precise
value for this temperature rise because some of the heat developed is carried
away by the oil and some heat is transferred by conduction, convection, and
radiation. Figure 8-19 may be used to obtain conservative estimates _ tem-
perature rise.

ze. In Section 8-11we will discuss the question

oftemperature rise in Srestet detail, but for the moment we will use the chart
to obtain a first estimate of temperature rise.
For S’ = 0.12 and where Cy = specific heat, 0.42 Btu/(Ib)(°F); J= mechani-
cal equivalent of heat, 778 ft-Ib/Btu ;and y = weight density, 0.03 Ib/in.*
JyCoAt
P
= 1

Vs el 12 x 100
Nig
JyCyo 778 x 12 x 0.03 x 0.42
=102°F ©

SECTION 8-11
Heat Balancing of Bearings
By heat balancing a bearing, we simply mean the determination of the
balance that exists between the heat developed and the heat dissipated’in a
bearing. There are two general types of bearings to be considered, the self-
contained bearing and the pressure-fed bearing. We will devote our attention
first to the self-contained type. A self-contained bearing is one in which the
lubricant is contained in the bearing housing, which is sealed to prevent oil
loss. Obviously, the self-contained bearing is economically more desirable
because it does not require expensive cooling or lubricant-circulating systems.
Bearings of this type dissipate heat to the surrounding atmosphere by means
of conduction, convection, and radiation. They are widely used in industry,
and are known as pillow-block or pedestal bearings.
The important question to be answered from the designer’s point of view,
is what will be the equilibrium temperature of the oil film. In other words,
after an interval of operation, what will be the film temperature when the heat
developed is exactly equal to the heat dissipated.
average
temperaturefilm is 160°F. If the temperature rises above 200°F,
 435
Section 8-11: Heat Balancing of Bearings

harmful deterioration of the lubricant as well as damage to the bearing

material can take place.
The heat dissipated from the bearing housing may be approximated by
equation 8-18

PH = CA(t, =t,) (8-18)

where
H = heat lost, Btu/hr
C = heat transfer coefficient, Btu/(hr)(ft?)(°F)
as |= effective area of bearing housing through which heat is being
transferred, ft?
t, = temperature of the bearing housing, °F
t, = ambient air temperature, °F.
The heat transfer coefficient depends on many factors, such as the surface
condition of the housing, the material and shape of the housing, and the
velocity of the surrounding air. A value for © of 2 Btu/(hr)(ft7)(°F) for still
air is widely used: a value of 5.9 Btu/(hr)(ft*)(°F) is suggested for air having
a velocity of 500 ft/min.
The bearing housing area, which is dissipating heat, is sometimes difficult
to calculate. A reasonable estimate for this area is that suggested by Fuller [2].
For simple pillow-block bearings, the effective area in square inches
12.5DL and for pillow-blocks with separate shells 20DL is recom-
mended.
Another difficulty the designer faces is that, although equation (8-18) may
give the rate at which the housing is dissipating heat, the average oil film
temperature rise may be greater than this value. The construction of the
bearing and lubrication method used are the factors to be considered. Faires
[8] suggests the following approximations.
For oil-ring bearings in still air

t, — t, = Qt, — t,) (8-19a)

where 1, 1s the oil film temperature.

For oil-bath bearings in still air
ematnes lea (tat) (8-19b)
For waste-packed bearings in still air

Es < La oa D Ait, oH iis) (8-19c)

For air moving at 500 ft/min, the oil-ring bearing constant should be
increased by 15-20%. We will demonstrate the procedure that can be
followed to determine the oil film temperature in Example 8-4.
 434
Chapter 8: Journal Bearings and Lubrication

Example 8-4. An oil-ring full journal bearing is to operate in still air. The bearing
diameter is 3 in., and the length is also 3 in. The bearing is subjected to a load
of 1000 Ib and is rotating at 500 rpm, the radial clearance is 0.0025 in., the oil
is an SAE 30, and the ambient air temperature is 80°F. What is the value of the
viscosity of the oil.
Solution: Assume the average oil temperature is 160°F. Then from Figure
8-13
pw = 2.8 x 107° reyns
—-6 500
pair ak

ore Fs eA ay adee i ae * 60
TNH el OMICS: 1000
3% 3
= 0.0758
From Figure 8-16

if = 2
EA

eS2.4
ie eS
.002

Therefore, the heat developed is

Hf WV = 0004 100015 15008 min dane

rev
eae
12
= 1570 ft-lb/min
Then from equation (8-18)

H = CA(t, — t,)
2 Btu ft-lb l hr
Cae ae
nG2Ch ~ Bm ~ comin
= 25.93 [ft-Ib]/[(min)(ft?)(°F)}

A =12 SDL= 125 33 i= 12 San

pe
b aa 2
CA

1570”
= min

ft-lb me
03S ein ee
Gain) GCC His tea eae
= 781cR
t, = 78.1 + t, = 78.1 + 80 = 158.1°F
From equation (8-19a)
to = ba = 2b, — 0,2 x 78 =9156:2°R
t, = 156.2 + t, = 156.2 + 80 = 236.2°F
 435
Section 8-11: Heat Balancing of Bearings

This does not agree with the assumed value of 160°F. Therefore, assume
t, = 200°F. Then from Figure 8-13
p= 1.3 x 107° reyns
Ee 500
SA
Ke r, iat its a {led < NOY sx 60

RS ary2 0.0025 1000

3 3

= 0.0352
From Figure 8-16

i
Sele
a >
1S 0025
f= FG 0025
1.5
Therefore

2n ;
H = fWV = 0.0025 x 1000 x 500 x D xls

= 982 ft-lb/min

Sg

MS OB) sx :
144
t, = 48.5 + 80 = 128.5°F
= th S Ate, =) =H BK AES = OB
fin = OH ce 0) Ie
This is closer to the assumed value of 200°F.
The next trial would be 190°F. In any event, because the film temperature
is less than 200°F, external cooling will not be required.

We will now turn our attention to the heat balance in forced-feed bearings.
As was stated in Example 8-4, in those cases when oil film temperature
exceeds 200°F, some external source of cooling is required. The usual method
of solving this problem is to supply oil under pressure to the bearing. As the
oil flows through the bearing, it picks up heat from the bearing. The oil is
then returned to a reservoir or sump where it is cooled before being recir-
culated.
A common method used for pressure lubrication 1s the so-called circum-
ferential groove method. Figure 8-20 is an illustration of a circumferential
groove bearing. As the name implies, a circumferential groove is cut at the
center ofthe bearing, the oil is forced into the groove through an oil supply
hole placed in the groove, opposite the portion ofthe oil film that is supporting
the load. The groove, however, has the disadvantage of breaking up the active
length of the oil film. As a result, the pressure distribution does not vary as
 436
Chapter 8: Journal Bearings and Lubrication

VMMMMMMMMMdbbddbbb
0 @CZTZ@M0000bbbbbo0
bd
Figure 8-20 Circumferential groove going completely around
the circumference of the bearing gives this bearing its name.

the smooth curve in Figure 8-21 but rather is more like the curve b in the
same figure.
Despite breaking the oil film in half, the efficient cooling obtained allows
larger loads to be applied to the bearing. Equation (8-20) can be used for
determining the theoretical flow through a 360-deg circumferentially grooved
bearing.

(8-20)

where the symbols have the same meaning as in Section 8-10.

b
Grooved
bearing

Ungrooved
bearing

Film
pressure

Length of bearing

Figure 8-21 Comparison of pressure distribution along the length of a journal

bearing with and without a groove.
 437
Section 8-12: Methods of Lubrication

Other types of pressurized oil supply bearings are the axial groove and
inlet hole. In the inlet hole type, the oil is introduced under pressure by means
of a single hole in the bearing, whereas the axial groove is similar to the
circumferential groove type except that the groove is placed 90 deg away, or in
other words, parallel to the axis of the bearing. There are many types that
involve combinations of axial and circumferential grooves as well as a wide
variety of groove types.

SECTION 8-12
Methods of Lubrication
There are a variety of methods used to lubricate bearings. The method
chosen for a specific problem depends to a large extent upon the type of
service the bearing is to perform and also how important the bearing is to
the system of which it is a part.
Bearings that are to be used in low speed, light load applications can be
lubricated by the hand oiling method. It is not a very desirable method
because it tends to result in an excessive supply of oil, with the resultant
leakage problem, at the time of application and later, before the next oiling,
there may not be sufficient lubricant present to maintain the oil film. A good
general rule to follow is that hand oiling should be used only if no other
alternative is available, and then only if the bearing plays a relatively minor
role in the operation of the system it is a part of.
A wick-feed oiler, as the name implies, depends upon an absorbant
material acting as a wick to supply the oil to the bearing. While one end
of the wick is immersed in a reservoir of oil, the other end is in contact with
the journal. The oil is thus drawn from the reservoir, through the wick, to
the bearing. The flow rate can be controlled because it will depend on the
kind of material and construction of the wick. The wick-feed oil method is
better than the hand oiling method because of the more uniform supply
of oil.
A drop-feed oiler allows oil from a reservoir to flow through a needle
valve to the bearing. Clearly, the valve can be adjusted to control the amount
of oil flow. While this method is also better than hand-oiling because of a
more uniform rate of flow, it does have the danger of the valve clogging,
if any impurities are present in the oil.
The ring-oiled bearing makes use of a ring or chain that is placed over the
journal near or at the center of the bearing. As the ring rotates, it dips into
a reservoir and carries the oil to the top of the bearing. The method is
quite reliable, and is widely used in medium to relatively high speed applica-
tions.
When a machine has a rotating part such as a crank or gear enclosed in a
housing, the splash system of lubrication is very useful. The moving part
runs through a reservoir of oil formed in the enclosed casing. This causes a
 438
Chapter 8: Journal Bearings and Lubrication

spray of oil to drench the casing, thus lubricating the bearing. It is also
possible to have channels and ducts that lead directly to the bearing. The
oil, in this system, is sprayed into these channels. As long as the proper oil
level is maintained, a continuous supply of oil is ensured.
In pressure fed systems, a continuous supply of oil is furnished to the
bearing by a pump. This method was discussed earlier, and it will be remem-
bered that oil is returned to a reservoir after being circulated through the
bearing. As a result, if sufficient flow is available, much of the heat developed
in the bearing is carried away by the oil. The pressure lubricated system is
the most extensively used method.
In those applications involving heavy loads, low speeds, high temperature,
and designs for which it would be difficult to retain oil, grease is recommended
as a lubricant. The grease must usually be applied under pressure to the
surface to be lubricated. An example everyone is familiar with is the use
of grease to lubricate the chassis of an automobile.

SECTION 8-13
Bearing Materials

The operating conditions for bearings are such that rather stringent
requirements must be set forth in regard to the material to be used. Some
of the properties a good bearing material should possess are mentioned in
the following list.
1. The material should have good compatibility, which means good
antiweld and antiscoring properties. Since most journal bearings will
experience occasional metal-to-metal contact, it is important that the
bearing material be one that does not readily weld itself to the shaft
material.
2. Conformability is a measure of the bearing’s ability to adjust to
misalignment or other geometric errors. Materials having a low
modulus of elasticity usually have good conformability.
3. Embeddability, means the material is soft enough to absorb foreign
materials that are too large to pass through the oil film. By allowing
this embedding, wear and scoring can be prevented.
4. The material should have good fatigue strength because of the many
applications for which the bearing loads change directions.
5. Resistance to corrosion is required in those applications where the
lubricating oil used may oxidize and thus corrode the bearing.
6. Clearly, from the discussion in Section 8-11, good heat conductance
is desirable.
7. Last, but certainly not least important, the cost should be as low as
possible.
 439
Section 8-13: Bearing Materials

Let us now consider some of the more important types of materials.

1. Babbitt alloys, of either tin or lead base, are perhaps the most commonly
used material. They have excellent embeddability and compatibility
characteristics. However, babbitt alloys have the disadvantage of being
relatively weaker than other bearing materials, lose strength rapidly with
increasing temperature, and have lower fatigue strength. These short-
comings can generally be remedied by using a thin internal babbitt surface
on steel backing.
2. Copper alloys are also widely used and are primarily bronzes. They
are generally stronger, have a greater load capacity, have good fatigue
strength, are harder, but less score resistant than babbitt bearings.
3. Aluminum alloys, because of their fatigue strength, load carrying
capacity, thermal conductivity, corrosion resistance, and low cost, are widely
used for bearings in internal combustion engines. Their disadvantages are that
they require hardened journals, and have poor compatibility, conformability,
and embeddability. When a thin layer of babbitt is placed inside an aluminum
bearing, the antiscoring and embeddability characteristics are improved.
4. Porous metal bearings of either bronze or iron have found wide
acceptance. These sintered-metal, self-lubricated bearings have inter-
connecting pores in which oil is stored and then fed to the bearing surface.
The low cost and long time between lubricant supply are their prime
advantages.
5. Silver bearings are extremely useful for heavy duty applications. The
bearings are made by electrodepositing silver on a steel backing with an
overlay of lead 0.001—0.005 in. thick. A thin layer of indium is then deposited
on the lead to provide corrosion protection.
6. Cast iron and steel are used in bearings subjected to relatively light
loads. The prime advantage of these bearings is low cost. However, because
of poor embeddability and conformability, these bearings require very good
alignment and freedom from foreign matter.
7. Various plastics,such as phenolics, nylon, fluorocarbons, acetal, and
so on, are used as bearing materials. The advantages of using plastics are
no corrosion, quiet operation, moldability, and excellent compatibility.
The last named characteristic usually means that no lubrication is required.
However, when used in heavy load applications, their low heat conductivity
may result in buildup of heat which will then require external cooling.
8. Rubber and other elastomeric materials are used in such applications
as ship propellor shafts and rudders, hydraulic pumps, sand dredges, and
so on. They provide vibration isolation, can compensate for misalignment,
and have good conformability. They are generally made fluted and placed
inside a noncorrodible metal shell. A water flush will thus permit sand and
grit to pass through the bearing without scoring it.
9. Other materials, which are used in specialty applications, are wood
(lignum vitae and oil-impregnated oak), ceramics and cermets, sapphire, and
glass.
 440
Chapter 8: Journal Bearings and Lubrication

SECTION 8-14

Design Considerations
When a bearing is still in the design stage, the effects of varying dimensions
and parameters should be investigated. A general design procedure would
involve choosing values for those quantities which are under the control
of the designer or which are given. These quantities are usually designated
as independent variables and include the viscosity of the lubricant, load,
speed of rotation, radius and length of bearing, clearance, and whether it is
to be a full or partial bearing. These values are then used to determine the
so-called dependent variables such as minimum film thickness, temperature
rise, coefficient of friction, and oil flow. We will now briefly discuss these
quantities as well as some of the more important design parameters.
1. Lubricants. These can be either solid, liquid, or gas. Designers usually
base their choice upon such factors as type of machine, method of lubricant
supply, and load characteristics.
2. Bearing Load. The load acting on a bearing is usually specified, But
by choosing the length and diameter of the bearing, the designer can choose
the value of the load per projected area. The desired life of the bearing
will usually dictate the magnitude of the load per projected area. Clearly,
the smaller the load per projected area, the greater the bearing life.
3. Length/Diameter Ratio. The L/D or length-to-diameter ratio is an
extremely important bearing parameter. The usual practice is to choose
L/D values from 0.8 to 1.5, with a value of unity most generally used. As a
guide in choosing the magnitude of an L/D ratio the following factors
should be considered.
Bearings with L/D ratios greater than | (long bearing) are usually used
in applications where misalignment must be avoided and a reduced load
carrying capacity can be tolerated. When the L/D ratio is less than | (short
bearing), the danger of metal-to-metal contact because of large shaft
deflections is greatly reduced. There are other factors to be considered, but
a rough rule to be followed might be as follows: use an L/D of unity, decrease
it if deflections are expected to be severe, increase it if shaft alignment is
important.
4. Clearance. Proper values to be used for journal bearing clearance
depend on factors such as materials, manufacturing accuracy, load carrying
capacity, minimum film clearance, oil flow, film temperature, and so on.
Clearly a complete analysis is needed to obtain a proper value for the
clearance. If a reasonable value is desired with a minimum of analysis, a
clearance ratio, c/r, of 0.001 has been successfully used for many years.
Clearance value can be obtained by optimization techniques. The problem
is that large clearances will permit foreign materials to pass easily through
the bearings, the increased flow will reduce film temperature and thus
increase bearing life. But too large a clearance will result in a loose, noisy
bearing and a resulting decrease in minimum film thickness.
 44]
Problems

PROBLEMS
1. Two storage tanks are connected by a horizontal capillary tube. The liquid stored
in the tanks is an SAE 10 oil at a temperature of 70°F. The tube has a bore of 0.03 in.
and is 90 in. long. If the difference in pressure in the two tanks (at the points where
the tube is connected to the tanks) is 10 psi, what is the flow in gallons per minute
if the flow is laminar?
2. What will be the absolute viscosity in reyns and the kinematic viscosity in SUS for
a fluid having a viscosity of 30 cs at 90°F and a specific gravity of 0.9 at 60°F.
3. What is the viscosity index of an oil with an SUS of 2900 at 100°F and 140 SUS at
DNOwER
4. A lightly loaded (Petroff’s equation may be used) journal bearing is 3 in. long, has
a 3 in. diameter, is acted on by a 400-lb radial load, the radial clearance is 0.003 in.,
is rotating at 30,000 rpm and is supplied with a lubricant having a viscosity of
0.5 x 10°° reyns. Determine (a) the frictional torque developed, (b) the frictional
horsepower, and (c) the coefficient of friction.
5. A lightly loaded full journal bearing has a 5-in. length, a 5-in. diameter, a 0.001 5-in.
radial clearance, a rotative speed of 500 rpm, and a radial load of 1500 lb. For a
coefficient of friction of 0.01, determine the average viscosity of the oil. If the average
film temperature is 155°F, what is the approximate grade ofoil being used?
6. A hydrostatic step bearing, similar to the one shown in Section 8-8, has a thrust
load of 100,000Ib applied to it. The bearing dimensions are as follows: journal
radius = 10in., recess radius = 5in., viscosity of lubricant = 2.5 x 107° reyns,
minimum film thickness = 0.004 in. What will be the vlaue of the supply pressure
required, and what is the required flow? The load carrying capacity is given by
ye
aoe Bote
2 | In(R,/R;)
and the required flow is given by
rth?
Q=_*
6u In (R,/R;)
7. The following problem is presented as a means of introducing an elementary
concept of hydrostatic lift.
A 180-deg partial journal bearing is found to require a hydrostatic lift. The
journal, of length 5 in., has a 2.500-in. radius, a bearing radius of 2.508-in., and a
5000-!b load applied to it. Oil having a viscosity of 20 x 10~° reyns is supplied to
a 3-in. long rectangular groove in the bearing. What will be the required supply
pressure in order to assure a minimum film thickness of 0.003 in.?
The load carrying capacity equation is given by

_ Qbpe*r7[2 + 32 — e°]
Ww ax(l — €2)

and the flow equation is

p,ba® |e&(4 — &?) 2+ 6 OEP \|

= tan» 5 {$=
@ SAGA =e) i (1 = 27)? i e2
where a = eccentricity; ¢ = eccentricy ratio = a/(r, — rj); b = length of oil slot in
bearing.
 442
Chapter 8: Journal Bearings and Lubrication

8. A journal bearing has the following specifications: journal radius = +in.; radial
clearance = 0.0005 in.;bearing length = 1 in.; viscosity of oil = 15 x 107’ reyns;
load acting on bearing = 1000 Ib. If the load on the bearing is stopped for 0.1 sec
and the minimum film thickness is to be at least 0.00045 in., will a hydrostatic lift
be needed before the load is again applied?
The equation for determining the time in seconds, for the eccentricity ratio to go
from a value of €, to €, is
1/2 € £2
24ubr? 1
Ae ee P|tan ie
We? l-—e _ =|
where b = bearing length; and c = radial clearance.
. A 360-deg journal bearing with a diameter of 2 in. and a length of2 in. consumes
0.116 hp in friction at an operating speed of 1432 rpm. SAE 10 oil is used, and the
film temperature is 173°F. The radial clearance of the bearing equals 0.001 in.
Determine (a) the eccentricity ratio. (b) The minimum film thickness. (c) The total
load for the bearing. :
10. To demonstrate the use of the Boyd and Raimondi design charts, determine the
quantities asked for in regard to the following journal bearing problem. A 180-deg
partial journal bearing has an L/D ratio of 1, a 4-in. diameter, a radial clearance of
0.003 in., an operating speed of 600 rpm and an applied load of 1500 lb. If the film
thickness is to be no less than 0.0009 in., find (a) required viscosity, (b) frictional
force developed, (c) amount of oil flow through bearing, (d) the end flow, (e) tem-
perature rise of the oil, (f) the heat developed.
11. A full journal bearing having a 5-in. diameter, a 5-in. length, and a radial clearance
of 0.001 in. is subjected to a radial load of 2500 lb. When the rotative speed is
200 rpm, a frictional force of 15 lb is developed. If the average film temperature is
maintained at 160°F, what will be the viscosity of the oil? (a) Using Petroff’s
equation (b) using the design charts. .
12. In order to gain more insight into the differences between full and partial bearings,
consider the following problem. A journal bearing has an L/D ratio of 1, a 3-in.
diameter, an oil viscosity of 4 x 107° reyns, and an r/c ratio of 1000. A minimum
film thickness of 0.0006 in. is to be maintained and the rotative speed is 3000 rpm.
Determine the following quantities for a 360-deg bearing. (a) Flow entering in
cubic inches per second. (b) End leakage in cubic inches per second. (c) Load.
(d) Coefficient of friction. (e) Frictional horsepower. (f) Temperature rise of the oil.

13. Same as Problem 12 for a 180 deg bearing.

14. Same as Problem 12 for a 120 deg bearing.
15. Same as Problem 12 for a 60 deg bearing.
16. A 4000 Ib load is applied to a 180 deg partial bearing having a 2-in. diameter and
length, and a 0.001-in. radial clearance. The journal is rotating at 4000 rpm, and uses
an SAE 30 oil with an inlet temperature of 100°F. What will be the average film
temperature?
17. An oil bath journal bearing rotating at 1000 rpm in an ambient air temperature of
90°F is subjected to a load of 6000 Ib. The 180 deg partial bearing has an L/D ratio
of 1, a journal radius of 2in., a bearing radius of 2.006 in. and a minimum film
thickness of 0.0024 in. Assume pillow-blocks with separate shells and a surround-
ing air velocity of 500 ft/min. Determine (a) the viscosity (b) the frictional loss
(c) the proper SAE grade of oil to use, and (d) the increase in temperature of the
oil between entrance and exit from the bearing.
 443
References

18. A full journal bearing is rotating at 500 rpm, and is supporting a load of 4000 lb.
The L/D ratio is 1, the journal radius 3 in., and the bearing radius is 3.006 in. A
minimum film thickness of 0.0024 in. is to be maintained. Oil is supplied to the
bearing by means of a circumferential groove at the center of the bearing at a supply
pressure of 60 psi gage. What is the average temperature rise of the oil?
19. A 360 deg journal bearing with a diameter of 4 in. and a length of 4 in. consumes
0.2 hp in friction at an operating speed of 1500 rpm. SAE 20 oil is used, and the film
temperature is 175°F. The radial clearance of the bearing equals 0.0015 in.
Determine: (a) The eccentricity ratio. (b) The minimum film thickness. (c) The total
load for the bearing.

REFERENCES
[1] E. W. Dean and G. H. B. Davis: Viscosity variation of oils with temperature.
Chem. Met. Eng., 36: 618-619 (1929).
[2] D. D. Fuller: Theory and practice of lubrication for engineers. John Wiley & Sons,
New York, 1956.
[3] O. Reynolds: On the theory of lubrication and its application to Mr. Beauchamp
Tower’s experiments. Phil. Trans. Roy. Soc. (London), 177: 157—234 (1886).
[4] A. A. Raimondi and J. Boyd: A solution for the finite journal bearing and its
application to analysis and design, Parts I, II, and III. Trans. ASLE, 1(1): 159-209
(1958).
[5] A. Sommerfeld: Zur Hydrodynamischen Theorie der Schmiermittelreibung, 2.
Math. Phys., 50: 97-155 (1904).
[6] J. Boyd and A. A. Raimondi: Bearing theory in analysis and design of journal
bearings. J. Appl. Mech., 73: 298-316 (1951).
[7] G. B. Karelitz: Performance of ring bearings. Trans. ASME, J2: 57—70 (1930).
[8] V. M. Faires: Design of Machine Elements. The Macmillan Company, New York,
1965.
[9] Seals Handbook, Machine Design. Penton, Cleveland, Ohio, 45 (1973).
[10] R. R. Slaymaker: Bearing Lubrication Analysis. John Wiley & Sons, New York,
LOSS:
[11] M. C. Shaw and E. F. Macks: Analysis and Lubrication of Bearings. McGraw-Hill
Book Co., 1949.
[12] The Bearings Book, Machine Design Penton, Cleveland, Ohio, 44 (1972).
[13] E. I. Radzimovsky: Lubrication of Bearings. The Ronald Press, New York, 1959.
[14] P. R. Trumpler: Design of Film Bearings. The Macmillan Company, New York,
1966.
Rolling Bearings
SYMBOLS

A = a value depending on the =e =| horsepower

ra"
bearing type (see Table 9-11) ~ number of rows of balls
a, = life adjustment reliability factor = rating life in millions of
a, = life adjustment material factor revolutions (based on 10%
a3 = life adjustment application failure)
factor rpm during the mth fraction of
b = an exponent a cycle
C = basic dynamic load rating, lb = number of balls
Co = basic static load rating, Ib shaft speed, rpm
D = bearing outer diameter, mm or equivalent load, Ib
in. = equivalent load during the mth
x |= bearing bore, mm or in. fraction of a cycle
d,, = mean bearing diameter, mm mth fraction of a cycle
e = aconstant (see Tables 9-5, 9-6, viscosity in Saybolt Universal
or 9-7) Seconds
F,, = thrust load, lb = frictional torque, Ib-in.
F, = radial load, Ib = a factor determined by whether
F, = bearing service factor the inner or outer ring is
jf, = a coefficient depending on rotating
bearing size (see Figure 9-47) a radial load factor
fo = a coefficient depending on a thrust load factor
relative bearing load (see coefficient of friction
Figure 9-48)
H = bearing height for a thrust
bearing, mm

The advent of the automobile, high speed engines, and automatic produc-
tion machinery provided the impetus for extensive research and development
of the rolling bearing (also called the antifriction bearing) see Figure 9-1.
Asa result, the Anti-Friction Bearing Manufacturers Association (AFBMA)
standardized bearing dimensions and the basis for their selection. Thus, it is

444
 445
Chapter 9: Rolling Bearings

| Snap ring
| groove
ie =
S5
Corner radius
Seal or shield
athe Shoulders

Seal or shield
notch

Corner radius

Inner ring
raceway
Outside
diameter Inner ring

Inner ring
land

Separator
Inner ring
face Outer ring raceway

Outer ring land

Outer ring
face Outer ring

Figure 9-1 Ball bearing nomenclature. [Courtesy New Departure-Hyatt

Bearings Division, General Motors Corporation. ]

possible for a designer to select a bearing from the catalog of one manu-
facturer and successfully replace it with a bearing having identical dimensions
from a different manufacturer. The designer will also find that, although
the AFBMA has standardized the method for selecting a bearing based on
load and life requirements, not all manufacturers interpret these standards
uniformly. However, the catalogs contain sufficient explanation so that one
can always relate his results to equivalent AFBMA values.
Before deciding to use rolling bearings, the designer should give careful
consideration to their relative advantages and disadvantages in comparison
to journal bearings.

Advantages of rolling bearings compared to journal bearings

1. Starting friction torque is low—not much greater than the running
frictional torque (see Section 9.7).
2. Ease of lubrication either with prepacked grease or with relatively
simple oil systems.
3. Less axial space for a comparable shaft diameter.
 446
Chapter 9: Rolling Bearings

4. With the exception of straight rollers, capable of supporting both

radial and thrust loads. (Journal bearings can support only a radial
load, if a thrust bearing is not also provided.)
. Early warning of impeding failure signalled by increasing noisiness
at the same speed of rotation. (Journal bearings do not provide such
a signal and can suddenly fail.)
Readily replaceable as stock items (except for some special applica-
tions).
Standardization and the employment of close tolerances make
preferable their use in the satisfactory operation of cams and gears.
. Can be preloaded causing a shaft to become stiffer—very important
in machine tool applications.
9 Can be used for mounting a shaft placed in any position in space.
10. Wide versatility with respect to mounting because they are supplied
in special housings (for example, pillow-blocks, flanges, and so on).
Disadvantages of rolling bearings compared to journal bearings
HE Greater diametral space required for a comparable shaft diameter.
2s Initial cost is usually higher.
33 Noisier in normal operation.
4. Dirt, metal chips, and so on, entering the bearings can limit their
life causing early failure. (Journal bearings do not suffer from this
malady because foreign matter is either washed away by the lubricant
or becomes embedded in the softer bearing material.)
. Finite life due to eventual failure by fatigue. (Journal bearings,
properly maintained, can run “‘forever.’’)
6. Lesser capacity to withstand shock.

The stresses developed between the ball(s) and the races or the roller(s)
and the races are very complex and were first studied by Stribeck [1] who
based his analysis on the earlier work of H. Hertz [2] (that is, contact stresses).
~ Space limitations preclude further discussion here, but the interested reader
is directed to references [3], [4], and [5] for a mathematical treatment of the
stress in rolling bearings.
In this chapter, we shall describe the most common types of bearings and
explain the procedure for selecting a bearing based upon load and life
requirements. In addition, we shall also discuss and illustrate some important
“schemes” for mounting bearings as well as their methods of lubrication.

SECTION 9-1

Bearing Nomenclature and Types of Ball Bearings

Figure 9-1 is an illustration naming the various parts, surfaces, and
edges of a ball bearing. Note that the basic bearing consists of four parts:
(1) the outer ring, (2) the inner ring, (3) the balls, and (4) the separator
 447
Section 9-1: Bearing Nomenclature and Types of Ball Bearings

One Two One Two Shield and Snap

shield shields seal seals seal ring

pak kaka ks
Snap ring
and
Snap ring
and
Snap ring
and
Snap ring
and
Snap ring
shield
one shield two shields one seal two seals and seal

Figure 9-2 Bearings with seals and shields. [Courtesy New Departure-Hyatt
Bearings Division, General Motors Corporation. ]

(that is, ball retainer). The separator serves the purpose of always keeping
the balls separated and thereby preventing them from rubbing against
each other. Some special ball (and also needle) bearings do not have an inner
ring. Instead, the balls are permitted to “ride” in a race directly cut in a
hardened shaft.
Bearings are available with (1) shields to prevent dirt from entering and
to retain grease, (2) shields and seals to contain lubricant for self-lubrication,
and (3) snap rings and flanges that provide for simple bearing containment
(see Figure 9-2).
One of the important aspects to successful bearing design is the conformity
of the ball radius to the raceway radius. Figure 9-3 is a photograph illustrating

Figure 9-3 Conformity of ball radius

to that of the raceway. [Courtesy New
Departure-Hyatt Bearings Division,
General Motors Corporation. ]
 448
Chapter 9: Rolling Bearings

what is meant by aaa Thecael shows anpgs of low

ee eae:

ronnity increases penute Prciion: ane the resultant “fit” of the ball and
raceway curvature becomes a matter of design compromise. Bearing
manufacturers establish their own conformity values based upon their
individual experiences and research data. Because of conformity, and elastic
and plastic deformation of the balls and races when loaded, the balls do
not have pure rolling motion. Instead, a small
which affects both the frictional loss and life of the Beating.
Ball bearings are categorized into three areas, namely, (1) radial ball
bearings, (2) angular-contact ball bearings, and (3) thrust ball bearings.

Radial ball bearings

Figure 9-4 illustrates the various types of radial ball bearings commonly
available. The most widely used radial ball bearing is the Conrad or deep
groove bearing. This bearing, primarily designed to support radial loads,
has deep raceways that are continuous (that is, there are no openings,
recesses, and so on) over all of the ring circumference. This type of construc-
tion permits the bearing also to support relatively high thrust loads in
either direction. In fact, the thrust load capacity is about 70% of the radial
load capacity, as noted in Figure 9-4. The reason that a ball bearing designed
primarily for a radial load can also support a high thrust load is that only
several balls (sometimes only one) carry the radial load, whereas all the
balls act to withstand the thrust load. Figure 9-5 shows, in an exaggerated
manner, how only some of the balls sustain a radial load.
As shown in Figure 9-6, Conrad bearings are assembled in four steps.
Conrad bearings that are intended for high speed operation are usually
made with machined bronze or phenolic-cloth laminate separators.
To increase the radial load carrying capacity of the ball bearing, more
balls are used. This is the concept behind the maximum capacity or filling
notch type of bearing. These bearings have the same basic radial construction
as the Conrad type. However, as shown in Figure 9-4, a filling notch (or
loading groove) permits more balls to be used than a Conrad bearing of
comparable size.
Because the loading groove in each ring interrupts the ball raceway
shoulders, the thrust capacity of the bearing is limited. As noted in Figure 9-4,
the thrust capacity of the filling notch type is only 20% of the thrust capacity
of the Conrad bearing, whereas the radial capacity is 20-40% higher than
the radial bearing.
The magneto or counterbored type bearing is similar to ine deep groove
bearing except that the outer ring has but one shoulder, which makes the
bearing separable. Thus, this type of bearing permits the inner and outer
rings to be mounted separately. Magneto bearings are extra-light, being
 449
Section 9-1: Bearing Nomenclature and Types of Ball Bearings

Figure 9-4 Common types of radial ball bearings and some of their important
characteristics. (1) The approximate range of bore sizes, although stated in inches,
is given in millimeters in manufacturers’ catalogs along with the equivalent inch
dimension. To accommodate American availability of bar stock for shafting,
manufacturers also supply some bearings with bores having inch dimensions.
(2) The columns “Relative Capacity” and “Limiting Speed Factor” are based on
the value of unity for the Conrad or deep groove bearing since this type of bearing
is the one most widely selected by designers. [From Machine Design, 1970 Bearings
Reference Issue). The Penton Publishing Co., Cleveland, Ohio.]

=| Type Approx Range

of Bore Sizes (in.) Relative Capacity
Limiting
Speed
Tolerance
to
F Misali
Min Max Radial Thrust actor NEUEN
or
Conrad
O4g!
or 0.1181 41.732 1,00 0.7 1.0 +0°15
deep (2-direction)
groove

Maximum
capacity Se;
or 0.3937 5.1181 1,2-1.4 0,2 1.0 +0°3
filling (2-direction)
notch
ee
Magneto
,
or 0.1181 1.181 0,9-1.3 0.5-0.9 1.0 +0°5
counterbored 0.3937 7.874 (1-direction)
outer

Airframe
or High static 0.5 °
aircraft 1200 1250 capacity (2-direction) we y
control

Selfe 0.2 Rell

aligning 0.1969 4.7244 0.7 (1-direction) 1.0 +2°30 |
internal

Self-
aligning, — em 1.0 07 1.0 High
external (2-direction)

Double row, 0.3937 4.3307 15 0.2 1.0 +0°3'

maximum (2-direction)

Double row,
deep 0.3937 4.3307 15 1.4 1.0 0°
groove (2-direction)

supplied to a bore size of only 1.181 in. Counterbored bearings, however,

are available in much larger bore sizes (see Figure 9-4). It is apparent that
these bearings can only support unidirectional thrust loads. Counterbored
types are widely employed in fractional horsepower motors.
 450
Chapter 9: Rolling Bearings

(a) (b)
Figure 9-5 (a) Radial displacement in a ball bearing when carrying a radial
load on the inner ring. [Courtesy of Link-Belt Division, FMC Corp.]
(b) Nonfluctuating load distribution within a bearing. [Courtesy SKF
Industries, Inc.]

Airframe or aircraft control bearings are designed to withstand heavy

radial loads in oscillating or slow turning applications. These bearings are
often supplied with a full complement of balls to provide high load capacity.
However, where torque requirements are not too severe and the operation
is smooth, the bearing can be obtained with ball separators. The bearings
are grease-packed and sealed with special synthetic rubber to withstand
adverse environmental conditions. The external surfaces are usually cadmium
plated to resist corrosion, depending upon specifications (usually military).
Airframe bearings are also available in double row, extra-wide, self-
aligning, and rod end types.
Self-aligning bearings are intended to compensate for misalignments
arising from shaft deflection, frame distortion, or foundation deflection.
They are available in two types: self-aligning internal and self-aligning

Figure 9-6 Steps in the assembly of the Conrad or deep groove type ball bearing.
[Courtesy of New Departure-Hyatt Bearings Division, General Motors
Corporation.]
 45]
Section 9-1: Bearing Nomenclature and Types of Bali Bearings

external. The basic difference between the two types is made clear by inspec-
tion of Figure 9-4. Because of conformity between the balls and the outer
raceway, higher contact stresses are produced in the internal type of self-
aligning bearing. The external self-aligning bearing does not suffer from
this handicap because alignment is achieved by grinding a spherical surface
on the outside diameter of the outer ring and matching it to a similarly
ground surface of a housing. The penalty paid in using the external type of
bearing is the greater mounting space required in the radial direction.
Double row ball bearings are available as double row, deep groove bearings
and as double row maximum ball bearings. The latter type of bearing is of
the filling notch type, whereas the former bearing is of the Conrad or deep
groove type.
The double row maximum bearing possesses a larger ball capacity than
the deep groove type, giving it good radial load capacity at the expense of
thrust load capacity. Note from Figure 9-4 that the lines of contact converge
with the envelope of the bearing.' This type of geometry reduces the resistance
of the bearing to the angular deflection, and therefore it should not be used
in a design requiring a single bearing mounting. In fact, this bearing is
particularly suited for applications having heavy loads and where slight
misalignments are expected (for example, gear reducers).
The double row deep groove bearing can be used not only with high radial
loads, but also with high thrust loads. Figure 9-4 shows the lines of contact
for this type of bearing converging outside of the bearing envelope. This
type of geometry produces a rigid bearing. Thus, the double row deep groove
bearing has a propensity for effectively resisting axial, radial, and over-
turning deflections. Consequently, these bearings are often used singly in
mounting idler pulleys, gears, wheels, and so on. As indicated in Figure 9-4,
the permissible misalignment tolerance is 0 deg, and they cannot be used
where angular shaft deflection is permitted.
Both types of bearings can be supplied with a light, medium, or heavy
preload, thereby providing a wide range of loading and stiffness.

Angular-contact ball bearings

A second category of ball bearings is the angular-contact ball bearing.
Figure 9-7 illustrates the various types of angular-contact bearings available,
their range of bore sizes, relative radial and thrust load capacities, limiting
speed factor, and misalignment tolerance.
The primary type of angular-contact bearing is the one-directional thrust
type, which has a single row of balls and is so designed that the centerline
of contact between the balls and the raceway is at an angle to a plane per-
pendicular to the axis of rotation. This angle, called the contact angle, is
shown in Figure 9-8.

' Theoretically, the lines of contact converge on the shaft center line.
 452
Chapter 9: Rolling Bearings

P ‘i are
eee Tolerance
i
Type Approx Range Relative Capacity
of Bore Sizes (in.) - Dee nee
Min Max Radial Thrust Eaton Misalignment

One-directional 0.3937 12.5984 1,00-1,15* 1.5-2,.3" 1.1-3,0" BO

thrust (1-direction)

Duplex,
back-
0.3937 12,5984 1.85 oie AS” 3.0 0°
(2-direction)
to-back

Duplex, 1.5 °
face- 0.3937 12.5984 1.85 (direction) 3.0 ie)
to-face

24 °
Duplex, 0.3937 12,5984 1.85 (ediection) 3.0 0
tandem

Two directional 1.5

0.3937 4.3307 +0°2'
or 1.15 (2-direction) se
split ring

1.85 3
=
Double
0.3937 5.5118 1.5 (2-direction) 0.8 S

0.5
Double (in 1 direction) 3
row, 0.3937 4.3307 1.65 15 0.7 0
maximum (in other direction)

*Depends on contact angle.

Figure 9-7 Typical forms of angular-contact bearings and some oftheir

important characteristics. [From Machine Design, 1970 Bearing Reference Issue.
The Penton Publishing Co., Cleveland, Ohio.]

The reader can observe from Figure 9-8 that the outer ring has one “heavy”
raceway shoulder and another raceway shoulder which has been removed by
counterboring. The counterbore is cut to within a few thousandths of the
bottom of the raceway. The bearing is assembled by thermally expanding the
Thrust face

Figure 9-8 Section through an

angular-contact bearing illustrating the
contact angle. [Courtesy New Departure-
Hyatt Bearings Division, General Motors
Corporation. ]

Contact
angle
 453
Section 9-1: Bearing Nomenclature and Types of Ball Bearings

outer ring so that the inner ring, a full complement of balls, and the ball
separator are “snapped” into place. Upon cooling the outer ring, the bearing
assembly becomes nonseparable, being held together by the small “‘snap”’
shoulder formed by the counterboring operation. This design, which uses a
full complement of balls, permits the bearing to carry a higher radial load
than a comparable deep groove bearing and a high unidirectional thrust
load.
These bearings are made with an initial internal looseness which is
necessary to obtain angular contact. They become rigid when loaded
axially and have contact angles ranging between 15 and 40 deg.” For example,
New Departure-Hyatt Bearings supplies angular-contact bearings with 15-,
25-, or 35-deg contact angles. The higher contact angle bearings can support
greater axial loads but at reduced radial loads.
The two-directional or split ring angular-contact bearing was originally
developed for use in high speed turbines. The bearing consists of a solid
one-piece outer ring and a two-piece inner ring. Because the inner ring is
split, it is possible to assemble this bearing with a full complement of balls
for maximum load carrying capacity. The ball separator is of one-piece
construction, usually made of bronze.
This type of bearing can carry a substantial amount of radial load provided
a sufficiently high thrust load is being simultaneously carried. Should a
high thrust load be absent, the balls would slip and rapid bearing deteriora-
tion would follow.
Duplexing of angular-contact ball bearings arises because there is a need
for maintaining very rigid bearing supports that act to minimize both
axial and radial shaft deflections. This need is most obvious in machine
tool applications (for example, lathe headstock spindles, grinding wheel
spindles, boring bar spindles, and so on), instruments, or in any machine
or device where shaft rigidity is a necessity for accurate performance.
Rigidity of the bearing assembly can be achieved by introducing a con-
trolled axial preload. This intentional preload is introduced in one of three
ways: (1) Using a set of matched ball bearings having a specific ring “‘stickout”’

Duplex DB Duplex DF Duplex DT

back-to-back face-to-face tandem

mounting mounting mounting

Figure 9-9 Duplex mounting arrangements. [Courtesy New Departure-Hyatt

Bearings Division, General Motors Corporation. ]

? For high speed applications, the contact angle should not exceed 30 deg.
Figure 9-10 Duplex mounting arrangements
showing face stickouts before assembly.
[Courtesy New Departure-Hyatt Division,
General Motors Corporation. ]

Single bearing showing equal

face stickout on either side.

DB mounting before clamping.

When inner rings are abutted,
a preload corresponding to an
axial deflection of ‘’A” will exist.

DF mounting before clamping.

When outer rings are abutted,
a preload corresponding to axial
deflection of ‘‘A”’ will exist.

DT mounting in which both

bearings share thrust load. DT
Pairs may be preloaded against
single angular contact bearings
or duplex (DT) pairs.

454
 Figure 9-11 DB mounting showing the effective
spread as greater than the distance between the
bearings. [Courtesy New Departure-Hyatt Bearings
Division, General Motors Corporation. ]

/ es \
(Effective spread

Figure 9-12 DF mounting showing the effective

spread as less than the distance between the bearings.
[Courtesy New Departure-Hyatt Bearings Division,
General Motors Corporation. ]

Re Effective spread

Figure 9-13 Maximum resistance to the misalignment of cocking loads is

obtained by utilizing a set of angular-contact bearings in the DB (back-to-back)
style of mounting. The figure illustrates a DB mounting that 1s free to float axially
in the housing. The preload predetermined for the duplex set is applied by locking
the inner rings of the pair together. A DF mounting cannot be similarly floating
because the preload is obtained by clamping the outer rings. Note that the rigidity
of the DB mounting necessitates an accurate alignment of the housing bore with
respect to the spindle and careful control of the squareness of the spindle shoulders.
[Courtesy New Departure-Hyatt Bearings Division, General Motors Corporation. ]

455
 ”A56

Chapter 9: Rolling Bearings

fj Clearance

== TM
ae =
cls

py
> 4

Figure 9-14 A fixed mounting in which the inner rings of the pair of DF
bearings are clamped between the spindle shoulder and a locknut. The
predetermined preload is applied by tightening the housing end cap until the two
outer rings are locked together. The DF mounting tolerates slightly more operating
misalignment than the DB mounting. Whenever optimum results are expected,
misalignment should be held to minimum values. [Courtesy New Departure-Hyatt
Bearings Division, General Motors Corporation. }

on the faces of the inner or outer rings. When the gap between the rings
without ‘“‘stickout” is closed, the bearing assembly receives a controlled
axial preload. This method of introducing a preload is called duplexing.
(2) Using a nut on the shaft or housing which, when turned, causes the rings
of the angular-contact bearing to move relative to one another in an axial
direction. All looseness within the bearing is removed and the axial preload
is determined by the number of turns made by the nut. (3) Using a thrust
washer, shims, or a spring to introduce the relative ring displacement
resulting in an axial load.
The first method (that is, duplexing) is the most reliable because it depends
only upon the dimensional accuracy of the duplex bearing sets that are
used. As in the manufacture of all bearing components, paired sets of
duplex bearings are carefully controlled dimensionally, and so this method
of preloading is the one which is most widely used.
Figure 9-7 shows three ways in which angular-contact ball bearings can
be duplexed. To assist in explaining the three bearing arrangements in
Figure 9-7, we shall refer to Figures 9-9 and 9-10,
As shown in Figures 9-7 and 9-9, duplex bearings may be mounted as
follows: (1) back-to-back (that is, DB), (2) face-to-face (that is, DF) or
(3) in tandem (that is, DT). Figure 9-10 illustrates how “‘stickout” (that is,
 457
Section 9-1: Bearing Nomenclature and Types of Ball Bearings

AY
Vl
Y
MS WW °
I
TF
Ek

XA SAN
Y @ TAX
Oil level AY

Geszs
AL
INNS
A
VAM ZL.
V2
CALIXdICAARK

“WN
LLL:
LLL
LLLL
dr
lille
LLidllel
LL
SEE,
“GEE
EA,
GEA,
WEG,
ESE
|,
=

LL
FLL MA
VL

Figure 9-15 The DT mounting is intended for combined loads with very heavy
thrust in one direction (indicated by arrow). When correctly and accurately applied,
two bearings of a DT set share the applied load. For most applications, it 1s usually
desired to preload the DT set against an opposed angular contact bearing or an
opposed DT set at the other end of the shaft. However, in most vertical mountings
where the applied thrust load is constant in one direction, preloading may not be
necessary. [Courtesy New Departure-Hyatt Bearings Division, General Motors
Corporation. ]

offset) is used to achieve one of the mounting arrangements shown in

Figure 9-9. The amount of “‘stickout” needed depends upon the magnitude
of the axial preload required. As a result, bearing manufacturers supply
duplex pairs for light, medium, or heavy preloads.
The selection of aDB, DF, or DT mounting depends on the rigidity and
thrust load design requirements. In the DB arrangement (see Figure 9-11),
contact lines converge outside of the bearing envelope resulting in an
“effective spread” that is greater than the axial distance between both
bearings. Consequently, this arrangement results in a very rigid assembly
and provides greater resistance to shaft bending moments and deflections.’
Thus, where a designer is aware of unavoidable shaft bending or housing
misalignments that must be tolerated, he should select the DF arrangement
(see Figure 9-12).

3 The reader may readily verify this argument by imagining a loaded shaft in each arrangement
and drawing the free body forces acting on the balls.
 458
Chapter 9: Rolling Bearings

The DT arrangement (see Figure 9-10) is used for applications requiring

the bearings to withstand extremely high thrust loads in one direction
where high speeds or space limitations preclude the use of a larger bearing
or simpler arrangement. For unusually high uniaxial thrust loads, it is
possible to use three (or more) DT bearings. Where a heavy thrust load is
carried in one direction and a lesser, but considerable, reversed load also
exists, then an arrangement using a pair of DT bearings mounted in a DB
arrangement with a single bearing should be used. When severe axial
loads are expected in both directions, two pairs of DT bearings mounted in
a DB arrangement should be used.
Representative arrangements showing the application of DB, DF, and
DT mounted bearings are shown in Figures 9-13, 9-14, 9-15 and 9-16.
The double row angular-contact bearings shown in Figure 9-7 act in the
same manner as two single opposed angular-contact bearings. They are
available with the contact lines converging outside the bearing envelope
(as with DB bearings), making the bearing more rigid, or with the contact
lines converging within the bearing envelope (as with DF bearings), which
allows for minor shaft and bearing support misalignment.
The double row maximum angular-contact bearing is of the filling notch
type and thus has a reduced thrust capacity in one direction but a higher
radial load capacity in comparison to the regular double row bearing.
These bearings are normally manufactured with a built-in preload.

vas:

>
ape = ysa
gin

BE
Me Wi). SSS
N Do
Ss g
Figure9-16 A modified DB mounting at the nose ofa precision spindle. The bearings
of the duplex set are separated by spacers, increasing the spindle’s resistance
to shaft bending moments. Both the inner and outer ring spacers must be exactly
the same length to obtain the predetermined preload for the duplex set of bearings.
[Courtesy New Departure-Hyatt Bearings Division, General Motors Corporation.]
 459
Section 9-1: Bearing Nomenclature and Types of Ball Bearings

Thrust ball bearings

As indicated in Figure 9-17, there are three types of thrust ball bearings,
(1) the one-directional flat race bearing, (2) the one-directional grooved
race bearing, and (3) the two-directional grooved race bearing.
Thrust ball bearings are designed to carry pure thrust loads and, if any
radial load is present, separate radial bearings must be used. From the
viewpoint of both economics and simplicity of design, it is wise to seek the
use of an angular-contact bearing where both radial and thrust loads are
present.
The one-directional flat race thrust ball bearing consists of two ungrooved
washers, balls, and a ball separator. The separator is kept in place by the
shaft, permitting the balls and the washers some slight radial movement
but no radial load support. Although the friction in this bearing is quite small,
its speed is limited by the centrifugal force acting on the balls, which under
high speeds cannot be contained by the separator.
The one-directional grooved race thrust ball bearing is different than the
flat race bearing in one respect, namely, the races are grooved, providing
containment for the balls. Because of this containment, this type of thrust
ball bearing can function under higher thrust loads and speeds than the
flat washer type of bearing. However, the frictional torque of this bearing
is higher than that of the flat washer type. The one-directional grooved
race thrust bearing is also available with one washer that has a convex
spherical seat that mates with a third washer having a concave spherical
seat. The whole bearing assembly thus becomes self-aligning.

Figure 9-17 Typical forms of ball thrust bearings and some of their important
characteristics. [From Machine Design, 1970 Bearings Reference Issue. The Penton
Publishing Co., Cleveland, Ohio.]

Approx Range Relative Capacity Limiting Tolerance

Type of Bore Sizes (in.) Radial Thrust Speed to
Min Max Factor Misalignment

One-directional, 0.254 3.500 0 10s 0.10 0°

flat race (1-direction)
(accepts eccen-
tricity)

One-directional, 0.254 46.4567 0 1.5 0.30 0°

Eraciediee (1-direction)

Two-directional 0.5906 8.6614 0 15 f 0.30 O°

grooved race (2-direction)
 460
Chapter 9: Rolling Bearings

The two-directional grooved race thrust ball bearing has two separators
and a middle grooved race that rotates with the balls. This bearing possesses
the same characteristics as the one-directional grooved bearing but can
withstand axial loads in two directions. These bearings can also be furnished
with spherically dished alignment washers.

SECTION 9-2
Roller Bearings
Roller bearings serve the same purpose as ball bearings, but they can
support much higher loads than comparably sized ball bearings because
they have line contact instead of point contact. Most types of radial roller
bearings cannot resist thrust loads of any significant magnitude and, with
the exception of the cylindrical type, operate at speeds lower than those for
ball bearings.* Roller bearings are also capable of withstanding moderate
to heavy shock loads depending on the roller size.
It is important to note that roller bearings are not supplied in sealed,
self-lubricated form as are the bearings of the Conrad type. Thus, careful
consideration must be given to their lubrication, although many can function
adequately under limited speed and load combinations with only a periodic
greasing. The manufacturer should be consulted before adopting this mode
of lubrication.
Roller bearings can be classified into four basic types: (1) cylindrical
roller bearings, (2) needle roller bearings, (3) tapered roller bearings, and
(4) spherical roller bearings. Figures 9-18, 9-19, 9-20 and 9-25 respectively
illustrate the different forms that exist in each class as well as their bore
sizes and other important characteristics.

Cylindrical roller bearings

Cylindrical roller bearings come in a variety of forms as indicated in
Figure 9-18. They are available in a wide range of bore sizes, and function
with rollers having length-to-diameter ratios from 1:1 to 3: 1. The outside
diameter of the roller is often crowned to increase the load carrying capacity
by eliminating any edge loading. The variety of types available permits the
consideration of a wide range of shaft and housing designs.
When the loads are high and a reduction in the diametral size of the
bearing is desirable, it is possible to permit the rollers to operate directly
on the shaft journal.° In these applications, it is imperative that high loads
be present to prevent the rollers from slipping in the races, which can lead
to early failure.
* Cylindrical roller bearings with roller length-to-diameter ratios of 1:1 can operate at the same
speeds as comparably sized ball bearings.
* The journal must be properly hardened (e.g., a minimum of Rockwell C58), and the reader is
advised to seek the counsel of the manufacturer before proceeding with this design.
 Figure 9-18 Typical forms of cylindrical roller bearings and some of their
important characteristics. [From Machine Design, 1970 Bearings Reference Issue.
The Penton Publishing Co., Cleveland, Ohio.]

Approx. range Relative capacity Limiting Tolerance

Type of bore sizes (in.) Radial Thrust speed to
Min. Max. factor misaligment

Separable
outer ring,
0.3937 12.5984 1°55 ¢) 1.20 +0°5
nonlocating
RN, RIN

Separable
inner ring,
nonlocating
0.4724 19.6850 1.55 0 1.20 +0°5'
RU, RIU

Separable
outer-ring,
Locating ocr
one-direction 1.5748 7.000 1.55 (itdirection| (Ru) +0°5
locating
RF, RIF

Separable
inner ring,
Locating Orr
one-direction 0.4724 12.5984 1.55 1,15 +0°5
(1-direction)
locating
FJ,RIJ

Self-
contained Locating
two-direction
0.4724 3.9370 1.35 (2-direction)
locating

Separable
inner ring,
Locating
two-direction 0.7874 12.5984 1.55 Oldrection)
locating
RT, RIT

Nonlocating,
full
0.6693 2.9528 2.10 0) 0.20 +0°5’
complement
RK, RIK

Double-row,
separable
outer ring, 1.1811 41.7323 1.85 0 1.00 0°
nonlocating
RD

Double row,
separable
inner ring,
2.7559 41.7323 1.85 0) 1.00 0°
nonlocating

461
 462
Chapter 9: Rolling Bearings

Needle roller bearings

Needle roller bearings are like cylindrical roller bearings in that they
can withstand high radial loads, but are different in that their rollers (called
needles) have a much greater length-to-diameter ratio. Also, needle bearings
have a much smaller diametral silhouette than cylindrical roller bearings.
It can be observed from Figure 9-19 that there are two basic forms of
needle bearings. In one form, the needles are not separated by a roller
cage, and in the other form the needles are separated by a cage. The bearing
that does not have the roller separator has a full complement of rollers and,
therefore, can handle a higher load than the bearing having a roller separator.
However, the latter type of needle bearing is capable of operating at much
higher speeds because the needles are kept from rubbing against one another
(causing skidding) by the separator.
Both forms of needle bearings can be mounted with the needles in direct
contact with the shaft, making for a rather small diametral bearing. As
with the cylindrical roller bearing, when the rollers (that is, needles) are in
direct contact with the shaft, the shaft should be heat treated to a minimum
hardness of Rockwell C 58.
Figure 9-19 Typical forms of needle roller bearings and some of their important
characteristics. [From Machine Design, 1970 Bearings Reference Issue. The Penton
Publishing Co., Cleveland, Ohio.]

Bore Sizes Relative Limiting

(in.) Load Capacity Speed Misalignment
Type Min Max Dynamic State Factor Tolerance

< Sea | ea 0.125 7.250 High Moderate 0.3 Low

needle |

Open end f Closed end

Drawn
cup,
needle, 0.156 1.000 High Moderate 0.3 Low
grease
retained

Drawn
Caer 0.187 2.750 Moderate Moderate 0.9 Moderate
roller

Heavy-
duty
roller 0.625 9,250 Very high Moderate 1.0 Moderate

Caged
roller 0.500 4.000 Very high High 1.0 Moderate

Cam M Moder
follower 0.500 6,000 WEUEUE
tohigh
oe
to high
0.3-0.9 Low

Needle
thrust 0.252 4.127 Very high Very high 0.7 Low
 463
Section 9-2: Roller Bearings

Needle bearings are mainly lubricated by grease. For high load or high
speed application, some form of oil lubrication is required if the bearing is
not to fail prematurely. Some bearings can be obtained with oil seals, but
the range of selection available is limited (for example, cam followers).
Where oil is mandatory in a needle bearing (for example, heavy duty) holes
are provided for lubrication supply.
A group of aircraft needle bearings specifically designed for heavy duty
operation and very low speeds will not be described in this text. It should
be noted, however, that all of these type bearings are provided with oil
holes for lubrication. For more detailed data concerning these bearings,
the reader is advised to refer to the manufacturers’ catalogs (for example,
Torrington Company).

Tapered roller bearings

Tapered roller bearings are available in the forms shown in Figure 9-20.
These bearings are designed specifically to withstand high radial loads, high
thrust loads, and combined high radial and thrust loads at moderate to
high speeds. The various parts that constitute a tapered bearing are illustrated
in Figure 9-21. It can be easily observed that the single-row tapered roller
bearing is capable of resisting thrust in one direction only. Consequently,
shafts mounting single-row tapered bearings must also mount an opposed
single-row tapered bearing, an opposed angular-contact ball bearing, or a
Conrad (that is, deep groove) type ball bearing. Where double- or four-row
tapered roller bearings are used, thrust in both directions can be withstood,
and the additional types of bearing just mentioned would not be needed.
The principle of operation of tapered roller bearings is similar to that of
bevel gear operation. Note that Figure 9-22 illustrates this idea. This figure
shows that the tapered surfaces of the cup, rollers, and cone represent
frustrums of cones, the elements of which converge to a common apex on
the bearing axis. As with bevel gears, pure rolling (or close to it because of
tolerances) action takes place. When under any type of load, whether
radial, thrust, or combined, the rollers are forced against the cone back
face rib (see Figure 9-21), which also provides guidance for the rollers.
Tapered roller bearings are ideally suited to withstand repeated shock
loads that can be expected from service applications (for example, auto-
motive wheels and transmissions, rolling mill shafts, railroad car trucks,
and so on). Also, the way in which a pair of single-row tapered bearings is
mounted plays a significant role in determining the rigidity of the bearing
assembly. This is best explained by referring to Figure 9-23 which shows
two mounting arrangements. The schematic arrangement of the two
bearings above the centerline is called an “indirect” mounting (note that
the cone elements converge towards apexes between the bearings). The
schematic arrangement below the centerline is called a “direct”? mounting
(note that in this case, the cone elements converge at apexes outside of the
bearings).
 464
Chapter 9: Rolling Bearings

Figure 9-20 Typical forms of tapered roller bearings and some of their
important characteristics. [From Machine Design, 1970 Bearings Reference Issue.
The Penton Publishing Co., Cleveland, Ohio.]

Approx Range of
Type Subtype Bore Sizes (in.)

Single row TST—tapered bore 0.3125 66.5000

iis TSS—steep angle 0.9375 17.0000
TS—pin cage 0.6250 50.0000
TSE, TSK—keyway cones =
TSF, TSSF—flanged cup 0.4720 14.8750
TSG-steering gear 0.3125 42.0000
(without cone)

Two row, double cone, single cups

Me TDIK, TDIT, TDITP—tapered Bore yee pita!

TDIE TDIKE-—slotted ine cone ice pe i
TDiS<stees ae 0.9375 27.0030
2.1650 20.5000

Two row, double cup, single cones,

adjustable
TDO 0.3125 72.0000
TDOC, TDOD (only one lubricant
hole in cup) 0.9600 56.3750
TDOS-steep angle 0.7500 56.3750

0.7500 24.0000
TNAD, TNADC (only one lubricant
Two row, double cup, single cones, hole in cup) 1.3125 24.0000
nonadjustable TNASW-slotted cones 1.1805 10.2500
TNA TNAU, TNASWE-extended ;
cone rib 0.7500 12.0000
TNAH, TNASWH-slotted cones,
sealed 0.3125 2.7559
TNAS-—steep angle 0.7500 4.5000
Four row, cup adjusted,
TAO |

2.7500 47.2500
TQOK, TQOT—tapered bore 9.7500 47.2500

Four row, cup adjusted,

TOI

TQOIK, TOlT—tapered bore —— —

In Figure 9-23, we see that perpendiculars have been drawn respectively

from point A, on bearing A, and point B, on bearing B to points A, and By
on the shaft centerline. Points A, and B, are the centerpoints of the respective
bearing roller cups. Thus, we obtain a distance a, (that is, effective bearing
spread) that is a measure of the rigidity of the bearing mounting. It is
 465
Section 9-2: Roller Bearings

Bearing width (7)____.»|

Cup width (C)

Cup back face radius (r)
Cup front face radius

crerce e SESS
D aN

VW,
\ Cup front face

Cage
a Cone back face rib
Va
Cone front face rib | Cone back face

Gone trontface Le GA Cone back face radius (A)

y »
Cone front face radius

Cone
(d)
bore

LL
Cup
(D)
(O.D.)
diameter
outside

EPRSS&ssf¥P
Cage clearance

Figure 9-21 Nomenclature for a tapered roller bearing. [Courtesy The Timken
Company. ]

readily seen that the actual bearing spacing a, for the “direct” mounting
(below the centerline) is longer than the actual bearing spacing a, for the
indirect mounting (above the line). From this geometric comparison, we
see that ‘“‘indirect” mountings provide greater rigidity when the pair of
bearings is closely spaced when used, for example, with idler pulleys, drums,
sheaves, spindles, front wheel automobile axles, stub gear shafts, and so on.
On the other hand, the “‘direct’’ mounting provides greater rigidity when
Figure 9-22 The basic principle for the correct operation of the
tapered roller bearing is for the tapered surfaces to converge to a
common apex on the axis of the bearing. [Courtesy The Timken
Company. |

:
l \i
Y I WG

[\
|
|\

ede B

WO
KHAN

Figure 9-23 Schematic arrangement of an “‘indirect”’ bearing mounting is

shown above the centerline, and that for “direct” mounting is shown below
the centerline. A complete explanation of the meaning and use of all the
letters shown can be found in the Timken Engineering Journal, Vol. 1.
[Courtesy The Timken Company.]

466
 467
Section 9-2: Roller Bearings

the bearings are not closely spaced and are used for mounting automobile
transmissions, machine tool gear boxes, speed reducers, machinery rollers,
and so on.
It is also of interest to observe that the two-row double-cone single-cup
and two-row double-cup single-cone adjustable bearings (see Figure 9-20),
which can withstand two-directional thrust loads, are similar to the DF and
DB duplex bearing mounting described in Section 9.1. These bearings can
be preloaded to provide greater rigidity at the shaft support. Preloading of
the double-cup bearing produces a stiffer (that is, high radial rigidity) shaft
assembly than the preloading of a double-cone bearing. The four-row cup-
adjusted bearings (Figure 9-20) are intended for extra-heavy duty service
(for example, rolling mills in the steel industry).
Tapered roller bearings to withstand pure thrust loads (not shown in
Figure 9-20) are also manufactured in standard forms. These are shown in
Figure 9-24.
For continuous reliable operation, the designer must provide for lubrica-
tion with grease or a less viscous oil. Before deciding on his final design, he
should carefully review the recommendations concerning proper lubrication
found in the manufacturers’ catalogs and consult the particular manufacturer
concerning any special or unusual requirements.

Figure 9-24 Typical forms of tapered roller thrust bearings. [Courtesy The
Timken Company. ]}

Approx. range
of bore size, in. Load range, |b
Rollers are spaced by a
TTSP :
stamped cage (i.e., separator).
6-—_}—__ >) The bearing is held together 2500-11.700*
SARIS —— as an assembly by a metal ‘
retainer. Used for oscillating
or slowly rotating installations.

1.2600-3.0100
Rollers are not separated
by a cage. The bearing is
held together as an assembly
7650-32,400t
by a metal retainer. Used
for oscillating or slowly
SS SSRN rotating motion.
1.2600-3.0100 Sa el
This is a heavy duty thrust
bearing designed for mod-
TTHD erately high speeds and cont-
inuous rotation. The rollers
operate between two
wos symm- 12000-1,930,000T
1.3750-48.0000
ot EI GET ssat etric raceways having inner
and outer ribs. The ground
rollers are held in a bronze
cage.

*For slow turning speeds (e.g. steering pivots)

+Load rating is at 50 rpm
 468
Chapter 9: Rolling Bearings

Spherical roller bearings

Spherical roller bearings (Figure 9-25) are available in single-row, double-
row or single-row thrust types. The important characteristic common to
all spherical roller bearings is their self-aligning property. The ability of
this type of bearing to adjust itself readily to shaft misalignment is found in
its fundamental construction, namely, grinding either the inner or outer
raceway to a spherical contour. Where the load is one that is predominantly
thrust with a small radial component, the thrust type spherical roller bearing
is recommended.
As in all the previous discussion relating to bearings, lubrication is of
primary importance in achieving reliable performance. Where the appro-
priate lubricant or method of lubrication is not listed in the manufacturer’s
catalog, the designer should discuss his particular problem with the manu-
facturer’s engineering department.
Figure 9- 25 Typical forms of spherical roller bearings and some of their
important characteristics. [From Machine Design, 1970 Bearings Reference Issue.
The Penton Publishing Co., Cleveland, Ohio.]
Approx Range Relative Limiting . Tolerance
of Bore Sizes (in.) Capacity Speed to
Min Max Radial Thrust Rector Misalannient

Single row,
bartel 0.7874 12.5984 2.10 0.20 0.50 +2°
or

convex

Double row
barrel 0.9843 49.2126 2.40 0.70 0.50 +1°30'
or

convex

3.3622 14.1732 0.10" 1.80° 0.35-0.50 ae

Thrust 0,10 2.40t

Double row, 1.9680 5.1171 2.40 0.70 0.50 +1°30'

concave

“Symmetric rollers. Tasymmetric rollers,

L

SECTION 9-3

Miscellaneous Bearing Configurations

All of the aforementioned types of bearings are available in a variety of
housing and mounting forms. In addition, special application type bearings
are also available. Figures 9-26 to 9-33 inclusive are illustrative of some of
the various types and forms available.
Figure 9-26 Fan and pump shaft bearings. [Courtesy New Departure-Hyatt
Bearings Division, General Motors Corporation. ]

Figure 9-27 Flange ball bearing.

[Courtesy Dodge Manufacturing
Division, Reliance Electric Co.]

(b)

(a)
Figure 9-28 Adapter bearing makes it possible to mount a ball bearing on
commercial steel shafting without machining a bearing seat. [Courtesy New
Departure-Hyatt Bearings Division, General Motors Corporation. ]

469
 Figure 9-29 Rod end bearing. It is used
for aircraft contro!s but has many machinery

KX cal
GQ Ts OT
and mechanism applications. [Courtesy The
Marlin-Rockwell Division of TRW, Inc.
V7
eeWA
KX
NN Jamestown, N.Y.]

Figure 9-30 Pillow block mounted ball

bearing. This type of mounting is also available /
for double row ball bearings, single and double
row roller bearings, and single and double row
tapered roller bearings. [Courtesy Dodge
Manufacturing Division, Reliance Electric Co.] ¥j

470
Figure 9-31 Recirculating roller chain bearing. [Courtesy Thomson Industries,
Inc., Manhasset, N.Y.]

Figure 9-32 Ball bushing. [Courtesy Thomson Industries, Inc., Manhasset, N.Y.]

471
 472
Chapter 9: Rolling Bearings

Figure 9-33 Ball bearing spline. [Courtesy Saginaw Steering Division, General
Motors Corporation. ]

SECTION 9-4
Standard Dimensions for Roller Bearings
For obvious reasons, standardization of common and widely used
mechanical elements is desirable in promoting economy in the design,
production, and maintenance of all kinds of equipment, machinery, and
devices. As mentioned at the beginning of this chapter, in the United States,
the Anti-Friction Bearing Manufacturers Association (AFBMA) is the
responsible organization for establishing and publishing the various roller
bearing standards and releasing revisions as required. The actual work in
preparing the standards was carried out by three engineering committees
appointed by the AFBMA, namely, The Annular Bearing Engineers
Committee (ABEC), The Roller Bearing Engineers Committee (RBEC),
and The Ball Manufacturers Engineers Committee (BMEC). The committees
worked in close cooperation with The American National Standards
Institute (ANSI) and The International Standards Organization (ISO). This
major contribution by the AFBMA has been most significant in promoting
the wide use and acceptance of uniform bearing standards. It is obvious
that we cannot here enter into a description of all the standards, and we
will therefore confine ourselves to data dealing with size specification.®
The dimensional system permits the interchangeability of the same size
bearing made by any manufacturer. However, there are special bearings,

° These specifications apply to ball and roller (straight, spherical, and needle) bearings in milli-
meter dimensions only. Standards for inch series bearings have been standardized in the United
States. To date, specifications for tapered roller bearings have not been established. However,
all the manufacturers have numbered tapered roller bearings similarly, and those numbers are
included in the AFBMA Standards, Section 5.
 473
Section 9-4: Standard Dimensions for Roller Bearings

limited application bearings, inch sizes, and so on, which are not standardized
but are subject entirely to the manufacturer’s control. In order to initiate a
uniformly standard system, it was first necessary to establish a range of
bore diameters. Starting with a diameter of 4mm, bearing bores increase
by | mm up to and including 10 mm after which they increase by 5mm to
120mm.’ Beyond this diameter, bearing bores increase by 10mm to a
maximum bore diameter of 500 mm.
In order to satisfy the different bearing capacities that might be needed,
a series of outside diameters was chosen for each bearing bore diameter.
These outside diameters (O.D.) called the diameter series are numbered
consecutively 8, 9, 0, 1, 2, 3, and 4, where 8 is the smallest O.D., and 4 is
the largest O.D. Thus, diameter series represent different size bearings having
the same section height to bore diameter (see Figure 9-34). Also, it can be
seen from Figure 9-34 that a width series 0, 1, 2, and 3 was established
whereby each successive number indicates and increasing ratio of width to
bearing section height.
The consequence of this system is that any two bearings belonging to the
same diameter and width series belong to the same dimension series (Figure
9-34). The dimension series is denoted by two digits—the first digit repre-
senting the width series, and the second digit representing the diameter
series. The system just described establishes standard external dimensions
without restricting the internal design of the bearing. The manufacturer is
thereby free to establish the ball or roller diameter, the quantity used,
raceway conformity, and so on, allowing the designer to choose from a
variety of bearing types that all have the same boundary dimensions.

Width Bp 0 1 2 3 4
series

Diameter
series

Dimension
series

Figure 9-34 Relative proportions of boundary dimensions of different bearing

dimension series.

7 Exceptions to this rule are the 12-mm and 17-mm bearing bore diameters. Also, manufacturers
list the 4-9-mm bore bearings as “‘extra small series,” and start with the 10-mm size when listing
the more commonly used bearings.
 474
Chapter 9: Rolling Bearings

To illustrate the AFBMA procedure, suppose a ball bearing is designated

as 10BC02. The letters BC are an indication of the type® of bearing. The
letters BC? also denote a single-row radial contact bearing having metric
dimensions. The prefix 10 indicates a bearing have a 10-mm bore diameter,
and the suffix 02 indicates that the bearing falls within the 02 dimension
series. Reference to Table 9-1 indicates that this bearing has an O.D.
1.1811 in., and is 0.3543 in. wide. This same size bearing as a filling slot
type (in metric dimensions) would be specified as 10BLO2.
Thus, any type of bearing having the same dimensions can be specified
by the appropriate letter designation separating the prefix and suffix numbers.
As indicated, these letter-bearing type associations are not listed in this
text. The designer is advised to consult the AFBMA Standards to obtain
complete information. However, he can specify a bearing size by using the
information in this chapter. For example, he may call for the use of a 10-mm
(0.3937-in.) single-row, deep-groove ball bearing of the 02 series.
Recognizing that designers do have available for their use the catalogs
of several bearing manufacturers, we will say a few words about their coding
procedure and how it relates to the AFBMA description. Although manu-
facturers adhere to AFBMA boundary dimensions, load and life ratings
have followed the older SAE standard of designating bearings as follows:
Extra light series or 100 series'® (equivalent to dimension series 01)
Light series or 200 series (equivalent to dimension series 02)
Medium series or 300 series (equivalent to dimension series 03)
Heavy series or 400 series (equivalent to dimension series 04)
Following traditional practice, manufacturers designate bearings by a
number-letter system that contains as a core one of the above number
series. The system is otherwise unique with respect to the company concerned.
The last two digits (that is, X00) of the core number are used to indicate the
bearing bore.'! Thus, starting with a bearing bore of 20 mm, the last two
digits when multiplied by 5 yield the bore diameter in millimeters. Other
prefix or suffix numbers and/or letters that appear in the manufacturer’s
8 The AFBMA has also established a bearing type identification code. This code has not been
included in this text because of the lack of space and also because the manufacturers persist in
identifying the type of bearing by their own systems. Nevertheless, the designation described
here should be familiar to the designer because various handbooks do contain the system. Also,
if the manufacturer’s catalog is not at hand a bearing designated by this code is easily translated
by the manufacturer or supplier to his own identification code.
° If the designation had been 10BICO02, an inch-size single-row radial contact bearing having a
te-in. bore diameter would be indicated. The 02 still signifies the same dimension series.
'°(1) New Departure-Hyatt calls this the 3L series. (2) Manufacturers’ catalogs also list an
extremely light series (equivalent to a diameter series 0). The numbering method for this series
is not consistent with all manufacturers. For example, New Departures-Hyatt uses 3LL00 as a
designation, and Marlin-Rockwell uses 1900. There is also lack of manufacturers’ (not AFBMA)
uniformity in designating instrument and extra small bearings.
'! This rule is not applicable to bearings with bores smaller than 20 mm.
 475
*a1dY PaqSt] JOU IIe PUk satsas UOISUAWIP Sty) 10J 210g WW (OI BIE ULdaq ssuLIeAg Japjos [eQUaYydS
“ZL6L WOK MAN “UONDIIOSSYSuaiNjaD{[ nun Sulsvag uoNILy-Huy ayfo spaopuolg *ASUNOS
000'ES | 00L‘6r oor‘te |008‘zz | 00L*6I |OOS‘LI | OOS‘ZZ |009‘0Z | OO8'8I | 009'SI | OOO'I! S6
009°8z | 009‘8z
| $9z0'% 000‘rr | 00r‘Or 009‘sz | 009‘0z |00s°81 | 002‘91 | 008‘0z | 00Z‘6I | CO9‘9I | OO9'EL | OFS | OFI'L| LIST I 06
009‘%Z | OOI‘zz
|009'EE Oor'zz | 006'SI |OOS‘'SI } OOI‘FI |OOL‘LI | OOE‘9T | OOF‘rI | COO'ZI | 0OS'8 | O6E'D | FZOII ss06's | OSI | S9PE'E | S8
OOL‘6I | OOS‘6I | SLEG'I | 76r | OOF'LE
| OOE'ZE | 000'6Z 00961 | 00z‘€! |oo6'e1 | OOE‘ZI | OOL‘S1 | OOI‘pt |OO9‘ZI |OOO'O! | Oz8°9 | DG's |9EZO'I | 9Z |6LO'O | BIIS'S | OFI | 96rIE | 08
OOI6I | OOE'8I | OOSL'I | pbb
| 007 8z | OOT sz 00z‘st | 00821 |ooez1 |00011 | ooo'hI | ooz‘z1 | oor‘t1 |OSz6 | OIL'9 | OT8'b | Ep8e0] sz] SOO | III'S | O€I | 8cS6C | SL
O0ELI |006'91 | OSZ9'I } E Lh
|oog‘ez |sozz't |te |oos‘b1 | 00z‘ot |ooo'zI |OOE‘OL | 006'Z1 | COI‘II |COS‘OI |OLH8 | 066'S | O22’ |6FF6' | Hz | 6S0'0 |EICOP | ScI | 6SSLC OL
ooc‘or |oorst |szost |cee |00z‘zz
|002'92 |oor'zz |sozz'1 | t¢ |oo6'r1 |00201 |o00I1 |OFS | oog‘II |COLO! |C066 | OL9L | O9E'S | OL8'E | SsO6O| Ez | 6S0'0 | PEL | OCI) lossc | $9
00st |ooo'rr |oo0s'1 | ge
|00€'2z |Ooz‘st |pzo1'1 |gz |oo9‘zt | 0968 | 00L‘'6 | O66L | COG! | OSI'6 | OL0'6 | OS69 | OOZ'S | OGr'E | 19980 | Zz | GSO } LOEE' | OII | cede 09
oos'er |ooc‘zt |suept | so¢
|000'st |erge'o |sz} oog‘ot |o1L'9 | o10'8 | OLr‘9 | 86's | OLHL | OOS'L | 09'S | O€9'r | O1O'E | B9z8'0] Iz | OSO'O | OLEG'E OOI | Ps9T'z} Ss
oor'tt |ooc'or | szie't | eee |00€'81
|00c‘bI |oo0'zt |ssoe'0 |€z | orr's | oce's | ozb‘9 | OII's | o9z‘L | oss’s | OLO'9 | OSr'h | O€6'E | OLE] PL8L'0] OZ | EDO | EEPSE | 06 $8961 | 0S
ozz'e | oct'g | szgtt |zo¢
| OOE TI ozo’s | o1o's | ogz‘9 | ozz'r | oz6'9 | opps | 99's | O1O'r | O8LZ'€ | O9I'7| O8PL0] 6I | 6E0'O| S9PEE] SB | LILL I} Sb
0z9'g | oze'L | SL8TT 007bI
009'ET | OL8‘6 oa‘. | o99'r |o9s‘s | o6t'r | o9t‘9 | osc'e | opo's | ozs‘e | o€e‘e | O€6'I | L80L'0} 81 | 6E0'O] 96FI'E | O8 | BPLST| OF
08%. | O€h'9 | SLIT
= = —! oes | osr'e | o'r | o6c'e | 091s | opg’e | Opry | oco'e | Op'z | OOS‘I | €699°0] LI | 6E00} SPER T] TL | OBLET | SE
089 | 009s | Sz90°I =
— - - — | ozee | ozz'z | osse | o6r'z | ole‘e | ozs‘z | o9e'€ | osz‘z | O1L'Z | OIE I | 66790} 91 | GEO} 60H 7c] 79 | II8I'I | O€
Oris | OZI'p | SLE6'O
| Sz18'°0 ose'z | oza't | o9s‘z | oct | ozs‘ | 0961 | ozr'z | o9s‘t | oor'z | $06 | 90650} SI} 6E0:0| LPO} ZS | EF860} SC
OoL‘e | 098'%
| Ops'z | SzIs‘0 = = = = = = or'z | o¢9't | o1z'z | oeetr | OIL | SIL | ZISSO]| p1 | GEO} POSSI] Lp | PL8L'0} 0%
Olv‘e
= = = org‘t | ogi‘ | osotr | O00! | oze'l | SPS | PZLhO| ZI} PZO'|O] BFLS'I | OF | £6990] LI
Ors'z | Ost | SL89°0
= = = OLPl | 076 OzET | 06L O6ZI | SP Ieev'0] 11 | pz7O'O} OSLE1| SE | 90650] SI
OEo'z | O€h'l | 0Sz79°0
= = a Ove'l SEL Osl'l | $89 $96 tee LE6E'O| O1 | PZOO| 86ST1T] ZTE | PeLyO] ZI
ozsl | OSZl | OSz9'0
= = = = - S08 Or 056 ooe | erseo] 6 | Pz7O';O| II8I1] 0€ | LE6EO]| OI
Opel | 008 $790
S8L Orr Sor soz | OSTEO] 8 | PCO'O] YECOI| 9% | EPSEO] 6
sos POE 09r srl = = || = — |— | Osteo) |3
9s vOe 09 Srl 9SL70] L | Z1O;O} 19980] 7 | 9SL7O} L
Si 117 Sep gil ZETO] 9 | ZIOO] O8PL'0| 61 | C9ETO] 9
SI 11Z — - 69610] S | 7100] 6679'0 | 9I | 69610] S$
te 897 871 — - 69610] S | TIOO] SIISO] Er | SLSTO}
| | ahLe =
Z Salias Ja}aWeId
Buueag [eq Suueog [12a Suyseag [ed Suurag [Ie a
Buresg J2[]0y YoION Bunesg s2]/0y Suueeg 19][0y aauy-daais asuy-ews Bee Suueag [eq “snipey
, Teouaydg -Bur!4-uoN jpuaydg Teoupur[AD yoR}U0D yoR WO) MOY-a[sutg SupUsye-J[2¢ BOLE |
MOY-2]Qnoq J UPI J UWIPIN ~reinsuy -1eynsuy : g PIN xeW

uoIsusUNG
q-eip P

suolsuawiq
apisino a10g Buueag

Saag
ZE Souag UOIsuaUTIG ZO sauag uoIsuswig

psepuels

‘Buljes
ZZ

sBuresg

sbuljey
Buljjoy

peo]
jeipey

(qI
peo7

pue
°4)
2 ‘q| ‘Burjes peoj oes dIseq =

jo
oiueuAp oIseq L-6
9I9eL
476
F1q4eL
7-6 piepueis suoisuewiq
pue peo7 s6uney
jo jeipey Buljjoy sbuueeg

uolsuauig
Saliag
£0
Buueag
asog apisino uoIsuawig
Sauag
£7 uolsuauig
Salag
£€
ee xe UPIM
a7 mou:ateuls ~rem8uy
y1e-195Surv MPI
7 MPIM
ag : Moy-a1qnog
eae joe1u09 [eoupUNAD jeousyds BujUste-sIag peousyds
| snipey [eg Sunieag Saale apuy-daais J2]0y Bulieag
haweaes yuoBuy
i Jayjoy Buueeg [jeg Bulseag sa0y Buueag aie cent
[eq Sureag yonTeg
uw ‘ul a Suueag
i

2 y | 9)
2

1a}
v ses}
[ 91 | 6629°0 eho. seeios|
lp i

€souas
S$ 69610] pes

0WIRIG
i

|
|
6I
|
TIO'O]

69610]
O8FL:0
T9ETO}
Osteo

C9ETO]

s 9 L 8
9SLTO}
6 EPSEO]
Ol Leeeo]
| SE | O8LE1 pz0'0}11] l€€r'O] - _ Spgs| OOP!
ZI p@Lb'0|
| LE| L9Sp'l
| 6€0'0
| C1 v7Ly'0| oEs| Joeol OPO!| 0891
SI 90650]
| Th| SES9'L
| 6€0°0 EL] 8LISO] 06s| |osol OzZI
| O96'l
LI €699°0]
| Lp | poss"!
| 6E0°0
| 1 TISS‘0] Oz| OLI‘z
| OLF'I
| OvE'Z O8bL'0 seg| 00s‘z
| — — |osevo|zz}
— = 069'7
| O£9E
OC PLEL'0|
| TS| ZLPO'7
| 6€0'0
| SI 90650] 006| Jost'z OSL| OSL'z
| Ozer
| ooo's - -
Sz
- - Iz 89z8°0]
| oort
| |OIze — — zzz}
| Oszgo
| OSL'I
| OSL‘
| OOT'E
| Oley
EFR6°0|
| 79| 6opr't
| 6€0'0
| LI €699°0]
| OSE'T
| OIE o6¢'z
| o99'e
| ozs‘z
| ozz'b
| ozs‘
| ers| o00'9
| 00z'9
| bz 6rb6'0}
| |oso'1 oiz‘r
| — — #sz7]
| Oooo!
| O6E'%
| 099'€
| o9E'p
| 099's
o¢ LIgt1|
| zz opes'z|
| 600 61} Osrz'0|
| Opel
| oOL‘'e
| Ove'e
| OsB'r
| Os'c
| OLE's
| Oboe
| 09L'9
| 005'8
| 0OS‘6
| Le O€90'T}
| Osz'z
| OCr'S
| — — zoe]
| sugrt
| obe'e
| Oss'p
| oor'9
| O6r'L
s¢ | ogzet
| 08 sere] 6s0:0|
| 12 89z9'0]
| orziz
| |ose'r OzO'r
| Osz‘s
| Oz9'r
| OvE'9
| OLE‘s
| oEs's
| 0086 OO8‘oI|
| 1€ SOzz'T}
| oLs'z
| 0289
| — — |ore| oseeT
| OzOr
| OSL'S
| Opes
|
oF spzs1|
| 06| eers'e oso] Ez] ssoeo}
| |\opz'z JorI's OzO's
| OFO'L
| OLL'S
| OvL'L ops‘9
| 098'8
oos‘or
| oo6'z1 ooz‘er|
| ee| zooz't
| oes‘e
| ose‘z
| oo1'91
| oor‘0z
| s-9¢| svev't
| OzO's
| OPO'L
| OLI'6
| OO8'OI
Sp | LIZLI COI}| OLE’
| 6500 Sz] €h86'0|
| O8s‘e
| 009‘9
| OEL9
| OZI'6
| O£L'c OOT‘O!
| O¢o'6
| OOZ‘rI
|} OOE'LI OOO'LI|
| 9€ ELIP1| OLb'b OEts
| 009'0z
| OOO'SZ
| L*6E
| szost
| OEL'9
| OzI‘6
OS $8961] O11]
| LoEE’b
| 6200 Lz] O€90'1}
| O€6'e
| OIS‘e
| O10'8 OOL‘OI}]
| 0026 OO8‘IT
| OOI‘I1
|
OOE'ZI}
| OOl'bt
OOO'LI
| OOE'8I OOE*6I|
| Ob| BrLS'1 O87's OOTII
| OOs‘9z
| OO9'IE
| tbr | ooszt
| OLo's
Ss S9I-z]}
| OzI | brZL'b 6200} 6z] LIFI'I|
| O90'S
| O08'8
| OOF'6 OOF'ZI}
| 008‘OI 009‘EI
O0z‘o1| 009r1}
| 00S‘91
| 009‘E1
| OOL‘oz 009'IZ} 008°ZZ}
| €b| 6769'I OOF‘
| ooO‘E!
| o09‘z¢
| O08'LE
| Z’6r|
09| zz9E7z O€1|
| I8II’s
| 6200
| Te soz]
| O86's
| |/098°6 O06‘OI OOI'PI}
|
see
| OOr'6
| OOr'cI OOz*L1])
| 000'61
00S‘Z1 009'SI
} OOS*S1
| OOE‘Ez
| 000'SZ 000°9Z|
| 9| OLI8l OFZ
| OoI‘st
| oo9‘g¢
|
so loss'z]} OPI}
| BLISS
|
OOE'tt
| o'bs| oszi'z 00601)
| OOI'bI 00651}
| 008'IZ
6L0'0
| €£| 76671 009'9
| OOZ‘oI
| OOS‘ZI
| OOO'91
| OOr'pI
| 009'LI
| OOI‘L1
| 009‘SZz
| OOS‘OE 000'1€} 8688"I
| O89'8
| 009'91
| OOT'Z
| OOL'8r
| L°8S| STIET aa = 006°7Z
| 009°bZ
OL 6SSLT}OSI| Sso6's 6L0'0}se} O8zEe1l|
| 0862 OO6'ZI|
| OOZ'rI
| 000'SI
| OOr'9I
| 008'61
| 008'0Z
| OOI‘OE
| 000'rE OOS‘bE
| IS 600°]
| 00001
| 006'81
| 00S*6r
| OO8'Ss S°€9] 000S°Z.| = =
SL 8CS6T]} O91| 2662'9
| 620° LE} L9Sb'1]
| Oz9‘8
| OOL‘EL
| 000'97
| 00L'LZ
OOI‘9I
| 009°61
| OOS‘sI OOS‘IZ]
| 009'Sz
| OOS*9E
| 000'8E OO0'6E
| Ss psoI'z|
| OOS‘o1
| 009'6I
| 00z'9s
|
O8| 96rI'€OLI|| 6z69°9
| 6L0'0
| 6¢ 00r'Z9 £89) SLE9T! = a 6z OOF‘
| 007'‘OE
pSES1|
| O6r'6
| OOE'sI
| 00'S! OOEIZ}
| OOL‘OZ
| OOE'EZ
| 009'SZ
| 00S*9E O00'Er OOO‘Er]
| 8S SeRz'z|
| OOO'E!
| 006'EZ
| OOP'b9
| OOF‘IL SL89°C = = 006'7E
| CE OOL
$8 | S9rE'E O81]
| 9980°L ZFI9'I
|; 006‘OI
| 00691
| OOT‘OZ 006'ZZ|
| OOIEZ 00%'sz]
| 00S‘0E
| 006% = — 09| zz9¢"z
| O08‘EI
| OOE'rz
| 008'89
| 00b'9L OSL8'T = =
06| E EPS £O8b'L [ 67269 00Z‘0Z
| OOO'ZE
| OOl'tE
OOE‘Zz OOL‘bz}
| 009"Sz 000'LZ]
| OOE'EE
| 00S‘9p a = LO1S"z
| OOF'SI
| OOF‘9Z
| 00z'sg
| 009'06 OSL8°T) O09'SE
| 009‘rE
:ADINOS Spsopuol§
fo ayl uowaig-iuy
sagSta jo0fnuopW
Suan
Y “UONDII0SS
MAN MOK “TLOL
 *a12Y past] 10U St pu 210g BulseIq WW-O| | & YIIM suTaq | ¢ SeLlas UOISUaUIP 10} sfuLag JaTIO4 [BoLaYds|
*a19Y Paist] JOU SI pur asoq SulsvAg WW-OZ |B YIM sufaq YE Salles UOIsUaUIP 105 s8uLTeaq 19]|01 [eoLIaYds|
TLO6I OA MON ‘MONTIIOSSY Siasniouf[nunyp BuLwag uoNIU.-UUp ayi fo spAvpurig -ADANOS

006°LZ
008°97
OOF‘OT
OOO'EI
|00871
00L‘9Z

OD0'SZ

OOS‘ZI

OOF'ZI

OOS‘OI
|

|
4
OOr‘Sz | OOE*EZ ; 00771 | OO8T1 |000'OT
O08‘1Z | OOT'0Z o0z‘OI | 166 | 09S°8
008‘0z | 009°81 0566 | Orr6 | OFZ8
006°9I | 000'ST 3 Ori’s | O%L°L | 0€8'9
000°L1 | 000'S! 4 089'L | 0869 | 08s°9
OO7EI | OOETI 0679 | 008'S | 08z's
oos‘Z1 | 00r‘ol Ores | OS7'S | 060'S
OOs‘II | Ors‘6 008°S | OL6r | 068'r
0s0°6 | O9T'L : Olrb | O9L‘€ } OLLE
oss | O1S‘9 Orl'h | O9E'€ | OENE
O9TL | OP'S O6re | O8L'z | 0067
OSvE
OS'Z
0L0°9
Olt
O9L‘Z
|
O19%r | O9T‘E _ — 0677
069°€ | ObP‘z — _ OrL'l
0791

O0'T
$96
06L088
S8s
sos
Ove

2a0015-daeq
jeoupur[A

Suresg
Suureg
Suleag

12/04
Bureag
‘snipey
[roueydgs
Teouayds

Ja]]0y
Ja]]Oy

[|eq
OUSIIENES
ye]U09
TEER
ERE

ILES
Surreog [1eq

-rejnBuy

TWIN
xey11TA

ie
WIM
4

:
apisino,
aid

UOIsUaIG,
H
uOIsusUIG

UOISUSIG,

qd
g

aPIsinO

Suuereg
p
aiog
Saliag
Saag
Seuag

Ql
O€
[€

M
| sauag 1aj9WIeIG Q salag 19} 9WIeIG
sBuueeg Burjjoy jeipey jo shuney peo7 pue suoisuewiq puepuelg €-6 BIqGeL
 478
Chapter 9: Rolling Bearings

designation of a bearing are used to indicate features such as bearing type,

self-lubricating, type of seals, snap ring for mounting, material differences,
and so on.'?
As an illustration, a New Departure-Hyatt bearing having a basic bearing
number 3212 indicates that it belongs to the 200 series (that is, light series),
has a bore diameter of 60-mm (12 x 5), and is a Conrad (that is, deep
groove) single-row bearing open on the sides. The last part of this description
is based on the use of 3000 by New Departure-Hyatt to indicate this type
of bearing. This same bearing is designated by the Marlin-Rockwell
Company as 212-S, where the 212 has the same meaning as described above.
However, the S is used to indicate a single-groove, Conrad type bearing.'*
The same bearing is designated by SKF as bearing number 6212. Notice
that SKF uses 6000 as the basic number to indicate single-groove, Conrad
bearings. Nevertheless, the important numerical characteristic to observe
is that in all cases the number 212 appears. This common basic number
makes it simple to substitute the bearing of one manufacturer for that of
another. As a service, many companies list bearing conversion tables at the
end of their catalogs. This same bearing designated by AFBMA standards
would be 60BC02.'*
It is obvious from Figure 9-34 that the number of bearings represented
by the dimension series is very extensive, and to reproduce them here is
not feasible. Therefore, Tables 9-1, 9-2, and 9-3 represent portions of the
more widely used bearings, and the reader is advised to consult the AFBMA
Standards or a manufacturer’s catalog for a complete listing.

SECTION 9-5
Bearing Tolerances
The Annular Bearing Engineers’ Committee (ABEC) of the AFBMA has
established five basic grades of ball bearing precision. These grades are
ABEC 1, 3, 5, 7, and 9. Grade | has the widest tolerance; as the numbers
increase, the tolerance becomes smaller. For most normal applications,
ABEC | is satisfactory. Where closer tolerances are needed, ABEC 3 and 5
can be specified. ABEC 7 and 9 are super precision grades reserved for
'? Here again, there are some exceptions to the practice of bearing designations by manufacturers.
For example, the New Departure-Hyatt basic bearing number 3212 becomes 7512 with one
shield, Z 97512 with one shield and one seal, 477512 with a snap ring and two shields, etc. Note
that the only numbers remaining of the basic bearing numbers are the last two digits (i.e. 12
which indicates a bearing bore of 60 mm).
"3 The Marlin-Rockwell Company is one that continues to use the same basic number regardless
of special features. For example, a 212-S bearing with one shield is designated as 212SF, as
212SF2 with one shield and one seal, as 212SFFG with two shields and a snap ring, etc.
'* The reader using the standard dimension and load rating tables of the AFBMA listed in this
text would, in selecting this bearing, call for a ‘“60-mm (2.3622-in.) bore, single-row deep-groove
(i.e., Conrad) ball bearing of the 02 dimension series.”
 479
Section 9-6: Rolling Bearing Materials

applications requiring extreme accuracy. Such applications are found in

the design of instruments, aircraft controls, servosystems, and machine tools.
Considering the discussion in Chapter 4 concerning tolerances and costs,
the designer must be able to justify specifying high ABEC grades. For
example, in the bearing bore range of 11 to 16 inclusive, the ABEC tolerances
are as follows:

ABEC Grade Tolerance

+0.0000
—0.0006

+0.0000
3 —0.0004
+0.0000
5 —0.0003

+0.0000
—0.0002

+0.0000
9 —0.00015

From this list, it is readily seen that the ABEC grade 9 has only one quarter
the allowance of the ABEC grade 1. Reference to Figure 4-69 shows the
relative time to produce the tolerance for ABEC 9 as 11.5, and the tolerance
for ABEC 1 as 7.9. Thus, to produce grade 9, a 45.6% longer period of
time (that is, increased cost) is required than to produce grade 1.
Manufacturers’ catalogs contain the complete ABEC tolerance tables
both for bore diameter and radial runnout. Tolerance grades have also
been established for cylindrical roller bearings by The Roller Bearing
Engineers’ Committee (RBEC) of the AFBMA. For cylindrical roller bear-
ings, there are only two grades, RBEC 1 and 5. These grades have the same
tolerance values that have been assigned to ABEC | and 5S. Tapered roller
bearing tolerances have also been graded. These grades are 4, 2, 3, 0, and 00
with 4 having the lowest precision and 00 having the highest precision.!*

SECTION 9-6
Rolling Bearing Materials
Ball bearings (except for the retainers, seals, and ball separators) are
commonly made of SAE 52100 (or AISI E-52100), a high carbon chromium
steel that is through hardened to Rockwell C 58-65. Due to the loss of
hardness with increasing temperatures, SAE 52100 cannot be used where
the temperature is above 350°F.
15 See Section 1 of The Timken Engineering Journal, The Timken Roller Bearing Company,
Canton, Ohio.
 480
Chapter 9: Rolling Bearings

Bearing steels are produced by the vacuum-melting (VM) process and, in

particular, by the consumable electrode vacuum-melted process (CVM) in
contrast to air melted steels. Steel produced by these methods is more
uniform in structure and contains fewer impurities. As a result, the reliability
of bearings has been improved.
High temperature bearing problems are not uncommon and in the past
years have served to prevent the development of both industrial and military
hardware. However, continuing research has led to the development of
materials capable of maintaining a through hardness greater than Rockwell
C 58 at higher than normal temperatures.
Figure 9-35 illustrates how some materials vary with temperature. The
maximum operating temperature of each material is that point at which the
curve crosses the Rockwell C 58 line.
For mildly corrosive applications, it is quite common to use grade 440C,
a martensitic stainless steel. Notice that in Figure 9-35 the anticorrosion
steel is “‘restricted” to an operating temperature no higher than 300°F.
At a temperature above 300°F, its hardness drops below Rockwell € 58
and the raceways may begin to “‘brinell.”’

66

64

62

60
Figure 9-35 Hardness versus
temperature of some through hardened
bearing steels. [From Machine Design,
58 1970 Bearings Reference Issue. The
Penton Publishing Co., Cleveland, Ohio.]

Hardness,
C
Rockwell
56
SAE 52100

54

AISI M-50

52

50
0 200 400 600 800 1000

Temperature, °F
 481
Section 9-6: Rolling Bearing Materials

Roller bearings are usually fabricated of case hardening (that is, carburiz-
ing grade) steels. These steels are surface hardened to Rockwell C 58-63 to
a depth of no greater than 0.015 in., past which there is a softer core with a
hardness of Rockwell C 25-40. Carburizing grades (for example, AISI 4620)
are limited to an operating temperature no higher than about 300°F.
Continuing metallurgical research, however, has led to the development of
carburizing grade bearing steels that are suitable for higher operating
temperatures. Figure 9-36 shows the hardness-temperature plot of two such
steels in comparison with AISI 4620. These curves illustrate the marked
increase in operating temperatures attainable by special steels.
Ball separators (that is, cages) are normally fabricated of 1010 steel.
However, for use at elevated temperatures (up to 700°F), separators made
of M-1, S-Monel, 440C, polyimide polymer (under 700°F), cobalt alloy,
and copper alloy (for example, bronze or silver-plated silicon iron bronze)
have been successfully used.
For mildly corrosive environments it is customary to use grade 302 or
grade 410 stainless steel as ball separator material. Most angular-contact
bearings and high speed bearings use ball separators made of cotton woven
fabric impregnated with a phenolic resin.
Seals used to retain the lubricant in “‘self-lubricating” bearings are stamped
from a low carbon steel (for example, 1010). Depending upon the manu-
facturer and the particular bearing, the seals are either completely molded
into a synthetic rubber compound or just edged with this material. Also a
seal made of felt pressed between two metal rings is quite popular.
Shields that help both to retain grease and to prevent chips, dirt, or large
particles from entering the bearing are also made of a low carbon steel
which is crimped into a groove in the outer ring and maintains a close
running fit with the inner ring.

fe>)oOo

aoa

oaao

Hardness,
C oao
Rockwell

45
100 200 300 400 500 600 700 800
Temperature, °F

Figure 9-36 Hardness versus temperature of three carburized bearing steels.

[Courtesy The Timken Company. ]
 482
Chapter 9: Rolling Bearings

Where extreme temperature (for example, 1500°F) and environmental

conditions have precluded the use of available stock bearings, materials
such as alumina, zirconium oxide, titanium carbide, pyroceram, 16 and so on,
have been used in place of steels.

SECTION 9-7
Rolling Bearing Friction
Bearing friction is greatly influenced, among other things (for example,
hysteresis, slip), by the coefficient of friction between the rolling members [3].
Table 9-4, based on years ofaccumulated test data, lists the average coefficient
of friction (uw) for several types of bearings measured at the shaft surface.
it VO} i ble 9-4
4
rea -

Table 9-4 Average coefficients of friction

Starting Running

Type of Bearing Radial Thrust Radial Thrust

Ball bearing 0.0025 0.0060 0.0015 0.0040

Spherical roller 0.0030 0.120 0.0018 0.0080

bearing

Cylindrical roller 0.0020 = 0.0011 =

bearing

SOURCE: Bearing Technical Journal. FMC Corporation, Link-

Belt Bearing Division, Indianapolis, Ind., 1970.

Actually, the greatest contribution to bearing friction is made by the seals

which can produce several times greater friction than the bearing torque.
Where the designer feels that the total bearing friction (that is, including seal
friction) is a significant factor affecting his design, then consultation with the
manufacturer becomes necessary.
Although insignificant in many designs, the horsepower lost due to
frictional torque is
eed Tn pk dn 9-1
> 63005) 106050 ees,
where

T = frictional torque, lb-in.

n = Shaft speed, rpm
d = bearing bore, in.
F,, = radial bearing load, Ib
'® Product of Corning Glass Company.
 483
Section 9-9: Life, Rating Life, and Basic Load Rating

SECTION 9-8
Basis for Bearing Failure
Carefully installed bearings that are properly lubricated, kept free of
foreign matter, and are not overloaded will fail by fatigue. Evidence of this
failure appears as spalling or flaking of a raceway or rolling element(s). As
the rolling elements revolve within the raceways, they are subjected to a
repeating contact (that is, Hertzian) stress. In addition, those areas of the
raceways that support the shaft load as the rolling elements pass over them
are also subjected to a repeating contact stress. Because these stresses are
greater than the endurance limit of the material, it is obvious that the
bearing will have a limited life. In fact, if one considers that pure rolling does
not occur in the bearing but that some sliding of the balls or rollers takes
place, it is readily understood why particles of metal are removed from the
raceway surfaces. The removal of this metal is called spalling and, when this
condition arises in a bearing, the bearing has failed.
The statistical analysis concerning the fatigue failure of rolling bearings
presented in 1947 by Lundberg and Palmgrem [9, 10] provided the ground
work for the adoption in 1950 by the AFBMA of a “Standard method of
evaluating the load ratings of annular ball bearings.’ This standard has also
been approved by the American National Standards Institute (ANSI
B3.11—1959) and is the basis for selecting a bearing as described in this text.
A similar standard for evaluating roller bearings was also adopted.

SECTION 9-9
Life, Rating Life, and Basic Load Rating
Notwithstanding modern manufacturing methods and the use of quality
control procedures, no group of seemingly identical bearings is exactly alike.
Each may differ, however slightly, UME Neal!in ENE ony soars
of balls, and so on. as acoulis no twe within th > family ma

dite al 1d Thus, the life ofa bearinens

established on a purely statistical basis (see reference [3]) and the AFBMA
has established the following definitions to ensure a uniform understanding
and procedure when selecting a bearing:
1. The life’’ ofa radial ball bearing is the number of revolutions (or hours,
at some given constant speed) that the bearing runs before the first
evidence of fatigue develops in the material of either ring or in a ball.
2. The rating life of a group of apparently identical radial ball bearings,
thrust ball bearings, radial roller bearings, or thrust roller bearings is

17 This same definition is also applicable to thrust ball bearings, radial roller bearings and thrust
roller bearings by just replacing “‘ball” with “roller,” and “‘ring” with “‘washer.”
 484
Chapter 9: Rolling Bearings

defined as the number of revolutions (or hours at some given constant

speed) that 90% of a group of apparently identical bearings will
complete or exceed before the first evidence offatigue develops. (Rating
life is commonly referred to as Lo life, B-10 life, or minimum life.) As
presently determined, the median life’® that this same group of bearings
would complete or exceed is 5 times the rating life (that is, median
lifes aL):
3. The basic load rating’? (also called the dynamic load rating, basic
dynamic capacity, and the specific dynamic capacity) of radial ball
bearings is that constant stationary radial load which a group of ap-
parently identical ball bearings with stationary outer rings can endure
for a rating life of 1 million (that is, 10°) revolutions of the inner rings.
In single-row angular-contact bearings, the basic load rating relates to
the radial component of the load which results in a purely radial dis-
placement of the bearing rings in relation to each other. Load ratings,
if given for specific speeds, are to be based on a rating life of 500 hr. (This
is equivalent to running a bearing for 500 hr at 333 rpm or for a total of
10° revolutions. On the basis of median life, this is equivalent to having
run a bearing for 2500 hr at 335 rpm for a total of 5 x 10° revolutions.)
4. The basic static load rating!? is defined as that radial load which corre-
sponds to a total permanent deformation of rolling element and race (or
washer) at the most heavily stressed contact of 0.0001 of the ball or roller
diameter. In single-row angular-contact ball bearings, the basic static
load rating relates to the radial component of that load which causes a
purely radial displacement ofthe bearing rings in relation to each other.

Basic static load rating

The basic static load ratings for different types of bearings, symbolized by
Co, are given in Tables 9-1, 9-2, and 9-3. The value of Cy) depends on the
bearing material, the number of rows of rolling elements in a bearing, the
number ofrolling elements per row, the bearing contact angle (see Figure 9-8),
and the ball or roller diameter. Those readers wishing to know how C, is
calculated should refer to Section 9 of reference [7] for specific equations and
data.
Normally the basic static load HOE has little influence in the Sa of
a rolling bearing. However ifa bearit r ed
PpoTiseitta tein ocala suena AER eRtterese This is Ree caaly
'8 Various manufacturers’ catalogs and texts mistakenly call this “average life.” The designer
must be careful in using this term since this usage is quite prevalent. The AFBMA defines
average life of an antifriction bearing as the summation of all bearing lives in a series of life
tests divided by the number of life tests. Some manufacturers define average life as being equal
to 7 times the rating life.
'? This same definition applies to radial roller bearings, thrust ball bearings, and thrust roller
bearings. However, for thrust bearings, the loads are not radial but are thrust loads, and the
word “‘ring”’ is replaced by the word ‘‘washer.”
 485
Section 9-10: Equivalent Load

true where the load exceeds C,). Where operating speeds are very low, higher
loads than Cy can be allowed but should be checked with the bearing
manufacturer.

Basic load rating

The basic load ratings for different bearings, symbolized by C, are listed in
Tables 9-1, 9-2, and 9-3. The value of C depends on the same factors as those
which determine Cy, except for an additional parameter concerning the
loading geometry. As with Cy, Section 9 of reference [7] contains the specific
relationships and data needed to compute C.
The basic load rating C enters directly into the process of selecting a
bearing, which will subsequently be explained.
Extensive testing of rolling bearings and subsequent statistical analysis
have shown that load and life of a bearing are related statistically [3]. In
particular this relationship is defined by the expression

Lio= (5 (0-27
where L,, = rating life in millions of revolutions (that is, the number of
revolutions resulting in 10% failure); C = basic load rating in pounds
(obtained from Tables 9-1, 9-2, 9-3, or a manufacturer’s catalog); P = equiva-
lent load in pounds (see Section 9-10); b = 3.0 for ball bearings*! and 42 for
roller bearings.
In terms of hours oflife, equation (9-2) becomes
10° (s)
erotEs ApoleP (9-3)
where n = rotational speed in revolutions per minute.
Again, we urge the designer to read carefully that portion of the manu-
facturer’s catalog that explains how load rating and life are established.
Although most manufacturers adhere to the recommendations of the
AFBMA, there are some exceptions. For example, New Departure-Hyatt
list their load ratings on an average (that is, median) life of 3800 hr instead
of 2500 hr as specified by the AFBMA. Regardless of catalog variations,
however, the designer can always convert the data to the AFBMA standard.

SECTION 9-10
Equivalent Load

2 Equation (9-2) is not applicable to filling slot type bearings.

?1 New Departure-Hyatt uses an exponent of 4.
 456
Chapter 9: Rolling Bearings

In many applications, bearings have to carry a load composed of axial and

radial components. In addition, they are sometimes required to operate with
a rotating outer and a stationary inner ring. It then becomes necessary to
express these conditions by an equivalent load satisfying the above definition.
The expression used to define this equivalent load, which is applicable to ball
(that is, radial and angular-contact) bearings and roller bearings (not
applicable to filling-slot bearings) is

P = XVF_+ YF, (9-4)

where

P = equivalent load, lb
F. = radial load, lb
F = thrust (that is, axial) load, lb
V =a rotation factor: 1.0 for inner ring rotation, 1.2 for outer ring
rotation. For self-aligning ball bearing use 1 for inner or outer ring
rotation.
X = aradial load factor (see Table 9-5, 9-6, or 9-7 as applicable).
Y = a thrust factor (see Table 9-5, 9-6, or 9-7 as applicable).
Looking at Table 9-5, we observe that the radial factor X can in some cases
be less than unity. Thus, it is possible to have a combination where, the
if

P = VF, (9-5)
The

Equations (9-4) and (9-5) are also used to determine the equivalent load
for single-row and double-row spherical bearings. For single-row spherical
, values of and
X Yare found ne =
bearings, values of and
X Yare found in

Equivalent loads for thrust

ball bearings and thrust roller bearings
can be
calculated from equation
(9-4) where V is unity.
Space limitations make it impossible to list the V factors for straight
cylindrical roller bearings, the X, Y factors for both types of thrust bearings,
and the respective bearing tables which give the values for Cy and C (that is,
tables like Tables 9-1, 9-2, and 9-3). Such information is readily available in
the AFBMA Standards, engineering handbooks, and manufacturers’ catalogs.

°? For those straight cylindrical roller bearings that have a locating flange or flanges (see
Figure 9-18), the maximum thrust load should not exceed the value determined by, F, < 0.2F,.
Table 9-5 Factor X and Y for ball and roller bearings
Single-Row Bearing? Double-Row Bearing?
Contact Angle !
a, deg F,/VF,>e Fq/VF, Se Fy/VF, >e e

Radial-contact groove ball bearing*

Angular-Contact groove ball bearings?

and e val-
ues appli-
cable to
single-row
radial- |
contact
bearings

SoHnNwEUIO
CO
O-OWh—N—

NNNY

SCOO—H—-NwWaar
OWNOO~
CONN WHRANWO—-wWO
ARNnDAwDOO—NWwW

Self-aligning ball bearings

OC WD
Roller-bearings, self-aligning, tapered>

SE CR
SOURCE: Frank W. Wilson (ed.): Tool Engineers Handbook, 2nd ed. McGraw-Hill Book Co,, New York, 1959.
NOTE; Values of X, Y, and e for a load or contact angle other than shown are obtained by linear interpolation,
1 Co is the static basic load rating; / is the number of rows of balls (not used for radial contact bearings).

When calculating the equivalent load for a unit consisting of two similar single-row angular-contact ball bearings in a duplex mounting,
“face to face” or “back to back,” the pair is considered as one double-row angular-contact ball bearing.
When calculating the equivalent load for a unit consisting of two or more single-row radial or angular-contact ball bearings mounted
“in tandem,” the bearings are calculated individually as single-row ball bearings.
Double-row bearings are presumed to be symmetrical.
4 Permissible maximum value of F,/C, depends on the bearing design.
5 Fora = 0; F,=0and X¥=1.

487
Table 9-6 Load ratings and values of X and Y for single row, spherical roller bearings
(inch size) (C, = basic static load rating; C = basic load rating, i.e. basic dynamic
load rating)

SOURCE: Data from Bearing Technical Journal. FMC Corporation, Link-Belt

Bearing Division, Indianapolis, Ind., 1970, The data applies to Link-Belt series A 2000S.
“For F,/VF, <e, x =1.0 and y =0.

Table 9-7 Value of Y for double row, spherical roller bearings”

Dimension Series 22 Dimension Series 23 | Dimension Seis
Seis 30
30 | Dimension Series 31 Dimension Series 32
Bearing
Bore,

SOURCE: Data from Bearing Technical Journal. F MC Corporation, Link-Belt Bearing Division, Indianapolis, Ind., 1970.
“For all dimension series, X = 1.00 for KalyR <e and X¥ = 0.67 for "aiyp,>e.

488
 489
Section 9-10: Equivalent Load

Table 9-8 Ball bearing service factors, F.

Multiply Calculated Load

Type of Service
by Following Factors
S \
Ball Roller ae /
Bearings Bearings

Uniform and steady load

Light shock load 1.0
Moderate shock load 3
Heavy shock load RAG
Extreme and indeterminate shock load _

Nevertheless the following procedure, used to illustrate the selection of ball

and spherical roller bearings, is the same as that which would be used to
select a straight cylindrical roller bearing or a thrust bearing.
The equivalent load calculated from equations (9-4) or (9-5) does not
account for any shock or impact forces a bearing may experience. As a pre-
caution against possible damage or early bearing failure, itis suggested that the

Table 9-9 Guide to values of life requirements for different classes of machines

Life in Hours
Type of Machine of Operation

Instruments and apparatus that are only infrequently used

Ex.: Demonstration apparatus, devices for operation of sliding doors 500

Aircraft engines 500-2,000

Machines for service of short duration or intermittent operation, where service interruptions are
of minor importance
Ex.: Hand tools, lifting tackle in machinery shops, hand driven machines in general, farm
machinery, assembly cranes, charging machines, foundry cranes, household machines 4,000-8,000

Machines for intermittent service where dependable operation is of great importance

Ex.: Auxiliary machines in power stations, conveying equipment in production lines, ele-
vators, general cargo cranes, machine tools less frequently used 8,000-12,000

Machines for 8-hr service that are not always fully utilized
Ex.: Stationary electric motors, gear drives for general purposes 12,000-20,000

Machines for 8-hr service that are fully utilized

Ex.: Machines in general in the mechanical industries, cranes for continuous service, blowers,
jackshafts 20,000-30,000

Machines for continuous operation (24-hr service)

Ex.: Separators, compressors, pumps, main line shafting, roller beds and conveyor rollers,
mine hoists, stationary electric motors 40,000-60,000

Machines for 24-hr service where dependability is of great importance

Ex.: Pulp and paper machines, public power stations, mine pumps, ‘public pumping stations,
machines for continuous service aborad ship 100,000-200,000

SOURCE: General Catalogue and Engineering Data, SKF Industries, Phila., Pa.
 490
Chapter 9: Rolling Bearings

from Table 9-8. The factors in this table are based upon accumulated years
of experience for a variety of design conditions.
In choosing a bearing, a designer must base his selection on life and load
requirements. As is often the case, data concerning the expected life of a
bearing to be used in a particular application are not readily available. Table
9-9, based upon accumulated industrial experience, can be used to assist the
designer in selecting a reasonable design life for a 90% probability of
survival.

Example 9.1. A 309 single-row deep groove ball bearing which is to operate at
1500 rpm, is acted on by a 1890-lb radial load and a 1250-lb thrust load. The
inner ring rotates, the load is steady, and the service is continuous. Determine
the rating life and the median life of the bearing.
Solution: A 309 bearing belongs to the medium series (that is, 300 series) and
is equivalent to dimension series 03. Referring to Table 9-2, we find that for a
45-mm (that is, 9 x 5) bore bearing: C = 9120 1b, and Cy = 6730 1b.
To determine the equivalent load P, we need to know the value of X and Y
from Table 9-5.
De. e 1(1250) _ 6 186 sik é 1250
= 0.672
G 6730 VF. 1(1890)
We see that F,/VF, is larger than e, so we must obtain X and Y from Table 9-5.
(Note: X = 1 and Y = 0 for F./VF. < e). By interpolating values of iF,/Cowe
obtain Y = 1.287. The value of X is constant and equals 0.56.
Applying equation (9-4)
P= XVF, + YF, = (0.56)(1)(1890) + (1.287)(1250) = 2668 Ib
and applying equation (9-5)
P= VF = 1(1890) = 1890 Ib
Since 2668 Ib is larger than 1890 lb, the larger value is used for life calculations.
Applying equation (9-2)
G\e 9120) pee :
Lo=(=|-= = 40 million revolutions = 40 x 10° rev
IP 2668
or
40 x 10°
ig = T500 x 60 > 445 hr

median life = 5(445) = 222Shr @

Example 9-2. For the bearing in Example 9-1, how does the rating life and median
life change if the bearing is subjected to a light shock load as the outer ring is
rotating?

Solution: For a rotating outer ring, V = 1.2. Thus, we have

B 1250
= 0.560
a

VF. (1.2)(1890)
we see that F,/VF, is still larger than any value of e applicable to a single-row
ball bearing in Table 9-5; thus, we still must choose an X and Y from Table
 49]
Section 9-10: Equivalent Load

9-5. With the ratio (iF ,/C, remaining the same, ¥ = 0.56 and Y = 1.287. From
Table 9-8 and equation (9-4)

P= F(XVE + YF,) = 1.5(0.56 x 1.2 x 1890 + 1.287 x 1250) = 4320 1b

From equation (9-5)
P=FVE = 15(122 x 1890) = 3400 1b
Since 4320 Ib is larger than 3400 lb, we use P = 4320 |b for calculating the
rating life. Now, using equation (9-2)
b 12 3
Lio = (| = i; ; = 9.37 x 10° rev
P 4320
f : 9.37 x 10°
rating iter FisoOsse6o! = 104hr

median life = 5(104) = 520 hr

Note: for a 62% increase in equivalent load, the rating life was reduced by
76.5%. @

Example 9-3. As another example, consider the case of selecting a bearing for a 23-in.
diameter shaft that rotates uniformly and steadily at 1800 rpm. Due to a bevel
gear mounted on the shaft, the bearing will have to withstand a 1000-Ib radial
load and a 350-lb thrust load. Without any restriction as to the type of bearing,
select one that is satisfactory for a rating life of at least 20,000 hr.
Solution: Observing that the thrust load is greater than 20°% of the radial
load, we can eliminate the possibility of using a straight roller bearing. We can
also eliminate from consideration the tapered roller bearing. This type of
bearing is primarily used for applications where there is high thrust load and
a moderately high radial load. We will, therefore, consider using one of the
following types of bearings: (1) a single-row deep groove bearing, (2) a single-
row spherical roller bearing, and (3) a double-row spherical roller bearing. The
bearing bore diameter is selected so that it will be compatible for mounting
on the 23-in. diameter shaft (that is, the shaft diameter can be “turned down”
to fit the bearing bore and provide a shoulder for the inner ring). In this case,
we shall specify a bearing bore of 50 mm (1.9685 in.).
Part 1. Consider a 310 single-row, deep groove bearing. From Table 9-2
we establish that C = 10,700 lb, and Cy = 8010 lb. We now need to determine
the equivalent load P. To do so, the values of X and Y are needed from Table
9-5. To obtain the values of X and _Y, it is necessary to calculate iF,/C) and
find e. Thus
iF, (350)
02 424 by linear interpolation
8010 = 0.0437;
Cy = ———
—4 e =

3
a = “T1000 = 0350: > 0:2424)—

Thus, from Table 9-5, X = 0.56, Y = 1.833 by linear interpolation. Applying

equation (9-4), we obtain
P= XVF, + YF, = (0.56)(1)(1000) + 1.833(350) = 1202 Ib
Then applying equation (9-5)
P = VF, = 1(1000) = 1000 Ib
 492
Chapter 9: Rolling Bearings

Using the higher equivalent load we proceed directly to calculate rating life.
By means of equation (9-2), we get

Ge 10,700) 3
1 (| = |ne |= 705 x 10° rev

or by using equation (9-3) we obtain the rating life L;, in hours

sols) 10°(705)
bog = = = 652 hr.
10 60n\P ~ 60(1800) ae

Part 2. Consider a single-row spherical roller bearing. This bearing is

ordinarily available in bore sizes that are in inches. To fit the shaft diameter
and to be comparable with other bearings reviewed in this problem, we shall
select a spherical roller bearing from Table 9-6 with a bore diameter of
1.9680 in., an O.D. of 3.5433 in., and a width of 0.787 in. We also note that
C = 11,100 lb, and Cy = 9700 lb. Following the same procedure as with the
deep groove ball bearing, we have

VF = 0.350 < 0.40 = e (from Table 9-6)

r

and so we use X = land Y = 0

P = XVF, + YF, = (1)(1)(1000) + 0(350) = 1000 Ib

P = VF, = 10001b
Cy? P1100) +42
ibaa = 5| = |ae | Z = 3020 x 10° rev
Ba 6

or,
10°(3020)
one = 60(1 800) = 27,900 hr

Part 3. Consider a double-row spherical roller bearing of the dimension

series 22. From Table 9-1 we note that C = 14,700 1b, and Cy = 12.000 lb.
Values of X and Y are obtained from Table 9-7 for a 50-mm bearing bore

VE = O350) > OA = Y = 3.83, X = 0.67 (see footnote of Table 9-7)

la

P = XVF. + YF, = (0.67)(1)(1000) + 3.83(350) = 2010 Ib

P = VF, = 10001b
Using the larger of the two calculated equivalent loads,
b 3133)
Lio = A = ead = 758.5 x 10° rev
le 2010
or

10°(758.5)
 493
Section 9-11: Bearing Survival with Probabilities Higher than 90%

Part 4. Consider a double-row spherical roller bearing of dimension series

23. From Table 9-2, we note that C = 31,600 lb, and C, = 26,500 lb. Values
of X and Y are obtained from Table 9-7 for a 50-mm bearing bore.
F
VE = 0,350) =< O39 Y = 1.73, and X = 1.00 (see footnote of Table 9-7)
r

P= XVF_+ YF, = (1)(1)(1000) + 1.73(350) = 1605 lb

P = VF. = 1000 |b
Using the larger of the two calculated equivalent loads,

CN 31,600) 3:33 z
JOG (5 = |1605 | = 20,418 x 10° rev

or

10°(20,418)
10 = ~60(1800) = 189,000 hr

From the above calculations, it is readily seen that the single-row deep
groove ball bearing and the double-row spherical roller bearing (dimension
series 22) do not satisfy the life requirement of 20,000 hr. Therefore, they are
eliminated from consideration. Both the single-row spherical roller bearing
and the double-row spherical roller bearing (dimension series 23) satisfy the
life requirement. However, the double-row spherical bearing is 9.45 times
overdesigned. In addition, it is more expensive than the single-row spherical
roller bearing. Our choice, therefore, is the single-row, spherical bearing. @

SECTION 9-11

Bearing Survival with Probabilities

Higher than 90%
As previously indicated, the definition of rating life was based upon a 10%
failure rate (or a 90% survival rate). The AFBMA and manufacturers’
catalogs list the bearing capacities to be compatible with this definition. The
question that arises is ‘““howcan the designer select a bearing witha probability
of survival greater than 90°’? Certainly, there are many applications in
which the likelihood of failure or catalog survival rate cannot be tolerated
(for example, manned space vehicles, medical and hospital equipment, nuclear
power plant controls, and so on).
The AFBMA has recognized the need for providing reliability adjustment
factors and has published such data in its latest (1972) standards. These
data are given for reliabilities of 90, 95, 96, 97, 98, and 99 %. To avoid the need
for interpolation, we have plotted these values in Figure 9-37. This curve can
be applied to both ball and roller bearings but is restricted to reliabilities no
greater than 99 %.
 474
Chapter 9: Rolling Bearings

To illustrate the AFBMA procedure, suppose a ball bearing is designated

as 10BC02. The letters BC are an indication of the type® of bearing. The
letters BC? also denote a single-row radial contact bearing having metric
dimensions. The prefix 10 indicates a bearing have a 10-mm bore diameter,
and the suffix 02 indicates that the bearing falls within the 02 dimension
series. Reference to Table 9-1 indicates that this bearing has an O.D.
1.1811 in., and is 0.3543 in. wide. This same size bearing as a filling slot
type (in metric dimensions) would be specified as 10BLO2.
Thus, any type of bearing having the same dimensions can be specified
by the appropriate letter designation separating the prefix and suffix numbers.
As indicated, these letter-bearing type associations are not listed in this
text. The designer is advised to consult the AFBMA Standards to obtain
complete information. However, he can specify a bearing size by using the
information in this chapter. For example, he may call for the use of a 10-mm
(0.3937-in.) single-row, deep-groove ball bearing of the 02 series.
Recognizing that designers do have available for their use the catalogs
of several bearing manufacturers, we will say a few words about their coding
procedure and how it relates to the AFBMA description. Although manu-
facturers adhere to AFBMA boundary dimensions, load and life ratings
have followed the older SAE standard of designating bearings as follows:
Extra light series or 100 series'® (equivalent to dimension series 01)
Light series or 200 series (equivalent to dimension series 02)
Medium series or 300 series (equivalent to dimension series 03)
Heavy series or 400 series (equivalent to dimension series 04)
Following traditional practice, manufacturers designate bearings by a
number-letter system that contains as a core one of the above number
series. The system is otherwise unique with respect to the company concerned.
The last two digits (that is, X00) of the core number are used to indicate the
bearing bore.'! Thus, starting with a bearing bore of 20 mm, the last two
digits when multiplied by 5 yield the bore diameter in millimeters. Other
prefix or suffix numbers and/or letters that appear in the manufacturer’s
8 The AFBMA has also established a bearing type identification code. This code has not been
included in this text because of the lack of space and also because the manufacturers persist in
identifying the type of bearing by their own systems. Nevertheless, the designation described
here should be familiar to the designer because various handbooks do contain the system. Also,
if the manufacturer’s catalog is not at hand a bearing designated by this code is easily translated
by the manufacturer or supplier to his own identification code.
° If the designation had been 10BIC02, an inch-size single-row radial contact bearing having a
te-in. bore diameter would be indicated. The 02 still signifies the same dimension series.
0(1) New Departure-Hyatt calls this the 3L series. (2) Manufacturers’ catalogs also list an
extremely light series (equivalent to a diameter series 0). The numbering method for this series
is not consistent with all manufacturers. For example, New Departures-Hyatt uses 3LL00 as a
designation, and Marlin-Rockwell uses 1900. There is also lack of manufacturers’ (not AFBMA)
uniformity in designating instrument and extra small bearings.
'! This rule is not applicable to bearings with bores smaller than 20 mm.
 475
“a1a4 Paist] JOU IIe PUL Savas UOISUAWIP Sty) JOJ a10qQ WW QO] & Ie Uaq S8uLIeDq JaqJos [e>UaydS
“TL6L MOA MAN “YON OIIOSSY Sdainion[nuop Sulsvag Uo} ILs-UUy ay fo spanpunlg *AIANOS
009'8z |009'gz | SL8I'z | 9°ss |O00'ES | OOL‘6r |67691 | EF | OOP‘TE | OOS'ZZ | OOL*6I | OOS‘LI |OOS‘ZZ |009‘OZ | 008‘8I | 009'SI | 000'I 0878 | 86SZ'I | TE | 6LO'O | 67699 | OLI | ZOPLE | $6
009'zz |oor'zz | s9zo'z | v'zs | 000'br | OOr‘Or | 8hL5'I |OF | CO9'8z | 009‘0Z | OOS'sI | OOL‘91 |OOS‘OZ | 00Z‘6I | 00991 |OO9‘EL | O86 | OFI'L] II8I'I | OE | 6L0°0 | C669 | O9I €EPS€ | 06
00461 |oos‘6t | sLeet | Zeb |OOF‘LE |O09‘EE | ELIH'I |9E |OOF'TZ | 006'ST |OOS‘SI | OOI‘FI |OOL‘LI | OOE‘9I | OOF'rI | 000'ZI | 00S'8 06€9 | pZOI'I | 8z | 6200 | SS06'S | OST | S9PEE | $8
oor6l | ooe'st |OOSL'I | p'br |OOE'ZE | 000'6z | co6z'1 |E¢ |009'6I | OOL‘E! |OOG'ET | OOE'ZI | OOL‘SI | OOI‘FI |OO9'ZI |OOO‘O! | Oz8'9 | O6zs |9EZO'I | 9z |6LO'0 | BITS’S | OPI | 9orIE | 08
oo¢'L1 |006'91 |osz9°1 | Ete |0OZ'8Z |OOT'SZ |sozz'1 | TE |OOZ'8I | OO8'ZI |OOE'ZI | OOO'IT | OOO'FI | OOL‘ZI |OOF IT |OSz'6 | OIL'9 OL8'b | €v860 | SZ] 6SO'O | IBIS | OI | BCSET |}SL
ooc‘9t | oor'st |szost | vee |007‘Lz | oog‘ez |sozz1 |1€ |COSbI | OOZ‘OI | COO'ZI | OOE‘OI | 006'71 | OOI‘TI | COS‘OL | OIH'8 | 066'S | OZZ'P | GhHOO | Hz | 6500 | EICOH | Sc} 6SSL7 | OL
ooo'st |ooo'rt |ooos'1 | I'g¢ | 07°92 | OOr‘ez |sozz1 | Le | CO6'PI | 0OZ‘OI |OOO‘IT | OOF'6 | COs‘IT | OOI‘OI |006 | OL9'L | O9E'S | OLE | Ss06:0 | Ez | 6S0°0 brcLvy | OTI | 16Ssz} $9
| seer’ | soe |OOE'ZZ | OOL'8I |PZOI'T |8z | OO9‘ZI | O98 | OOL‘6 | 066'L | CDG‘OI | OST'6 | OLO.S | OS69 | 007s O6r'€ | 19980 | TZ | 6500 | LOEE'Y | OII | 779ET} 09
oos‘er |ooc‘z1
oortt |ooc‘or |szie | eee |O0€'8I | OO0ST | Er86'0 |sz | OoE‘o! | OIL'9 | C108 | OLr'9 | O86'8 | OIHL | OOS‘ | OE9'S | OES | OLO'E | 89780} Iz | OSO'O | OLEG'E OOT |} PS9T7]| ss
ozz‘6 | o€l’s | Ssegt1 | zo¢ | OOL‘rI | O0O'ZI | $s06'°0 | EZ | Or's | OLES | OLP'9 | OTIS | O97L oss’s | OL0'9 | OSr'r | O€OE OLET | PL8L0 | OZ | 6E00 | EEPS'E | 06 $8961 | OS
oz9‘g | oze'z | SL8T1 | to¢ | OO FI | OOE'II | Ss06'0 | Ez | OLO'8 OOS | O€Z'9 | OLL‘r | 0769 | Orr's | 99'S | OLO'h | O8L'E O97 | O8PL'0 | 61 | GEO'O | S9PE'E | $8 ELCEST ESP
O89. | O€h9 | SL8TI} ZOE] O09'EI | OL8'6 | $5060 |EZ | OL9'L | 099'F | O9S‘S o6t'r | 0919 | O8L‘b | OvO's | Ozs‘e | O€E'E | OEGT | L80L0] 81 | 6€00] 96FIE} 08 8PLs'l | OF
O8Z'9 | 009°S | Sz790°l | OL - as = —| ooe‘s | osr‘e | 069'r | ooe'e | o91's | ops‘e | Orr'p | OLO'E | OFL'Z | OOS'1 | £6990] LI} GEDO| 9PEBT] TL | OBLET| SE
Ops | O@I'p | SLEGO] BET] — - - — | o6‘e | ozzz | oss‘e | Oor'z | Ol6E | OzB'z | O9EE | OSz'Z | OIL‘ | OLE | 66790} 91 | 6E0'O| 6OFHT] 79 | TIBI] OF
OOL'E | 098°% | SZI8'0] 9'0z ose‘z | ozo‘t | o9s‘z | cect | 078% | 0961 | O@h'Z | 09ST | OOTZ | $06 906S0 | SI | 6€00] ZLPO'7 | ZS | EP86'0] SZ
Olpe | OPS'T | S780] 9°02 = = = = = = = = OLY Ca NOES MONG NROGSeTem ONL len site ZISSO} pI | 6€0'0} POSSI | Lp | PL8L'0| 0%
Orsz | OPsl | SL89°0| SLI = = = = = = = = ops‘t | ogt‘t | oso‘t | OOO! | ovE‘l | SPS | HZLhO}] ZI} pZOO] BPLS1| OF | £6990] LI
OE? | O€hI | 05790} 6'SI - - = = = - = = OLr‘I | 076 OzeT | 06L O6z‘I | csr I€er'0] IL | PZO'O} OSLET| SE | 90650] SI
ozs‘l | oszl | 0Sz9°0} 6'SI - - - = — Ore |) Sez O8I'l | S89 $96 zee | LE6EO| OI | PZOO] B6STI] ZE | HZLHO] ZI
Opz'l | 008 s79s‘0| € v1 - = - = = = _ oa - - $08 Oy 0S6 ooe | evseo| 6 | PZO'O| II8I1} O€ | LE6EO} OL
S8L Obr Sob 80z | OSTEO] 8 | HZOO] JECOT] 9% | EPSED] 6
sos poe 09¢ Srl = ll el OSLO) |S
9s vO€ 09¢ Srl 9SLTO} L | ZIOO] 19980] 7 | 9SL7O] L
SI 117 Sep Sil Z9ETO] 9 | TIOO]| OSbL'0} 61 | C9ETO] 9
Sip 117 - - 69610] S | ZI0'0] 667970 | 9I | 69610] S
897 871 - - 69610] S$ | TIO'O] 8IISO] El | SLSTO} +
| ae
Z Salles 1o}oWIeIG
9) )
Fe 2
uu “ul wu
Suueag [12g Suueag [eq eunesanied “ut
Sureeg [[eq
Supreag 191/0% ajauy-doaig a[suy-[jews ANS tele Suuesg [jeg ‘snipey
Surzeag 12/104 YION Suueag
12]/]0y
jeouayds TeoupuyAa }9B)U0D yoeUoD me Serre Suiuare-j[9g ial
+ [ueyds “Bur -UON
MoY-2]GnN0g 7 WPI PIM -aeinsuy aenduy EES @ UPI xe
gd eip Pp
apisino alog sueag
ZE Sauag vOIsUaWIGg ZZ Satag UoIsueUNG ZO saliag uoIsUaWIG
(q|
‘Burnes peo;
oiweuAp sIseq ‘Buljes
:q| 2 = oIseq dNe}s peo] sBuleag
°4) = Buljjoy jeipey sbuney
jo suoisuewiq
pue peo] piepuelS ajqeL
L-G
476
aiqeL7-6 piepuels suoisuawig
pue peo7 sbuljey
jo jeipey Buljjoy sHuueag

uorsuawig
Satag
£0 uolsuawig
Sauag
£7
Suueaga1og apisino vorsuauig
Saag£€
eae xew | gun eS -rgjn8uy MPI
A UPA
a :
e dU Suruaye-j12S< aeaes y9e1U09, Aa leoupurl
Moy-a1qnog
Teouayds Surusie-s]2g jeoueyds MOY-218U05 “SurYUON
snipey [12g Burseag eoeules aBuy-doaig 191104 Bursvog 191104 Supseag [eg Supeag soy Suueag
11eq uaeag Saag YIONleg
3 G a Suueag

“ur
ww “UL wu

€ Salias 1a} oueIq

t 0 SLs| 91 | 66790
| Z10'0
| S$ 69610]
s 69610
| 61 | O8bL'0 TIO'O]
| 9 C9ETO]
9 CIETO]
L 9SLTO}
8 Osl€o]

6 £PSE0|
O1 LE6E'0}
| SE | O8LEI pz0'0|Il] I€€p'0] ca = SPs| OOr'l
ZI PZLr'O]
| LE| LOSHI
| 6E0'O
| Cl PTLb'0] OES| Joes! OPO'T
| 0891
SI 90650}
| th| SESMI
| 6E0O El] 8IISO] 065| }OSol OZZI
| O96'T
LI €699°0]
| Lb| POS8'l
| GEO'O
| HI TISSO| 078| OLI‘Z
| OLb'l
| Ove’c =< a = — = 6I| O8PL'0 S68| oos‘z = Oszs8'0 = = 0697
| OE9'€
OF HL8L°0}
| TS| ZLPO'7
| 6€0°O
| SI 906S'0| 006| OST'Z
| OSL*T
| OSL‘
| OZ6'I
| OOO'€ - ~_ aaa IZ 89780]
| O6IT
| OIZ‘E - OSzEs'O
| OSL‘I
| OSL‘Z
| OOZ'E
| Olt
sz Ergoo|
| 79 | Gorr'z 6E0'0|21] £699°0}
| ose!
| forte o6€'z
| |099'e ozs'z
| Ozz'r
| oz8'z
| oEI's
| O00'9
| 00L'9
| bz 6rr6'0}
| ogo'r
| 0lzp
| — CODD!
| O6E'Z
| 099'E
| O9€'b
| 0995's
OF | LISI
| cL OPEs'z| 6E0'0}]
| 61 O8b'0]
| ObL'I
| OOL‘E
| OEE| OS8°p
| OFB'E
| OLE'S
| Ovo'E
| O9L‘9
| OOS'8
| 00S‘6
| LZ O€90'I]
| osz'z
| 02's
| —
SE | Oszet cvgt't
| OpE'e
| Oss'p
| ooI'9
| or’
| Og| 96rI'e 6S0':0|Iz] 89780]
| OIZ'z
| JOSE’ OZO'P
| OSL's
| Oz9'p
| OPED
| OLE'S
| OE8's
| 0086 OO8‘OI}
| 1€ SOzz'I]
| |oL8‘z 0789
| — Osuet
| O20
|
OF BbzS'1|
| 06| EErs'e OS'S
| ores
| 098's
6so'0|€z| sso6'o|
| |lorz'z orI‘s
| OzO'Ss
| OFO'L
| OLL‘s
| OFL'L
| OrE'9
| OOE'oI
| OO6'ZI OOL'EI]
| CE Zoez'1]
| OEs‘e
| seize
| COL'9I
| O0r'0z suEP't
} OZO'S
| OFO'L
| OLI‘6
| OO8'OI
Sb | LIZ“ OO}| ozee'e
| 6S0'0sz| Eps6'0]
| ogs‘e
| 009‘
| 0€2'9
| OzI‘6
| oEc‘z
| OOI‘OI
| o€o'6
| OOz'rI OOELI} OOO'LI]
| 9E ELIP'1]
| OLb'p
| O€b'6
| 009'07
| 000'Sz sz9s"1
| OEL‘9
| OZI'S OOE'TZI]
| OOIFl
OS| $8961 OL1}| Loge'r
| 6200 LZ] O€90'1]
| O€6'€
| OIS'L
| OLO'8 OOL'OI}
| 0076
| OO8'IT O01| OO0'LI
| OOE'8I OOE‘6I|
| OP 8bLS1]
| O8Z‘'s
| OOIIT
| OOS*9Z
| 009'1E OOSL'I
| O10'S
| OOL'OI
| OO9FI
| 00S*9I
Ss PS9I'Z|Oz]| pHZL'b
| 6L0°0 6z| LIFI'1]
| O90's
| OO8's
| COr.6
| OOF'ZI
| OOS‘OI
| 009‘EI
| OO9'EI
| OOL'0z OO9'IZ] OO‘ZZ]
| EP 6Z69'I]
| OOE'9
| OOO‘!
| 009'E
| 008'LE SZE6'I
| OOF6
| OOPZI
| OOT'LI
| 000°6I
09| zz9¢7 o€1|| Igtt's
| 6200te} sozz't]
| og6's
| |lo98°6 O06‘! OOI'FI}
| OOS‘zI
| DO9'sT
| OOS‘sI
| OOE'ez
| OO0'sz 000'9Z]
| 9b L181]
| OObL
| OOT'ST
| OO9'gE
| O0E'bh OszI'z 0060}
| OOLbI
| 006'a1
Iz 008°
so Lossz}
| OI| BIIS"S
| 6L0'0
| €€ zo6z'1|
| 009'9
| OOZ'oI
| 00S'ZI
| 000'91
| OOb‘bI 0O9'LI|
| OOI‘LI
| O09‘sz
| OOS‘oE COO'TE]
| 8h B68Bl]
| O89'R
| CO9*9I
| OOL'Zp
| OOI'Sh STIET = = 006'7Z
| 009*rz
OL| 6SSL'7 OSI]
| $S06'S
| 6L0'0 SE] O8LE1]
| O86'L 006'ZI
| 00Z'tI
| 000'8I
| OOP‘91 CO8‘6I|
| OO8‘0z
| OOI‘OE
| OO0'PE OOS‘PE]}
| IS 6L00'Z}
| 000'0I
| 0O6'8I
| OOS'6r
| 008'ss 000S°2 as = 000'97
| OOL‘LZ
SL | 8zs67 091]
| Z66z'9
| 6200 LE} L9SPI]
| Oz9'8
| OOL‘EL
| OOT‘9I
| 009'6I
| OOS‘gI OOS'IZ]
| OO9'sz
| OOS'9E
| O00'8E OOO'6E|
| Ss PS9I'z]
| OOS‘OL
| 009‘6I
| 00Z'9S
| 0029 SL897! om = OOF'6z
| 00Z*OE
08| 96rI€ OLI|
| 6769°9
| 6L0°0
| 6€ PSEST|
| 066 OOE'sI
| OO0'SI OOE‘IZ}
| OOL‘0z OOE'ez| 009'sz|
| OOS‘9E
| OO0‘EF 0O0'Eb]
| 8S SERZ'Z}
| OOO'EI
| 0O6'EZ
| OOF'P9
| OOr'IL SL89T: = = 006‘7E
| TE OOL
$8| Sore’ O81]
| 9980°L
| 860°0Ib} }\ZHI9'I] 006'OI
| 00691
| OOT*OZ 006'ZZ|
| OOI‘EZ 00%'Sz]
| 00S‘OE
| 006'%p — 09| Zz9E"Z
| OO8‘EI
| OOE'Pz
| 008'89
| 00b‘9L OSL8T, = — OoO'zE]
| OOlze
06| EEbS'EOI]| EO8r'L
| 860°
| Eb | 6269°I 009'ZI}
| 00Z'0Z
| OOE‘ZZ OOL‘hZ]
| 009'Sz
| 000'LZ
| OOE'EE
| 00S‘9b — +9| L6IS'Z
| COPS
| OOF‘9Z
| OOT'S8
| 009'06
| O'EL
|| OSL8°C v= = 009‘SE
| 009‘rE
:ADNNOS spsopuvIsS
fo ays uorr44-11Up
Buisvag Ssasnjo0[nuopW
MIOSSY
“UO MAN MOK “TLOL
 JOU SI puke asoq Julreaq WUI-)] | B YIIM sUIsaq |¢ Salles UOISUaWIP 10) sdULRAG J9]]O1 [eoLIaYds|
JOU SI pue asoq BulsvAag WU-0Z | F YIM SUIaag C€ SolseS UOIsUaUNIP Jo) s8ULTeAQ Jo]JO1 [eoLIaydg|
“TL61 “OK MAN “UOUDIIOSSY Siainjov[nuoy Supwag uoNIuUy-uUp ayi fo spAvpunlg :ADUNOS
006‘LZ |
008‘9z 00s‘ZI 00r‘O!
00L‘9Z |
000'Sz 00S‘*7I 00S‘OI
OOP | sz
O0E"EZ ; 00771
008‘IZ OOT‘0Z
| 00Z‘OI
008‘0z 009°81
| 0566
000°SI
006‘91
OrI's
|
000°LI |
000'SI ¥ 089°L
f
TI
OOE
O0z‘EI
0679
|
00r'Ol i Ov6's
OOS*II |
OFS*6 F 008°S
4
Olt'r
0S0°6
O9T'L
O1s‘9 i Ori‘
Olr's i O6r'e
Oly Ostve
ool'e
Opr'z
1 ul ul un

MOY-2[BUIS
MOY-219N

yoeUOD
3
r ie Suueag 19]]0y 1 Surreed . Suuesg TPE |[eq =ni te

-rejnsuy
1aT]OY uLeag Suleag 12]]0y

Md
aS apuy-jews aa0019-daaq

es
pea
4 Teoueyds 191T , [eoueyds

UIPIM

UPI
xe

xe
uolsuauiig
apisino

apisino
uolsuaWIg
uolsuawiq

qq
J
7

Suuveg
sales
saueg

p
Saves

a1og
QI
O€
I€
] satiag 1ajoweIg Q satiag 19}WeIG
sBuueeg Buijjoy jeipey jo sbunjey peo pue suoisuawiq piepuels ¢€-G BIGeL
 478
Chapter 9: Rolling Bearings

designation of a bearing are used to indicate features such as bearing type,

self-lubricating, type of seals, snap ring for mounting, material differences,
and so on.'”
As an illustration, a New Departure-Hyatt bearing having a basic bearing
number 3212 indicates that it belongs to the 200 series (that is, light series),
has a bore diameter of 60-mm (12 x 5), and is a Conrad (that is, deep
groove) single-row bearing open on the sides. The last part of this description
is based on the use of 3000 by New Departure-Hyatt to indicate this type
of bearing. This same bearing is designated by the Marlin-Rockwell
Company as 212-S, where the 212 has the same meaning as described above.
However, the S is used to indicate a single-groove, Conrad type bearing.'*
The same bearing is designated by SKF as bearing number 6212. Notice
that SKF uses 6000 as the basic number to indicate single-groove, Conrad
bearings. Nevertheless, the important numerical characteristic to observe
is that in all cases the number 212 appears. This common basic number
makes it simple to substitute the bearing of one manufacturer for that of
another. As a service, many companies list bearing conversion tables at the
end of their catalogs. This same bearing designated by AFBMA standards
would be 60BC02.'*
It is obvious from Figure 9-34 that the number of bearings represented
by the dimension series is very extensive, and to reproduce them here is
not feasible. Therefore, Tables 9-1, 9-2, and 9-3 represent portions of the
more widely used bearings, and the reader is advised to consult the AFBMA
Standards or a manufacturer’s catalog for a complete listing.

SECTION 9-5
Bearing Tolerances
The Annular Bearing Engineers’ Committee (ABEC) of the AFBMA has
established five basic grades of ball bearing precision. These grades are
ABEC 1, 3, 5, 7, and 9. Grade | has the widest tolerance; as the numbers
increase, the tolerance becomes smaller. For most normal applications,
ABEC 1 is satisfactory. Where closer tolerances are needed, ABEC 3 and 5
can be specified. ABEC 7 and 9 are super precision grades reserved for
'? Here again, there are some exceptions to the practice of bearing designations by manufacturers.
For example, the New Departure-Hyatt basic bearing number 3212 becomes 7512 with one
shield, Z 97512 with one shield and one seal, 477512 with a snap ring and two shields, etc. Note
that the only numbers remaining of the basic bearing numbers are the last two digits (i.e., 12
which indicates a bearing bore of 60 mm).
'3 The Marlin-Rockwell Company is one that continues to use the same basic number regardless
of special features. For example, a 212-S bearing with one shield is designated as 212SF, as
212SF2 with one shield and one seal, as 212SFFG with two shields and a snap ring, etc.
'* The reader using the standard dimension and load rating tables of the AFBMA listed in this
text would, in selecting this bearing, call for a ‘‘60-mm (2.3622-in.) bore, single-row deep-groove
(i.e, Conrad) ball bearing of the 02 dimension series.”
 479
Section 9-6: Rolling Bearing Materials

applications requiring extreme accuracy. Such applications are found in

the design of instruments, aircraft controls, servosystems, and machine tools.
Considering the discussion in Chapter 4 concerning tolerances and costs,
the designer must be able to justify specifying high ABEC grades. For
example, in the bearing bore range of 11 to 16 inclusive, the ABEC tolerances
are as follows:

ABEC Grade Tolerance

1 +0.0000
—0.0006

+ 0.0000
3 —0.0004
+0.0000
5 —0,0003
7 + 0.0000
—0.0002

+0.0000
e —0.00015

From this list, it is readily seen that the ABEC grade 9 has only one quarter
the allowance of the ABEC grade 1. Reference to Figure 4-69 shows the
relative time to produce the tolerance for ABEC 9 as 11.5, and the tolerance
for ABEC 1 as 7.9. Thus, to produce grade 9, a 45.6% longer period of
time (that is, increased cost) is required than to produce grade 1.
Manufacturers’ catalogs contain the complete ABEC tolerance tables
both for bore diameter and radial runnout. Tolerance grades have also
been established for cylindrical roller bearings by The Roller Bearing
Engineers’ Committee (RBEC) of the AFBMA. For cylindrical roller bear-
ings, there are only two grades, RBEC 1 and 5. These grades have the same
tolerance values that have been assigned to ABEC 1 and S. Tapered roller
bearing tolerances have also been graded. These grades are 4, 2, 3, 0, and 00
with 4 having the lowest precision and 00 having the highest precision.'®

SECTION 9-6
Rolling Bearing Materials
Ball bearings (except for the retainers, seals, and ball separators) are
commonly made of SAE 52100 (or AISI E-52100), a high carbon chromium
steel that is through hardened to Rockwell C 58-65. Due to the loss of
hardness with increasing temperatures, SAE 52100 cannot be used where
the temperature is above 350°F.
'S See Section 1 of The Timken Engineering Journal, The Timken Roller Bearing Company,
Canton, Ohio.
 480
Chapter 9: Rolling Bearings

Bearing steels are produced by the vacuum-melting (VM) process and, in

particular, by the consumable electrode vacuum-melted process (CVM) in
contrast to air melted steels. Steel produced by these methods is more
uniform in structure and contains fewer impurities. As a result, the reliability
of bearings has been improved.
High temperature bearing problems are not uncommon and in the past
years have served to prevent the development of both industrial and military
hardware. However, continuing research has led to the development of
materials capable of maintaining a through hardness greater than Rockwell
C 58 at higher than normal temperatures.
Figure 9-35 illustrates how some materials vary with temperature. The
maximum operating temperature of each material is that point at which the
curve crosses the Rockwell C 58 line.
For mildly corrosive applications, it is quite common to use grade 440C,
a martensitic stainless steel. Notice that in Figure 9-35 the anticorrosion
steel is “‘restricted” to an operating temperature no higher than 300°F.
At a temperature above 300°F, its hardness drops below Rockwell € 58
and the raceways may begin to “‘brinell.”

Figure 9-35 Hardness versus

temperature of some through hardened
bearing steels. [From Machine Design,
1970 Bearings Reference Issue. The
Penton Publishing Co., Cleveland, Ohio.]

Hardness,
C
Rockwell

0 200 400 600 800 1000

Temperature, °F
 48]
Section 9-6: Rolling Bearing Materials

Roller bearings are usually fabricated of case hardening (that is, carburiz-
ing grade) steels. These steels are surface hardened to Rockwell C 58-63 to
a depth of no greater than 0.015 in., past which there is a softer core with a
hardness of Rockwell C 25—40. Carburizing grades (for example, AISI 4620)
are limited to an operating temperature no higher than about 300°F.
Continuing metallurgical research, however, has led to the development of
carburizing grade bearing steels that are suitable for higher operating
temperatures. Figure 9-36 shows the hardness-temperature plot of two such
steels in comparison with AISI 4620. These curves illustrate the marked
increase in operating temperatures attainable by special steels.
Ball separators (that is, cages) are normally fabricated of 1010 steel.
However, for use at elevated temperatures (up to 700°F), separators made
of M-1, S-Monel, 440C, polyimide polymer (under 700°F), cobalt alloy,
and copper alloy (for example, bronze or silver-plated silicon iron bronze)
have been successfully used.
For mildly corrosive environments it is customary to use grade 302 or
grade 410 stainless steel as ball separator material. Most angular-contact
bearings and high speed bearings use ball separators made of cotton woven
fabric impregnated with a phenolic resin.
Seals used to retain the lubricant in “‘self-lubricating” bearings are stamped
from a low carbon steel (for example, 1010). Depending upon the manu-
facturer and the particular bearing, the seals are either completely molded
into a synthetic rubber compound or just edged with this material. Also a
seal made of felt pressed between two metal rings is quite popular.
Shields that help both to retain grease and to prevent chips, dirt, or large
particles from entering the bearing are also made of a low carbon steel
which is crimped into a groove in the outer ring and maintains a close
running fit with the inner ring.

o>)Oo

oaee}

olo

Rockwell
Hardness,
C

100 200 300 400 500 600 700 800

Temperature, °F

Figure 9-36 Hardness versus temperature ofthree carburized bearing steels.

[Courtesy The Timken Company.]
 482
Chapter 9: Rolling Bearings

Where extreme temperature (for example, 1500°F) and environmental

conditions have precluded the use of available stock bearings, materials
such as alumina, zirconium oxide, titanium carbide, pyroceram,'° and so on,
have been used in place of steels.

SECTION 9-7
Rolling Bearing Friction
Bearing friction is greatly influenced, among other things (for example,
hysteresis, slip), by the coefficient of friction between the rolling members [3].
Table 9-4, based on years ofaccumulated test data, lists the average coefficient
of friction () for several types of bearings measured at the shaft surface.
[tis importar
[ 1] 1ze
Ze h Pe, ve ole 9-4 req

Table 9-4 Average coefficients of friction

Starting Running

Type of Bearing Radial Thrust Radial Thrust

Ball bearing 0.0025 0.0060 0.0015 0.0040

Spherical roller 0.0030 0.120 0.0018 0.0080

bearing

Cylindrical roller 0.0020 => 0.001 1 az

bearing

SOURCE: Bearing Technical Journal. FMC Corporation, Link-

Belt Bearing Division, Indianapolis, Ind., 1970.

Actually, the greatest contribution to bearing friction is made by the seals

which can produce several times greater friction than the bearing torque.
Where the designer feels that the total bearing friction (that is, including seal
friction) is a significant factor affecting his design, then consultation with the
manufacturer becomes necessary.
Although insignificant in many designs, the horsepower lost due to
frictional torque is

hp Tn ==
uF, dn
:
P= 63,025 126,050 ae
where

T = frictional torque, Ib-in.

n = Shaft speed, rpm
d = bearing bore, in.
F, = radial bearing load, |b
'© Product of Corning Glass Company.
 483
Section 9-9: Life, Rating Life, and Basic Load Rating

SECTION 9-8
Basis for Bearing Failure
Carefully installed bearings that are properly lubricated, kept free of
foreign matter, and are not overloaded will fail by fatigue. Evidence of this
failure appears as spalling or flaking of a raceway or rolling element(s). As
the rolling elements revolve within the raceways, they are subjected to a
repeating contact (that is, Hertzian) stress. In addition, those areas of the
raceways that support the shaft load as the rolling elements pass over them
are also subjected to a repeating contact stress. Because these stresses are
greater than the endurance limit of the material, it is obvious that the
bearing will have a limited life. In fact, if one considers that pure rolling does
not occur in the bearing but that some sliding of the balls or rollers takes
place, it is readily understood why particles of metal are removed from the
raceway surfaces. The removal of this metal is called spalling and, when this
condition arises in a bearing, the bearing has failed.
The statistical analysis concerning the fatigue failure of rolling bearings
presented in 1947 by Lundberg and Palmgrem [9, 10] provided the ground
work for the adoption in 1950 by the AFBMA of a “Standard method of
evaluating the load ratings of annular ball bearings.’ This standard has also
been approved by the American National Standards Institute (ANSI
B3.11—1959) and is the basis for selecting a bearing as described in this text.
A similar standard for evaluating roller bearings was also adopted.

SECTION 9-9
Life, Rating Life, and Basic Load Rating
Notwithstanding modern manufacturing methods and the use of quality
control procedures, no group of seemingly identical bearings is exactly alike.
Each may differ, however slightly, ee in suHaee finish, seuaeeee
of balls: and so on.les a Ss no two bez the same fam:

ct »t 1 Thus, the life aie Coane iS

established on a purely statistical basis (see reference [3]) and the AFBMA
has established the following definitions to ensure a uniform understanding
and procedure when selecting a bearing :
1. The life’’ ofa radial ball bearing is the number of revolutions (or hours,
at some given constant speed) that the bearing runs before the first
evidence of fatigue develops in the material of either ring or in a ball.
2. The rating life of a group of apparently identical radial ball bearings,
thrust ball bearings, radial roller bearings, or thrust roller bearings is

‘7 This same definition is also applicable to thrust ball bearings, radial roller bearings and thrust
roller bearings by just replacing “ball” with “roller,” and “ring” with “washer.”
 484
Chapter 9: Rolling Bearings

defined as the number of revolutions (or hours at some given constant

speed) that 90% of a group of apparently identical bearings will
complete or exceed before the first evidence offatigue develops. (Rating
life is commonly referred to as L,, life, B-10 life, or minimum life.) As
presently determined, the median life'® that this same group ofbearings
would complete or exceed is 5 times the rating life (that is, median
lifev— Sic, )
3. The basic load rating’® (also called the dynamic load rating, basic
dynamic capacity, and the specific dynamic capacity) of radial ball
bearings is that constant stationary radial load which a group of ap-
parently identical ball bearings with stationary outer rings can endure
for a rating life of 1 million (that is, 10°) revolutions of the inner rings.
In single-row angular-contact bearings, the basic load rating relates to
the radial component of the load which results in a purely radial dis-
placement of the bearing rings in relation to each other. Load ratings,
if given for specific speeds, are to be based on a rating life of 500 hr. (This
is equivalent to running a bearing for 500 hr at 334 rpm or for a total of
10° revolutions. On the basis of median life, this is equivalent to having
run a bearing for 2500 hr at 334 rpm for a total of 5 x 10° revolutions.)
4. The basic static load rating!° is defined as that radial load which corre-
sponds to a total permanent deformation of rolling element and race (or
washer) at the most heavily stressed contact of 0.0001 of the ball or roller
diameter. In single-row angular-contact ball bearings, the basic static
load rating relates to the radial component of that load which causes a
purely radial displacement ofthe bearing rings in relation to each other.

Basic static load rating

The basic static load ratings for different types of bearings, symbolized by
Co, are given in Tables 9-1, 9-2, and 9-3. The value of Cy depends on the
bearing material, the number of rows of rolling elements in a bearing, the
number ofrolling elements per row, the bearing contact angle (see Figure 9-8),
and the ball or roller diameter. Those readers wishing to know how C, is
calculated should refer to Section 9 of reference [7] for specific equations and
data.
Normally the basic static lone ey: has little influence in the see of
a rolling bearing. However if i de
patie soon escicanEMRAER SR
tienmere This is Piticularly
‘8 Various manufacturers’ catalogs and texts mistakenly call this “average life.” The designer
must be careful in using this term since this usage is quite prevalent. The AFBMA defines
average life of an antifriction bearing as the summation of all bearing lives in a series of life
tests divided by the number of life tests. Some manufacturers define average life as being equal
to 7 times the rating life.
'® This same definition applies to radial roller bearings, thrust ball bearings, and thrust roller
bearings. However, for thrust bearings, the loads are not radial but are thrust loads, and the
word “‘ring”’ is replaced by the word “‘washer.”
 485
Section 9-10: Equivalent Load

true where the load exceeds Cy. Where operating speeds are very low, higher
loads than C, can be allowed but should be checked with the bearing
manufacturer.

Basic load rating

The basic load ratings for different bearings, symbolized by C, are listed in
Tables 9-1, 9-2, and 9-3. The value of C depends on the same factors as those
which determine Cy, except for an additional parameter concerning the
loading geometry. As with Cy, Section 9 of reference [7] contains the specific
relationships and data needed to compute C.
The basic load rating C enters directly into the process of selecting a
bearing, which will subsequently be explained.
Extensive testing of rolling bearings and subsequent statistical analysis
have shown that load and life of a bearing are related statistically [3]. In
particular this relationship is defined by the expression

Lo = (5 (9-27
b

where L,, = rating life in millions of revolutions (that is, the number of
revolutions resulting in 10% failure); C = basic load rating in pounds
(obtained from Tables 9-1, 9-2, 9-3, or a manufacturer’s catalog); P = equiva-
lent load in pounds (see Section 9-10); b = 3.0 for ball bearings”! and 42 for
roller bearings.
In terms of hours oflife, equation (9-2) becomes
102 Fe\2
NE is GOn (| (9-3)
where n = rotational speed in revolutions per minute.
Again, we urge the designer to read carefully that portion of the manu-
facturer’s catalog that explains how load rating and life are established.
Although most manufacturers adhere to the recommendations of the
AFBMA, there are some exceptions. For example, New Departure-Hyatt
list their load ratings on an average (that is, median) life of 3800 hr instead
of 2500 hr as specified by the AFBMA. Regardless of catalog variations,
however, the designer can always convert the data to the AFBMA standard.

SECTION 9-10

Equivalent Load
As denped bythe URS the eq 1\

20 Equation (9-2) is not applicable to filling slot type bearings.

21 New Departure-Hyatt uses an exponent of 4.
 456
Chapter 9: Rolling Bearings

In many applications, bearings have to carry a load composed ofaxial and

radial components. In addition, they are sometimes required to operate with
a rotating outer and a stationary inner ring. It then becomes necessary to
express these conditions by an equivalent load satisfying the above definition.
The expression used to define this equivalent load, which is applicable to ball
(that is, radial and angular-contact) bearings and roller bearings (not
applicable to filling-slot bearings) is

P= XVF.+ YF, (9-4)

where

P = equivalent load, lb
F, = radial load, lb
F = thrust (that is, axial) load, lb
V =a rotation factor: 1.0 for inner ring rotation, 1.2 for outer ring
rotation. For self-aligning ball bearing use 1 for inner or outer ring
rotation.
X = aradial load factor (see Table 9-5, 9-6, or 9-7 as applicable).
Y = a thrust factor (see Table 9-5, 9-6, or 9-7 as applicable).
Looking at Table 9-5, we observe that the radial factor X can in some cases
be less than unity. Thus, it is possible to have a combination where, the
if

Equations (9-4) and (9-5) are also used to determine the equivalent load
for single-row and double-row spherical bearings. For single-row spherical
, values of X and Yare found inTable7. ——
bearings, values of and
X Yare found in

Equivalent loads for thrust

ball bearings and thrust roller bearings
can be
calculated from equation
(9-4) where V is unity.
Space limitations make it impossible to list the V factors for straight
cylindrical roller bearings, the X, Y factors for both types ofthrust bearings,
and the respective bearing tables which give the values for Cy and C (that is,
tables like Tables 9-1, 9-2, and 9-3). Such information is readily available in
the AFBM A Standards, engineering handbooks, and manufacturers’ catalogs.

2? For those straight cylindrical roller bearings that have a locating flange or flanges (see
Figure 9-18), the maximum thrust load should not exceed the value determined by, F, < 0.2F,.
Table 9-5 Factor X and Y for ball and roller bearings
Single-Row Bearing” Double-Row Bearing?

Contact Angle

and e val-
ues appli-
cable to
single-row
radial- |
contact
bearings

SCOHNWAUIO
S=OvuR=—vN—-&

So0-Lhbee
CONNOWOON ae
Sat
ee
NDNNADRWOO—NW
OU
WWANWO—

Self-aligning ball bearings

[aso] oars [rofoatcorefoas

[oscar] ene
Roller-bearings, self-aligning, tapered>

SO CR
SOURCE: Frank W., Wilson (ed,): Tool Engineers Handbook, 2nd ed, McGraw-Hill Book Co., New York, 1959,
NOTE: Values of X, Y, and e for a load or contact angle other than shown are obtained by linear interpolation,
1 Co is the static basic load rating; iis the number of rows of balls (not used for radial contact bearings),

When calculating the equivalent load for a unit consisting of two similar single-row angular-contact ball bearings in a duplex mounting,
“face to face” or “back to back,” the pair is considered as one double-row angular-contact ball bearing.
When calculating the equivalent load for a unit consisting of two or more single-row radial or angular-contact ball bearings mounted
“in tandem,” the bearings are calculated individually as single-row ball bearings.
3 Double-row bearings are presumed to be symmetrical.
4 Permissible maximum value of F,/Cg depends on the bearing design.
Fora =0;F,=Oand X=1,

487
Table 9-6 Load ratings and values of X and Y for single row, spherical roller bearings
(inch size) (C, = basic static load rating; C = basic load rating, i.e. basic dynamic
load rating)

Assembly
Width,

1.0000 He 2.5000} 4. R
1.1250 b : I 2.6250} 4.4677 | 0.948 3 P 18600
1.1875 | 2.6250] 0.812 0.48 1.25 5590 7170 } 2.6250] 5.0000} 1.094 | 0.42 1.44 | 20300 |} 22800
1.2500 | 2.6250] 0.812 | 0.48 1.25 | 5590 7170 | 2.7500} 5.1172} 1.062 | 0.33 1.59 | 20700 | 21700
1.3775 | 2.8345] 0.670 | 0.44 1.37 | 6300 7640 | 2.7553} 5.9045 | 1.188 | 0.40 1.49 | 26000 30000
1.5000 | 2.8345} 0.670 | 0.44 1.37 | 6300 7640 | 2.9522] 5.1172} 1.062 | 0.38 1.59 | 20700 21700
1.5743 | 3.1496] 0.828 | 0.47 1.27 | 7470 9130 } 3.0000} 5.1172] 1.062 | 0.38 1.59 | 20700 21700
1.7500 | 3.3464} 0.812 | 0.45 1.34 | 8650 10500 | 3.0000] 5.9045 | 1.109 | 0.38 1.57 | 25800 26200
1.7712 | 3.3464] 0.812 | 0.45 1.34 | 8650 10500 | 3.3457| 5.9045] 1.109 | 0.38 1.57 | 25800 26200
1.9375 | 3.5433] 0.787 | 0.40 | g.40]} 1.49 | 9700 11100 } 3.4375| 6.0000} 1.562 | 0.43 | 0,40 | 1.39 | 33000 | 31700
1.9680 | 3.5433] 0.787 | 0.40 1.49 | 9700 11100 | 3.5000} 6.0000} 1.562 | 0.43 1.39 | 33000 | 31700
1.9680 | 3.9370] 0.828 | 0.45 1.34 | 13600 15800 } 3.5100} 6.0000} 1.562 | 0.43 1.39 | 33000 31700
1.9680 | 4.3299] 1.063 | 0.48 1.24 | 16800 19200 | 3.5425] 7.4792] 1.419 | 0.38 1.60 | 50500} 48100
2.0000 | 3.9370] 0.828 | 0.40 1.49 | 12900 13800 | 3.6250} 6.0000} 1.562 | 0.43 1.39 | 33000 31700
2.1648 | 3.9370| 0.828 | 0.40 1.49 | 12900 13800 | 3.6250| 7.0866] 1.339 | 0.36 1.67 | 39700 40600
2.1875 | 3.9370] 0.828 | 0.40 1.49 | 12900 13800 } 3.7402] 7.0666] 1.339 | 0.36 1.67 | 39700 40600
2.2500 | 3.9370] 0.828 | 0.40 1.49 |12900 | 13800 | 3.9375] 7.0856] 1.495 | 0.36 1.67 | 39700} 40600
2.2500 | 4.3307] 0.866 | 0.38 1.60 | 15500 16400 | 4.0000} 7.4792} 1.419 | 0.38 1.60 | 50500 4800
2.3616 | 4.3307 | 0.866 4.1331] 7.4792 48100
2.4375 | 4.3307] 0.866 4.7236 59500

SOURCE: Data from Bearing Technical Journal. FMC Corporation, Link-Belt

Bearing Division, Indianapolis, Ind., 1970. The data applies to Link-Belt series A 2000S.
“Por F,/VF, <¢, x =1.0 and y=0.

Table 9-7 Value of Y for double row, spherical roller bearings*

Dimension Series 22 Dimension Series 23 Dimension Series 30 Dimension Series 31 Dimension Series 32
Bearing
Bore, 7am
Fi e
mm
[es

SOURCE; Data from Bearing Technical Journal, ¥ MC Corporation, Link-Belt Bearing Division, Indianapolis, Ind., 1970,
“For all dimension series, X = 1.00 for Falyp, <e and X¥ = 0.67 for Falyp,>e.

488
 489
Section 9-10: Equivalent Load

Table 9-8 Ball bearing service factors, F,

Multiply Calculated Load

by Following Factors
Type of Service \ YON ON es

Ball Roller <> 1/ \OC <

Bearings Bearings

Uniform and steady load

Light shock load
Moderate shock load
Heavy shock load
Extreme and indeterminate shock load _ OO
OO

Nevertheless the following procedure, used to illustrate the selection of ball

and spherical roller bearings, is the same as that which would be used to
select a straight cylindrical roller bearing or a thrust bearing.
The equivalent load calculated from equations (9-4) or (9-5) does not
account for any impact
shock or forces a bearing may experience. As a pre-
caution against possible damage or early bearing failure, it is suggested that the

Table 9-9 Guide to values of life requirements for different classes of machines

' : Life in Hours

Type of Machine of Operation

Instruments and apparatus that are only infrequently used

Ex.: Demonstration apparatus, devices for operation of sliding doors 500

Aircraft engines 500-2,000

Machines for service of short duration or intermittent operation, where service interruptions are
of minor importance
Ex.: Hand tools, lifting tackle in machinery shops, hand driven machines in general, farm
machinery, assembly cranes, charging machines, foundry cranes, household machines 4,000-8,000

Machines for intermittent service where dependable operation is of great importance

Ex.: Auxiliary machines in power stations, conveying equipment in production lines, ele-
vators, general cargo cranes, machine tools less frequently used 8,000-1 2,000

Machines for 8-hr service that are not always fully utilized
Ex.: Stationary electric motors, gear drives for general purposes 12,000-20,000

Machines for 8-hr service that are fully utilized

Ex.: Machines in general in the mechanical industries, cranes for continuous service, blowers,
jackshafts 20,000-30,000

Machines for continuous operation (24-hr service)

Ex.: Separators, compressors, pumps, main line shafting, roller beds and conveyor rollers,
mine hoists, stationary electric motors 40,000-60,000

Machines for 24-hr service where dependability is of great importance

Ex.: Pulp and paper machines, public power stations, mine pumps, ‘public pumping stations,
machines for continuous service aborad ship 100,000-200,000

SOURCE: General Catalogue and Engineering Data, SKF Industries, Phila., Pa.
 490
Chapter 9: Rolling Bearings

from Table 9-8. The factors in this table are based upon accumulated years
of experience for a variety of design conditions.
In choosing a bearing, a designer must base his selection on life and load
requirements. As is often the case, data concerning the expected life of a
bearing to be used in a particular application are not readily available. Table
9-9, based upon accumulated industrial experience, can be used to assist the
designer in selecting a reasonable design life for a 90% probability of
survival.

Example 9.1. A 309 single-row deep groove ball bearing which is to operate at
1500 rpm, is acted on by a 1890-Ib radial load and a 1250-lb thrust load. The
inner ring rotates, the load is steady, and the service is continuous. Determine
the rating life and the median life of the bearing.
Solution: A 309 bearing belongs to the medium series (that is, 300 series) and
is equivalent to dimension series 03. Referring to Table 9-2, we find that for a
45-mm (that is, 9 x 5) bore bearing: C = 9120 lb, and Cy = 6730 |b.
To determine the equivalent load P, we need to know the value of X and Y
from Table 9-5.
iF, 1(1250) Fi 250
Gouul6i30e
= =
So
Lal
atv a tES50)
— =
We:
if

We see that F./VF. is larger than e, so we must obtain X and Y from Table 9-5.
(Note: X = 1 and Y = 0 for F./VF. r < e). By interpolating values of iF,/C owe
obtain Y = 1.287. The value of X is constant and equals 0.56.
Applying equation (9-4)
P= XVF, + YF, = (0.56)(1)(1890) + (1.287)(1250) = 2668 Ib —
and applying equation (9-5)
P = VF, = 1(1890) = 1890 Ib
Since 2668 Ib is larger than 1890 lb, the larger value is used for life calculations.
Applying equation (9-2)
G\P 9120) ae .
Laa= | = | |= 40 million revolutions = 40 x 10° rev
P 2668
or

40 x 10°
IE, =>
oy ee
median life = 5(445) = 2225hr @

Example 9-2. For the bearing in Example 9-1, how does the rating life and median
life change if the bearing is subjected to a light shock load as the outer ring is
rotating?

Solution: For a rotating outer ring, V = 1.2. Thus, we have

F, -~» 1250
= 0.560
VF. (1.2)(1890)
we see that F,/VF, is still larger than any value of e applicable to a single-row
ball bearing in Table 9-5; thus, we still must choose an X and Y from Table
 49]
Section 9-10: Equivalent Load

9-5. With the ratio (iF ,/Cy remaining the same, X = 0.56 and Y = 1.287. From
Table 9-8 and equation (9-4)

P= F(XVE + YF.) = 1.5(0.56 x 1.2 x 1890 +4 1.287 x 1250) = 4320 1b

From equation (9-5)
P=FVF = 1.51.2 x 1890) = 3400 Ib
Since 4320 Ib is larger than 3400 lb, we use P = 4320 Ib for calculating the
rating life. Now, using equation (9-2)
b 1 3
Yoana (| “ a = 9.37 x 10° rev
IP 4320
: 9.37 x 10°
rating litte = 1500 x 60. = 104 hr

median life = 5(104) = 520 hr

Note: for a 62% increase in equivalent load, the rating life was reduced by
76.5%. @

Example 9-3. As another example, consider the case of selecting a bearing for a 25-in.
diameter shaft that rotates uniformly and steadily at 1800 rpm. Due to a bevel
gear mounted on the shaft, the bearing will have to withstand a 1000-Ib radial
load and a 350-lb thrust load. Without any restriction as to the type of bearing,
select one that is satisfactory for a rating life of at least 20,000 hr.
Solution: Observing that the thrust load is greater than 20% of the radial
load, we can eliminate the possibility of using a straight roller bearing. We can
also eliminate from consideration the tapered roller bearing. This type of
bearing is primarily used for applications where there is high thrust load and
a moderately high radial load. We will, therefore, consider using one of the
following types of bearings: (1) a single-row deep groove bearing, (2) a single-
row spherical roller bearing, and (3) a double-row spherical roller bearing. The
bearing bore diameter is selected so that it will be compatible for mounting
on the 25-in. diameter shaft (that is, the shaft diameter can be “turned down”
to fit the bearing bore and provide a shoulder for the inner ring). In this case,
we shall specify a bearing bore of 50 mm (1.9685 in.).
Part 1. Consider a 310 single-row, deep groove bearing. From Table 9-2
we establish that C = 10,700 lb, and Cy = 8010 lb. We now need to determine
the equivalent load P. To do so, the values of X and Y are needed from Table
9-5. To obtain the values of X and _Y, it is necessary to calculate iF,/Cy and
find e. Thus
5
e = Ane = 0.0437; e = 0.2424 by linear interpolation

3
a = ma = 250 = O24 = @

Thus, from Table 9-5, X = 0.56, Y = 1.833 by linear interpolation. Applying

equation (9-4), we obtain
P=XVE, + YF, = (0:56)(1)(1000) =F 1.833(350) = 1202 1b
Then applying equation (9-5)
P = VF, = 1(1000) = 1000 1b
 492
Chapter 9: Rolling Bearings

Using the higher equivalent load we proceed directly to calculate rating life.
By means of equation (9-2), we get
b 3
Lio = (5 = (2 |= 705 x 10° rev
ie 1202

or by using equation (9-3) we obtain the rating life L,, in hours

107 CV" Es
= 105G05) = 652 hr.
>| 60(1800) :
Ibs =
Ae 60n

Part 2. Consider a: single-row spherical roller bearing. This bearing is

ordinarily available in bore sizes that are in inches. To fit the shaft diameter
and to be comparable with other bearings reviewed in this problem, we shall
select a spherical roller bearing from Table 9-6 with a bore diameter of
1.9680 in., an O.D. of 3.5433 in., and a width of 0.787 in. We also note that
C = 11,100 lb, and Cy = 9700 Ib. Following the same procedure as with the
deep groove ball bearing, we have

VE = 0.350 < 0.40 = e (from Table 9-6)

r

and so we use X = 1 and Y=0

P= XVF. + YF, = (1)(1)(1000) + 0(350) = 1000 Ib

Pes ae OOO

GV? M100) ,
Lio= (5|= (Fiae = 3020
x 10° rev

Or,

10°(3020)
10 = ~60(1800) = 27,900 hr

Part 3. Consider a double-row spherical roller bearing of the dimension

series 22. From Table 9-1 we note that C = 14,700 Ib, and C, = 12.000 lb.
Values of X and Y are obtained from Table 9-7 for a 50-mm bearing bore

VE =O350) SS QAO = Y = 3.83, X = 0.67 (see footnote of Table 9-7)

r

P = XVF, + YF, = (0.67)(1)(1000) + 3.83(350) = 2010 Ib

P = VF, = 10001b
Using the larger of the two calculated equivalent loads,

Lie= (| = |
C\* _ {14,700\33
Fata | Sn 58.01 LO? TEV

or

_ee OO
10°(758.5) ah
10 = “60(1800) 3
 493
Section 9-11: Bearing Survival with Probabilities Higher than 90%

Part 4. Consider a double-row spherical roller bearing of dimension series

23. From Table 9-2, we note that C = 31,600 lb, and C, = 26,500 lb. Values
of X and Y are obtained from Table 9-7 for a 50-mm bearing bore.
F
VE == (sail) < (0338 Y = 1.73, and X = 1.00 (see footnote of Table 9-7)

P = XVF. + YF, = (1)(1)(1000) + 1.73(350) = 1605 Ib

P = VF, = 10001b
Using the larger of the two calculated equivalent loads,

GN 31,600\4 22 Ze
Dry 5| |1605 | = 20,418 x 10° rev

or

10°(20,418)
NO
Ie, => 60(1800)
——_ = 189,000 hr

From the above calculations, it is readily seen that the single-row deep
groove ball bearing and the double-row spherical roller bearing (dimension
series 22) do not satisfy the life requirement of 20,000 hr. Therefore, they are
eliminated from consideration. Both the single-row spherical roller bearing
and the double-row spherical roller bearing (dimension series 23) satisfy the
life requirement. However, the double-row spherical bearing is 9.45 times
overdesigned. In addition, it is more expensive than the single-row spherical
roller bearing. Our choice, therefore, is the single-row, spherical bearing. @

SECTION 9-11

Bearing Survival with Probabilities

Higher than 90%
As previously indicated, the definition of rating life was based upon a 10%
failure rate (or a 90% survival rate). The AFBMA and manufacturers’
catalogs list the bearing capacities to be compatible with this definition. The
question that arises 1s “how can the designer select a bearing with a probability
of survival greater than 90%’? Certainly, there are many applications in
which the likelihood of failure or catalog survival rate cannot be tolerated
(for example, manned space vehicles, medical and hospital equipment, nuclear
power plant controls, and so on).
The AFBMA has recognized the need for providing reliability adjustment
factors and has published such data in its latest (1972) standards. These
data are given for reliabilities of 90, 95, 96, 97, 98, and 99 %. To avoid the need
for interpolation, we have plotted these values in Figure 9-37. This curve can
be applied to both ball and roller bearings but is restricted to reliabilities no
greater than 99 %.
 494
Chapter 9: Rolling Bearings

factor,
reliability
a4
adjustment
Life

Reliability, percent

Figure 9-37 Life adjustment relating factor, a,.

Example 9-4. Let us assume that in Example 9-1 only a 5% probability of failure can
be permitted (that is, 95 % reliability). What can the designer expect the rating
life to be for this 309 bearing?
Solution: From Figure 9-37, for a reliability of 95.0%, a, = 0.62. Thus
CG 3

jig Ori bgp a P |= 0.62(40 x 10°) = 24.8 x 10° rev

or

Ls; = 0.62(445) = 276hr and _ median life = 5(276) = 1380 hr

In order to improve the reliability of the 309 bearing of Example 9-1 from 90
to 95.0%, a reduction of rating life from 40 x 10° rev to 24.8 x 10° rev is
required. @

Example 9-5. A single-row deep groove ball bearing with a bore diameter not to
exceed 45 mm (1.7717 in.) is to be selected. The bearing is to withstand a 400-lb
radial load and a 100-Ib thrust load at 1000 rpm. It is desired that the bearing
have a rating life of 5000 hr with a failure probability of 6.0%. Select the
bearing if the inner ring is rotating.

Solution:
 495
Section 9-12: Life Adjustment Factors for Materials

Inspection of Table 9-5 leads us to a first guess, namely that

VF. IA) 2 2
Ee

Mhusywesesx— lvand) y— 0

P= XVF, + YF, = (1)(1)(400) + (0)(100) = 400 Ib

P= VF, = 400 ib

From equation (9-3), we have

c= P|n x 60 x =e)" - 4g (0 x 60 x 5000) '/%

7 ie = 2680 Ib

From Figure 9-37 for a 6.0% probability of failure, we find a, = 0.70. Thus
we must find a single-row deep groove bearing with C = 2680/(0.70)'/? =
3018 lb. @

Inspecting Table 9-1 under the column for dimension series 02, we find a
bearing with the following data: bore diameter = 30 mm (1.1811 in.);
6 — 3360 Ib; Co = 2250 Ib:
As a check of our original guess, we calculate

iF, 1(100)
=—_ = 0.0444
Gu mtlI50
Inspection of Table 9-5 shows that for this value of iF,/Co, e is greater
than 0.250. Thus our original assumption is justified, namely X = 1 and
Y = 0 when F,/VF, < e. Thus the bearing selected is satisfactory.

SECTION 9-12
Life Adjustment Factors for Materials
The basic dynamic capacity ratings as defined by the AFBMA (see Section
9-9) are predicated on the use of air-melted AISI 52100 steel with a minimum
through hardness of 58 Rockwell C. However, an improved air-melting
process and the introduction of carbon-vacuum degassed (CVD)’? steel has
resulted in steels of higher quality.
Thus better steels, higher quality forming methods, improved control of
heat treatment and overall manufacturing techniques have resulted in life
increases of 3-8 times that of L,, [13]. In fact in using consumable electrode
vacuum-remelted (CVM)?* steel, extraordinarily long bearing lives were
experienced.
The AFBMA, has recognized this life improvement but considers all of
these improved steels as a special category and therefore suggests that the
?3 This is currently the most common process for producing bearing steels.
24 The process called CVM (consumable electrode vacuum remelting) results in a steel of even
higher quality than that produced by the CVD process (see reference [13)).
 496
Chapter 9: Rolling Bearings

designer obtain the materials adjustment factor a, from the manufacturer.

However, the AFBMA Standard does indicate that “the ball bearing
industry currently uses an a, factor of 3 for radial ball bearings of good
quality.’’?°
For CVM steels, reference [13] also recommends a factor of.3. For a
‘‘normal’’ bearing material (that is, air-melted and through-hardened steel),
reference [13] lists factors that range from 0.2 to 2.0. Of particular significance
is the fact that the most common bearing material, AISI 52100 is assigned the
factor of 2.0 based on the results of having fatigue tested a 207 ball bearing.
This results in a bearing life 100°% longer than would be normally calculated
from equation (9-2). The reader is advised to use this factor a, with caution
and only after careful consideration of all the parameters that can affect the
fatigue life of a bearing.

SECTION 9-13

Life Adjustment Factor for Application

Conditions
By deductive reasoning, it is not difficult to realize that bearing life can
be (and is) significantly affected by application conditions. Accordingly the
AFBMA suggests the following be considered in determining the life of a
bearing: (1) lubrication, (2) load distribution (including the effects of
clearance, misalignment, housing and shaft stiffness, type of loading, and
thermal gradients), and (3) temperature. The life factor determined by (2)
and (3) should be obtained from the bearing manufacturer.*°
Where lubrication (that is, item 1) is adequate, an a, factor of 1 is recom-
mended. However, a, might conceivably be less than 1 when any of the
following conditions prevail:
1. Exceptionally low values of nd,, (revolutions per minute times ball pitch
diameter in millimeters), for example, nd,, < 10,000. This requirement is
applicable to both ball and roller bearings.
2. Lubricant viscosity is less than 70 SUS (100 SUS for roller bearings) at
operating temperature.
3. Excessively high operating temperature.

If any of these conditions exist in a particular application, the designer

should obtain an appropriate value of a, from the manufacturer before
making a final bearing selection.

° Although not a part of the AFBMA Standard, reference [13] contains a list of adjustment
factors which makes a distinction between bearing materials and processes.
7° Reference [13] contains quantitative values for such parameters as lubrication, speed, and
misalignment. However, these values have not been sanctioned by the AFBMA and if used
should be applied with caution.
 497
Section 9-15: Rolling Element Bearings Subjected to Variable Loads

SECTION 9-14
Summary of Life Adjustment Factors
From the previous discussion an adjusted fatigue life equation can be
expressed as
GC b

Ly, = 44203 (<] (9-6)

where L/, = the adjusted fatigue life of the bearing with a failure probability
of n, and a,a,a, = factors as defined respectively in Sections 9-11, 9-12, and
9-13.
The designer should be forewarned not to apply the life adjustment factors
indiscriminately because this may lead to an overly optimistic estimation of
the bearing endurance life. Uncertain parameters such as shaft size, mis-
alignment, temperature gradients, and so on, can cause equation (9-6) to be
invalid.
Most situations requiring the selection of a bearing are routine. But where
special requirements must be satisfied, the designer would do well to discuss
his problem(s) with the manufacturer.

SECTION 9-15

Rolling Element Bearings Subjected to

Variable Loads
There are many applications where a bearing is subjected to both variable
loading and changing speed during each revolution. A designer could select
a bearing for the most severe loading condition. However, this viewpoint is
rather conservative and could be unnecessarily expensive, particularly if
production required the use of many such bearings. What is needed, therefore,
is a relationship that will relate load rating C and rating life L,) for varying
loads.
Rearranging equation (9-2), we have

_ 60nLjo
@
nae (9-7)

If during a fraction of cycle p,, the speed n,, and equivalent load P,, are
constant, we express equation (9-7) as a summation of the effect of n,, and
P,, on each fraction ofa cycle. Thus, we obtain

60Li5 <
Cc = Te Dep mom
Neesm (9-8)
m=1
 498
Chapter 9: Rolling Bearings

where

Il the required basic load rating, lb

rating life, hr (for a probability of failure of 16%)
II fraction of a cycle
= speed during fraction of a cycle, rpm
= equivalent load during ayfraction of a cycle, lb

Example 9-6. A 25-in. diameter shaft of a machine operates continuously for 8 hr

daily. Because of an overrunning clutch, one of the shaft bearings will be
subjected to a varying load and speed cycle as follows:

Fraction of the Cycle F,, lb E510, RPM Load Conditions

+ 800 400 1000 Steady

+5 400 400 1500 Light shock
A 1000 400 1200 _—_Light shock
a 500 400 1500 Steady

Find a ball or roller bearing that will satisfy the given conditions. The inner
ring rotates.
Solution: Referring to Table 9-9, we shall assume a design life of 25,000 hr
(based upon a reliability of 90%). Also, from Table 9-8, shock factors for the
given conditions are:
For ball bearings:
F (steady load) = 1.0
Flight shock) = 1.5
For roller bearings:
F(steady load) = 1.0
F(light shock) = 1.0
As a trial selection, we shall try a 312 single-row deep groove ball bearing.
From Table 9-2 fora 60-mm(12 x 5 = 60) bearing bore, we have C = 14,100 lb
and Cy = 10,900 lb. We now proceed to calculate the equivalent load for each
fraction ofthe cycle.
For the first tenth of the cycle:
lgpe ae 400 = 5

VF. (1)(800) noe

For the second tenth of the cycle:
F 400
——— = 100)
 499
Section 9-15: Rolling Element Bearings Subjected to Variable Loads

For the next half of the cycle:

F, 400
= = 04
VF, — (1)(1000) :
For the remaining three tenths of the cycle:
F, 400
VF. (1)(500) — Re
1)(4
S636)
C, 10,900
and by interpolation, e = 0.232.
We can readily see that all of the values of F,/VF, are greater than e = 0.232.
Thus, from Table 9-5, X = 0.56 and, by interploation, Y = 1.922. From
equation (9-4), we now can calculate the equivalent load P for each fraction of
the cycle.
P = XVF, + YF,
P(7b cycle)= (0.56)(1)(800)+ (1.922)(400)= 1217 1b
P(j5 cycle)= (0.56)(1)(400)+ (1.922)(400)= 993 Ib
P(E cycle)= (0.56)(1)(1000)+ (1.922)(400)= 1329 Ib
P(3 cycle) = (0.56)(1)(500)
1) (500)+ (1.922)(400
)(400) = 1049 Ib

Now, we can substitute into equation (9-8), remembering to account for the
shock factors, F..
60(25,000)
ay [¥o(1000)(1.0 x 1217)? + 7o(1500)(1.5 x 933)?
he 10°
+ 3(1200)(1.5 x 1329)? + 7%(1500)(1.0 x 1049)%]
C = 20,650 lb
Because the required basic load rating is larger than the load rating of the
312 bearing (that is, 20,650 > 14,100), we cannot use this bearing. Inspection of
Table 9-2 discloses that a 316 single-row deep groove bearing with a
C = 21,300 lb is the smallest one that can be used. But its bore diameter is
80mm (3.1496 in.), exceeding the 25-in. shaft diameter, so it cannot be
considered.
Referring again to Table 9-2 under the dimension series 23, we now try a
double-row spherical roller bearing with a 45-mm (1.7717 in.) bore. For this
bearing C = 25,000 1b and Cy= 20,600 lb. As above, we must calculate the
equivalent load for each fraction of the cycle.
The value of F,/VF, for each fraction of the cycle calculated above remains
the same for the spherical roller bearing. We see that each of the values for
F_/VF, is greater than the value of e(= 0.39) found in Table 9-7 for a 45-mm
bore bearing of the dimension series 23. Thus, X = 0.67 and Y = 2.59. The
equivalent load for each cycle fraction can now be calculated.
P(;'o cycle) = (0.67)(1)(800)+ (2.59)(400)= 1572 Ib
P(,'5 cycle) = (0.67)(1)(400) + (2.59)(400)= 1302 Ib
P(3 cycle) = (0.67)(1)(1000)+ (2.59)(400) = 1706 lb
P(;%5 cycle) = (0.67)(1)(S00) + (2.59)(400)= 1371 Ib
 500
Chapter 9: Rolling Bearings

Substituting these values into equation (9-8) and accounting for the shock
factors F,, we obtain

3 _ 60(25,000) (5(1000)(1.0 x 1572)? + 74(1500)(1.0 x 1302)°

G 10°
+ 4(200)(1.0 x 1706)? + 33(1500)(1.0 x 1371)*]
C = 19,400 lb
Because 19,4001b is less than 25,000 lb, the choice of the double-row
spherical roller bearing with a 45-mm bore diameter is satisfactory. @

SECTION 9-16
Lubrication of Rolling Bearings
Satisfactory performance of rolling bearings depends greatly upon the use
of the proper lubricant, lubrication frequency, and a housing design that
provides adequate lubricant flow. Although the frictional forces produced
by a rolling element are comparatively low (see Section 9-7), sliding between
the elements, races, and separator (that is, retainer or cage) is a major source
of frictional resistance. Two other factors contributing to frictional resistance
are the movement of the bearing elements through the lubricant and the
deformation of the rolling elements and race when under load. The latter
consideration has, in recent times, become a topic of great research interest,
particularly when the bearing Senet are highly loaded. It is called
elastohydrodynamic lubrication.
Specifically, a proper lubricant should satisfy the following requirements :
1. Form a film between the rolling elements and raceways and between the
rolling elements and separator.
2. Protect the bearing components against corrosion.
Ww. Remove the heat generated by the bearing when in operation.
4. Prevent the infiltration ofdirt and/or other foreign matter from entering
the bearing.
Depending upon load, speed, and temperature requirements, lubricants are
either greases or oils. Synthetic and dry lubricants are also widely used for
special applications (for example, extreme temperatures, inaccessibility to
continuous lubrication, and so on). Greases are suitable for low speed
operation and do not require complicated sealing or lubricating systems such
as when oils are used. Also, grease permits bearings to be prepacked.
Greases are a mixture ofa lubricating oil and a metallic soap or some other
vehicle that will keep the oil in suspension (for example, synthetic thickener).
*” The problems ofinterest in this area also involve gear tooth contact stress. A good introduction
to this subject is a book entitled Elasto-H ydrodynamic Lubrication, by D. Dawson and G. R.
Higginson, Pergamon Press, New York. It contains an excellent list of references. The reader
should also refer to recent papers published in the Transactions of the ASME and The Journal
of Lubrication Technology, published by the American Society of Lubrication Engineers.
 501
Section 9-16: Lubrication of Rolling Bearings

Thickeners used are calcium (that is, lime), sodium, lithium, barium, bentone,
aluminum, or synthetic soap bases. Greases are graded according to a
hardness rank scale established by the National Lubrication Grease Institute
(NGLI), which ranges from 0 to 6 in order of increasing hardness. Hardness
grade numbers are determined by standard ASTM penetration tests. Table
9-10 indicates the important operating characteristics and operating con-
ditions of greases having various soap bases.
To prevent “churning”’ of grease at operating temperature, the bearing
housing volume should not be filled with more than one third to one half of
its capacity. The housing and shaft design should be such as to “‘urge’’ the
grease into contact with the bearing. The time periods between regreasing
depend upon the load, speed, operating temperature, type of grease used, and
environment. These periods can range from 6 months to 2 years. Manu-
facturers’ catalogs usually give regreasing recommendations for normal
applications. Typical bearing mountings employing grease as a lubricant are
shown in Figures 9-38 and 9-39.
Where bearing speeds are higher or loads are severe, oil becomes the
preferred method oflubrication. The most common type of lubricating oil is
a mineral oil. Oil, depending upon the desired properties, contains different
kinds and quantities of extreme pressure additives, antifoaming, and anti-
oxidation agents. Viscosity ratings for the lubricating oils are usually stated
in Saybolt Universal Seconds (SUS) at 100°F, and are graded by SAE,
ASTM, or AGMA numbers.

Table 9-10 Important operating characteristics of various greases

Operating Conditions Characteristics

Point, ating Temp- ne

Recommended Number °F °F Resistance

7a
\

SY Lime base Good

N se 2, 4, 0r5§
a, 3,4,

ral Ses

a3
EN NS
\\

N
Sodium base |0.
0. moe
1or2||

lor2]
|0,1
Sra
325

012 Above
WW

— | AX_
Lithium base Good

TL
ae - Barium base ma > 400 300

= a | sean
SOURCE: Bearings Book Issue, Machine Design, The Penton
Publishing Co., Cleveland, Ohio, 1963.

NOTES: 1, Grease should contain oxidation inhibitor.

2. Grease should contain EP additives if loads
are abnormally high,
3. Under operating conditions, use areas that are cross hatched.
 502
Chapter 9: Rolling Bearings

Figure 9-38 Regreaseable mounting. Recommended for

moderate speeds and loads. Where prelubricated sealed
bearings are not suitable for a particular reason,
consideration must be given to the use of an open type
bearing with provision for relubrication. The grease plug
at the bottom should be removed while the grease is being
inserted through the fittings at the top. The direction of
flow tends to remove the old grease. [Courtesy Marlin-
Rockwell Division of TRW, Inc. Jamestown, N.Y.]

Methods of selecting a lubricating oil vary with different manufacturers.

For example, Link-Belt Company makes its selection based on speed, per
cent of basic load rating, and operating temperature. The Torrington
Company uses operating temperature and speed, where the speed is limited
by the bearing size and type. SKF Industries, Inc. attempts to select a
lubricant by considering the elastohydrodynamic film formed between the
rolling elements and the raceway. The procedure is rather involved and
cannot be entered into in this text. Interested readers are advised to contact
SKF for specific information. Figure 9-40 is offered here as a simple and
straightforward method of selecting a lubricating oil for a rolling bearing.
The heavy, dotted line in Figure 9-40 is an example of how to use the chart.
 Figure 9-39 High temperature grease
mounting. The life of permanently
lubricated installations at high temperatures
is a function of the volume of grease present,
and the design of the mounting. Note that
ample grease space has been provided, and
that the configuration of parts adjacent to
the bearing is such as to urge the lubricant
into contact with the active bearing parts.
(Courtesy Marlin-Rockwell, Division of
TRW, Inc. Jamestown, N.Y.]

°F
Operating
temperature,

Figure 9-40 Oil viscosity selection chart. The heavy dotted line is an example
(that is, a 50-mm bore ball bearing (D), running at a shaft speed of 5000 rpm (N) is
to operate at 150°F). Enter the chart at DN = 50 x 5000 = 250,000. Move upward
and onto the top surface until the operating temperature of 150°F is reached.
Move parallel and down to the right and then vertically down as shown to read the
required viscosity is 170 SUS. [Courtesy the Fafnir Bearing Company. ]

503
 504
Chapter 9: Rolling Bearings

The amount ofoil necessary for adequate lubrication is such that no more
than a thin film is provided. When the quantity of oil exceeds that which is
required to just form a film on the bearing, frictional torque increases. Only
in those cases where the heat generated is an important factor should increased
quantities of oil be used—then only where the oil can be circulated or mist
sprayed. There are six methods commonly employed for oil lubrication.
These are described by the Marlin-Rockwell Company as:
1. Jet Oil Lubrication. Oil, under pressure, is forced through an orifice into
the bearing. The oil jet is directed at the space between the bore of the cage
(that is, separator) and the outer diameter of the inner ring. This means of
application is particularly advantageous for heavily loaded, high speed
operation. For extremely high speeds, means for scavenging the oil should be
provided on each side of the bearing (see Figure 9-41).
2. Circulating Oil Lubrication. In this system, oil is circulated through the
bearing by slingers or other means. Circulating systems provide a reliable,
relatively low cost method of lubricating heavily loaded bearings. More
expensive systems use a circulating pump to assure a positive supply of
lubricant (see Figure 9-42).

Figure 9-41 Jet oil lubrication.

[Courtesy Marlin-Rockwell, Division of
TRW Inc., Jamestown, N.Y.
 ZZ

Figure 9-42 Circulating oil lubrication by means of a

slinger. [Courtesy SKF Industries, Inc.]

Figure 9-43 Oil splash lubrication. Note the feeder trails in the case
to direct the oil washdown into the bearings. [Courtesy Fafnir Bearing
Co.]

505
 506
Chapter 9: Rolling Bearings

Figure 9-44 Oil bath lubrication. Note that

the oil depth is maintained at approximately the
level of the midpoint of the lowest ball.
[Courtesy Marlin-Rockwell, Division of TRW,
Inc., Jamestown, N.Y.]

B@bal
Csths

3. Oil Splash Lubrication. The splash system is particularly suitable for

oil lubricated gear boxes. The splash from the gears serves to lubricate the
bearings. Because the oil carries the debris of wear, the use of filters and
magnetic drain plugs is helpful in reducing possible contamination of
bearings (see Figure 9-43).
4. Oil Bath Lubrication. This form of lubrication is suitable for slower
speed applications where an oil reservoir keeps the bearing partially
immersed. The oil level should never be higher than the midpoint of the
lowest ball (or roller) being lubricated. A larger amount of oil can cause
“churning,” which increases fluid friction and can cause excessive operating
temperatures (see Figure 9-44).

=
oth fees ed SYA 222227772
Of idLMT
Zz Ss Aye
opSS
et)
Li/S
Fae

Figure 9-45 Drop feed lubrication. [Courtesy Fafnir Bearing Co.]

 507
Section 9-17: Speed Limitations of Rolling Bearings

ll
N
LLITID

is
poe)
(
<

yy

—
Figure 9-46 Oil-air mist lubrication of a
machine tool spindle. The lubricant used is a
spindle oil rated at 100 SUS at 100°F. It is
delivered at the rate of 40 drops/min into the
system under 20 psi air pressure. Precleaning
of the air and lubricating lines and effective
air filtering is a must for successful operation.
[Courtesy Fafnir Bearing Co.]

ae
Eicmay
el
LLL
lll“Z
»
1K

5. Drop Feed Lubrication. This system uses a commercial drop feed oil. It
is used where moderate loads and speeds are encountered. The principal
disadvantages are the refilling of the oil cup and the provision for the disposal
of waste oil (see Figure 9-45).
6. Air-Oil Mist Lubrication. In this system, tiny droplets of oil are sus-
pended in an air stream that passes through the bearing. Commercial units
are available which provide excellent lubrication for high speed units such as
machine tool spindles (see Figure 9-46).

SECTION 9-17

‘Speed Limitations of Rolling Bearings

Some manufacturers’ catalogs provide information concerning the speed
limitation imposed on bearings. However, lacking such information, we can
determine the approximate limiting speed for rolling bearings. Such infor-
mation is important because excessive speed for an extended time period can
increase the bearing temperature and cause a breakdown in lubrication. This
situation would ultimately lead to bearing failure. The temperature rise,
caused by friction, is a function of the type of bearing construction, the size
of the bearing, and the load it has to carry.
The approximate limiting speed of a ball or roller bearing is given by [11]

_SifA
il
(9-9)
m
 508
Chapter 9: Rolling Bearings

0.2 i
0.1
5 Teas rea eeeeee ee Ms 2 Ta
50 100 200 300 500 1000

Bearing mean diameter, d,,,mm

Figure 9-47 Factorf, to be used in equations

(9-9) and (9-10). [From Frank W. Wilson (ed.): Tool
Engineers Handbook, 2nd ed. McGraw-Hill Book
Co. New York, 1959.]

and for a thrust bearing is

Sifra
n= 2 (9-10)
nyide
where

n = approximate speed, rpm, when using oil lubrication (when using

grease, reduce this value to 3 for ball bearings, and 4 for roller
bearings)
f; = a coefficient depending on bearing size, from Figure 9-47
ff, = a coefficient depending on relative bearing load, from Figure 9-48

fy
1
0.9
0.8
0.7
0.6 1000,000
0.5 nee 50,000
0.4 i in hours 20,000
0.3 10,000
0.2 |-
0.1
FO ee EN 1g) 6 aon Wesel ie ee ed
50 100 200 300 500 1000
Bearing mean diameter, d_,,mm

Figure 9-48 Factorf, to be used in equations

(9-9) and (9-10). [From Frank W. Wilson (ed.): Tool
Engineers Handbook, 2nd ed. McGraw-Hill Book
Co., New York, 1959.]
 509
Section 9-18: Seals

Table 9-11 Factor A for determining approximate

bearing speed limitation

Bearing Type Factor A

Radial bearings
Single-row deep-groove ball bearings
Ball-centered pressed-steel cage _____==—SC«SS.00,000
Ring-centered machined cage____—SC«é.,000
Angular-contact ball bearings
Large-angle ball-centered cage, single 500,000
Large-angle ball-centered cage, duplex ___—s 350,000
Small-angle ring-centered cage, single ____——- 800,000
Small-angle ring-centered cage, duplex _______——- 600,000
Single-row filling-slot ball bearings __E—SC«S.00,000
Double-row deep-groove ball bearings __________ 350,000
Self-aligning ball bearings
Dimension series 02, 12, 22, 03,04 ___—« 00,000
Dimension series 23 _____— SS CS«4S'0,000
Cylindrical roller bearings
Sing] C10 yy eeeee eee 0000
Double-1owe ee 45,000
Self-aligning roller bearings
Danvensiontsenicsulon2 a.) 5 =a nS (O00)
Dimensiontiseriess Ono leo. ED SOOO
Tapered roller bearings
Single10 j= 00,000
DOU bIe=3 0 eee EE RI)I() 00
Thrust bearings:
Ball thrust bearings
PTESSEG-SLCE CAD Cae LS OOO
Machined-b1ass\cace 200000
Spherical roller thrust bearings ___S- 200,000
SOURCE: Frank W. Wilson (ed.): Tool Engineers Handbook, 2nd ed. McGraw-Hill
Book Company, New York, 1959.

aN = a value depending on bearing type, from Table 9-11

d,,= = bearing mean diameter, mm
= bearing outer diameter, mm
H = bearing height for a thrust bearing, mm

SECTION 9-18

Seals
Seals perform two very important functions, containment of the lubricant
and keeping foreign matter from entering the bearing area. Earlier in our
description of bearings, we described how some of them are provided with
seals and shields to contain the lubricant and exclude contaminants. How-
ever, where bearings are not “‘self-sealed,” provision must be made for
sealing. There are many seal designs, and the selection of any one design
depends on what is expected of the seal, for example, (1) the nature of the
material to be excluded (that is, liquid, solid, powder, and so on, (2) shaft
 510
Chapter 9: Rolling Bearings

surface speed, (3) temperature conditions, (4) type of lubricant, (5) permissible
leakage, and (6) other environmental and operational conditions.
Due to limitations of space, we cannot enter into a detailed analysis and
description ofseals.*® Basically, there are three categories ofseals : (1) contact
seals, (2) labyrinth or clearance seals, and (3) combinations of contact and
labyrinth seals.
Contact seals are available from simple felt stripping to spring loaded lip
assemblies. Seals within this category are in intimate contact with the shaft,
and require quality shaft finishes. In some cases, surface flatness is within
twenty millionths of an inch. Contact seals limit the surface speed of the shaft
because of frictional drag and the resulting temperature increase in the
bearing area. Depending upon the shaft diameter, seal] material, and seal
design, shaft rubbing (that is, surface) speeds range between 500 and
1000 ft/min for felt, and from 2000 to 3000 ft/min for lip seals.
Lip seals are made of various materials such as leather, synthetic elastomers,
plastics, or laminates. Figure 9-49 illustrates how a felt seal is used in con-
junction with a lip seal. Felt seals are used in combination with lip seals and
labyrinth seals because of their “‘wicking effect,’ which tends to absorb a
contaminating liquid that is to be “sealed out”’ or an oil which is to be “sealed
in.” Felt seals are good for excluding dirt and foreign solid matter. Lip seals
are excellent for sealing solids, liquids, and gases at moderate pressures. For
sealing liquids and gases at medium to high pressures and high speeds, an
axial type of seal is often used. Although expensive and requiring very precise
alignment, the axial type seal is quite effective (see Figure 9-50).

Figure 9-49 Felt used in combination with a lip seal.

[Courtesy of Link-Belt Bearing Division, FMC
Corporation. ]

8 For an in-depth discussion, see the Seals Symposium Issue, Transactions of the ASME, Series
F, Vol. 90, No. 2, 1968. Also see Machine Design, 1971 Seals and Reference Issue, Vol. 43, No. 2.
 S17
Section 9-18: Seals

Removable Tension Spring tension ring

end plate i
Driving ring
—S=—} ,
Seal face ring
. —
ut —— 7]
——7 Anti-friction
Sere ring seat

| ] KRAgy RAY) AN Ly) |

Anti-friction

lat IZ
SS NY sealing ring

Figure 9-50 Mechanical axial seal. [Courtesy Link-

Belt Bearing Division, FMC Corporation. ]

Labyrinth or clearance type seals are deliberately designed to provide

clearance between the seal and the rotating shaft. Unlike the contact type
seal, friction due to rubbing and the subsequent heat generation present no
problem for the labyrinth type seal. Consequently, labyrinth seals do not
restrict the speed of the rotating shaft. Labyrinth seals are generally not used
to seal against pressure or for applications that require the bearing to be
submerged in liquids or solids. Labyrinth type seals range from simple low-
clearance shaft openings (see Figure 9-51) to a complex system of labyrinths,
flingers, and baffles (see Figure 9-52). Labyrinth seals are constructed of
metal or plastic, the choice depending upon corrosive environments,
temperature conditions, abrasive materials, and so on.

Figure 9-51 Felt used in combination with a

labyrinth seal. [Courtesy Link-Belt Bearing Division,
FMC Corporation. ]
 DZ
Chapter 9: Rolling Bearings

Figure 9-52 Variations of labyrinth seals. [Courtesy Link-Belt Division, FMC

Corporation. ]

SECTION 9-19
Typical Bearing Assemblies
It is important that the designer specify the correct shaft and housing
tolerances for mounting a bearing. For a rotating shaft, the inner ring of a
bearing has a tight (that is, interference) to a slightly loose fit with the shaft,
whereas the outer ring has a loose to slightly tight fit with the housing. For
a mounting with a rotating housing, the converse of these fits applies. The
type of fit depends upon the bearing size and the ABEC number. Manufac-
turers’ catalogs contain detailed tables specifying the shaft and housing
tolerances for each ABEC number. A design that does not use the data in
these tables invariably leads to early or even immediate bearing failure.
Actual mounting arrangements vary widely and depend upon the type of
bearing used and the application being considered. The Marlin-Rockwell
Company suggests five basic mounting designs for single-row and double-
row ball bearings which are used in the majority of applications. These
mounting designs are shown in Figures 9-53 to 9-57. Aspects of these designs
can also be used for rolling bearings. It would be wiser, however, to consult
the manufacturers’ catalogs for specific recommendations.
The standard mounting (Figure 9-53) is the ideal mounting for a shaft
supported by two ball bearings. It has the following advantages :
1. Permits one bearing to take an axial load in either direction.
2. Axial shaft expansion is provided for by the “‘floating”’ of the unclamped
bearing.
3. Bearings cannot be axially preloaded through improper adjustment of
the lock nuts. The lock nuts serve to clamp the bearing inner rings
against the shaft shoulder.
4. Mounting arrangement is suitable for a wide range of speed and
temperature conditions.
 cy
ASX
C+)

Figure 9-53 Standard mounting. [Courtesy Marlin-Rockwell Division

of TRW, Inc., Jamestown, N.Y.]

ar
SAY

L277)

Figure 9-54 Standard mounting with a spacer.

[Courtesy Marlin-Rockwell Division of TRW, Inc.,
Jamestown, N.Y.]

0.005 to 0.010

SIN
KD
C20

Figure 9-55 Mounting using no lock nut or snap-rings. [Courtesy

Marlin-Rockwell Division of TRW, Inc., Jamestown, N.Y.]

513
 514
Chapter 9: Rolling Bearings

Figure 9-56 Mounting with a lock nut and snap-ring.

[Courtesy Marlin-Rockwell Division of TRW, Inc.,
Jamestown, N.Y.]

An alternate bearing mounting to that shown in Figure 9-53 is the one

shown in Figure 9-54. This design is used where the bearings are assembled
from one end of the shaft. Note that the inner rings of both bearings, along
with the spacer which separates them, are clamped against the shaft shoulder
by means ofonly one lock nut at the end of the shaft. The advantages of this
mounting design are the same as the ones cited above.
The bearing mounting design shown in Figure 9-55 does not use nuts on
the shaft and permits a “through” bore in the housing. Because of tolerance
buildup of the various components, it is difficult to contro! the axial play of
the shaft unless shims are used. These shims are usually mounted between the
face of the bearing outer ring and the end cover shoulder. The axial play of
the shaft should be sufficiently large to eliminate any possibility of preloading
due to thermal expansions, yet small enough to eliminate the effects of any
reversing thrust loads. Preferably, this design should be avoided where
reversing thrust loads may occur. This mounting design is also adaptable for
locations where the shaft length is short.
Another alternate to the mounting illustrated in Figure 9-53 is the one
shown in Figure 9-56. Note that this design differs from that shown in Figure

Spring washer

ie}
IK
SQQ
KZZL

Figure 9-57 Mounting with snap rings and a spring washer. [Courtesy
Marlin-Rockwell Division of TRW, Inc., Jamestown, N.Y.]
 89 fs)
Problems

9-53 in that the bearing on the right end is held on the shaft by a snap-ring
which eliminates the lock nut and the necessity of threading the shaft. Also,
the use of a snap-ring bearing (shown on the left in Figure 9-56) makes
possible a “‘through”’ bore housing reducing, thereby, the cost of manufacture.
The last of the five most commonly used bearing mounting designs is
shown in Figure 9-57. This particular mounting can be used where both the
shaft and housing have shoulders and where the thrust load is minimal. A
commercially available spring washer is used to provide a small preload on
the bearings, which acts to eliminate shaft end play and contributes to quiet
bearing operation.

SECTION 9-20
Selection of Tapered Roller Bearings
Limitations of space preclude a discussion of the procedure required to
select a tapered roller bearing. However, bearing manufacturers such as The
Timken Bearing Company and The Torrington Company provide literature
which in “‘stepwise”’ fashion clearly illustrates the selection of this type of
bearing.

PROBLEMS
1. Explain why deep groove bearings are capable of supporting relatively high thrust
loads.
2. Explain why filling notch bearings can support higher radial loads than deep
groove bearings of comparable size.
3. Why is the thrust capacity of the filling notch bearing less than a deep grove
bearing of comparable size?
4. With double-row bearings, the lines of contact converge either inside or outside
the bearing envelope depending on the bearing type. Explain how this convergence
affects the rigidity of the shaft and how it may be used to advantage for particular
applications.
5. What is the primary characteristic difference between a deep groove bearing and
an angular-contact bearing?
6. Explain duplexing of angular-contact bearings. What is the purpose of duplexing?
7. Why are roller type bearings used? What magnitude of thrust load can they
withstand?
8. What is the most important advantage gained in using needle bearings? Can you
cite any examples?
9. List some applications where tapered roller bearings are used. Explain why they
are used for these applications.
10. What is the most common material used for making ball bearings? What materia]
is used for making roller bearings? For each of the above bearing types, what is
the approximate hardness for the material indicated? What temperature limits
the use of these materials?
 516
Chapter 9: Rolling Bearings

11. Explain the purpose of having seals on bearings. Explain the purpose of having
shields on bearings.
12. Why are ball separators used? Why is it detrimental when they are not in the
bearing?
13. Explain the basis for bearing failure. What is meant by spalling?
14. What is the meaning ofL, (or B-10) life of aroller bearing? How does this compare
with the median life of a bearing?
15. Explain rating life and basic load rating.
16. (a) A single-row deep-groove ball bearing with a 60-mm bore rotates at 1000 rpm
while having to withstand a 1000-lb radial load. Calculate the horsepower lost
due to friction. (b) What is the torque at starting? (c) Calculate the starting and
running friction horsepower if, in addition to the radial load, a 400-Ib thrust load
also had to be withstood by the bearing. (d) Repeat the calculation of part (c) for
a spherical roller bearing.
17. For a 60-mm bore single-row deep-groove ball bearing of dimension series 02,
Table 9.1 lists the following values for the basic dynamic and static load ratings as
C = 9070 Ib, and Cy = 6950 Ib. Two of these bearings are to be used to support
a shaft with a stationary center load of 2000 ib. The right hand bearing must also
support a stationary thrust load of 300 Ib. If the speed of the shaft is constant at
1000 rpm, determine (a) the L,9 rating life in revolutions and in hours for each
bearing, (b) the median life in hours for each bearing, (c) the expected life of each
bearing if the equivalent load for each were increased by 50%, (d) the equivalent
load for each bearing if operating life were increased 100% above the L,, rating
life, and (e) the basic number of this bearing that one finds listed in industrial
catalogs.
18. Repeat Problem 17 (except part e) using the same loading and speed conditions
but for a 1.5-in. bore single-row spherical bearing.
19. Repeat Problem 17 (except part e) using the same loading and speed conditions
for a 60-mm bore (dimension series 22), double-row spherical roller bearing.
20. Repeat Problem 17 (except part e) for a rotating outer ring, but stationary inner
ring.
21. A 30-mm bore (dimension series 02) single-row angular-contact ball bearing has a
30 deg contact angle (assume this as a steep angle). If the bearing carries a 700-lb
radial load and a 1200-lb thrust load and the inner ring rotates at 1500 rpm,
determine the rating life in hours.
22. Repeat Problem 21 if the contact angle is 5 deg (that is, considered small).
23. Repeat Problem 21 if the bearing is a single-row deep-groove type.
24. A vertically mounted centrifuge basket and shaft weighs 500 lb and can be filled
with as much as 1000 Ib of material. If the load is centered on the shaft, select an
angular contact ball bearing (from Table 9-2) which will fit a nominal shaft
diameter of 33 in. which rotates at 3000 rpm. Considering the application, select an
appropriate service factor from Table 9-8. What is the rating life of the bearing
selected? If the centrifuge is used intermittently, how does this life compare to that
which is recommended in Table 9-9? To increase the life of the bearing, could you
recommend another type of bearing from Table 9-2? If so, what would be the new
rating life? For each of the bearings selected, specify the bearing number that
would identify either of these bearings in any commercial catalog.
 7
Problems

25. For Problem 24, select a bearing to support the upper part of the vertical shaft.
Keep in mind that all of the thrust load is carried by the lower bearing.

26. (a) An overhead crane trolley runs at 60 ft/min on four 8-in. diameter wheels and is
to have a capacity of 4 tons. The wheels are to be mounted on cylindrical type roller
bearings which are supported by nonrotating shafts 13 in. in diameter. Assuming
that the type of service will, at times, subject the crane to moderate shock, select an
appropriate bearing for a minimum life of 8000 hr of operation. (b) What would be
the expected life of the bearing selected in part (a) if the following considerations
were explored: (1) Desired reliability is 96 %. (2) Full advantage is taken of the AISI
52100, CVM type steel. (3) Normal application conditions (however, check nd,,
value).

27. A spindle for a deburring machine is to be supported by two bearings. The burr,
located outside of the bearing supports, is to rotate at 4000 rpm. Under the most
severe conditions, the bearing closest to the burr will experience a radial load of
250 Ib and an axial load of 150 lb. The bearing furthest from the burr will experience
only a radial load of 200 Ib. If the shaft is subjected to a light shock loading
and cannot exceed 14in. in diameter, select suitable bearings for this operation.
What is the rating life in hours for the bearings selected? Compare this life to the
suggested hours of operation indicated in Table 9-9 assuming the machine will
be in continuous operation for 8hr/day. If the operating temperature of the
bearings is not to exceed 120°F, what must the viscosity of.the lubricating oil be?
Check the bearings selected to see if the required operating speed exceeds its
speed limitation.

28. A deep groove ball bearing is subjected to the following cyclic radial loading:
500 Ib for 4 of a revolution, and 250 Ib for 3 of a revolution. If the shaft is rotating
at a speed of SOO rpm, determine the basic load rating to be used in selecting a
bearing with a life of 10,000 hrs.

29. A shaft is subjected to the varying radial load per cycle as follows: 600 Ib for 1 sec,
200 Ib for 2sec, and 100 1b for 3 sec. If the shaft rotates at a constant speed of
500 rpm what basic load rating for each fraction of the cycle should be used in
selecting a bearing with a life of 15,000 hrs?

30. Determine the equivalent load for the following cyclic radial loading for a life
of 20,000 hr: 100 1b at 1000 rpm for 30% of the time, 80 1b at 5000 rpm for 60% of
the time, and 50 1b at 2000 rpm for 10° of the time. Select the smallest possible
deep groove ball bearing that will satisfy this load. What is the “core number” of
this bearing that would appear in a commercial bearing designation? Check the
bearing for speed limitation.

31. A number 312 deep groove ball bearing is subjected to the following loading cycle:
for 20 % ofthe time, the radial load is 1000 Ib and the axial load is 500 Ib at 700 rpm,
for 50% of the time, the radial load is 1200 1b, and the axial load is 300 lb at
1000 rpm, for 30 % of the time, the radial load is 600 Ib, and the axial load is 400 Ib
at 1500 rpm. Determine (a) the equivalent load, (b) the rating life, (c) the mean
life. The loads are steady and the inner ring rotates.

32. Repeat Problem 31 but this time account for mild shock for the initial 20% of
the time, and heavy shock for the last 30% of the time. In addition to answering
the same questions asked in Problem 31, determine the bearing life for 97%
reliability.
 518
Chapter 9: Rolling Bearings

REFERENCES
[1] Stribeck (H. Hess, tr.): Ball bearings for various loads. Trans. ASME, 29: 420-463
(1907).
[2] H. Hertz: Gesamlte Werke (Collected Works), Vol. 1. Leipzig, Germany, 1895,
p.157.
[3] Arvid Palmgren: Ball and Roller Bearing Engineering, 3rd ed. SKF Industries,
Inc., Philadelphia, Pa., 1959.
[4] Tedric A. Harris: Rolling Bearing Analysis: Theory and Analysis. John Wiley &
Sons, Inc., New York, 1966.
Harold A. Rothbart (ed.): Mechanical Design and Systems Handbook, Sect. 13.
McGraw-Hill Book Co., New York, 1964.
| 1970 Bearings Reference Issue, Machine Design. The Penton Publishing Co.,
Cleveland, Ohio.
Standards of the Anti-Friction Bearing Manufacturers Association. New York, 1972.
| Bearing Technical Journal. FMC Corp., Link-Belt Bearing Division, Indianapolis,
Ind., 1970.
G. Lundberg and A. Palmgren: Dynamic capacity of roller bearings, Acta
Polytech., Mech. Eng. Ser. (Stockholm), 1(3), 1947.
| G. Lundberg and A. Palmgren: Dynamic capacity of roller bearings. Acta Polytech.
Mech. Eng. Ser. (Stockholm), 2(4), 1947.
] Frank W. Wilson (ed.): Tool Engineers Handbook, 2nd ed. McGraw-Hill Book
Co., New York, 1959.
Central Catalogue and Engineering Data. SKF Industries, Philadelphia, Pa.,
] E.N. Bamber, et al.: Life Adjustment Factors for Ball and Roller Bearings. American
Society of Mechanical Engineers, New York, 1971.
| Bearings Book Issue, Machine Design. The Penton Publishing Co., Cleveland,
Ohio, 1963.
Spur Gears
SYMBOLS

a = addendum Kr = factor of safety (reliability

b = width of gear factor)
C = clearance K, = size correction factor
c = center distance K, = temperature factor
Cy, = surface condition factor K, = theoretical stress concentration
C,, = load distribution factor factor
C, = overload factor n = speed, rpm
C, = elastic properties coefficient N, = number of teeth
C, = size factor P = diametral pitch
C,, = dynamic factor p = circular pitch
= pitch diameter q = notch sensitivity factor
d, = base circle diameter R = carburized Rockwell hardness
d, = outside diameter number
E = modulus of elasticity r = pitch circle radius
F,, = bending load r, = addendum circle radius
F, = dynamic load r, = base circle radius
= normal load r. = radius of curvature
= radial load r, = velocity ratio
F, = tangential force S, = allowable static stress
F,, = wear load S, = allowable surface endurance
hp = horsepower limit
I = geometry factor for wear T = torque
geometry factor for bending t = tooth thickness
reollwear load factor V,, = pitch line velocity
K, = fatigue stress concentration Y, y = Lewis form factor
factor it = Poisson’s ratio
K, = life factor o = actual stress
nn m ll= load distribution factor o = pressure angle
K, = overload factor @ = angular velocity

This chapter will present a detailed analysis of spur gears. A brief review
of terminology and kinematics will be followed by a complete discussion of
the stresses and other factors that influence spur gear design.

oS
 520
Chapter 10: Spur Gears

SECTION 10-1

Spur Gear Terminology

Because it is assumed that the reader already has an understanding of the
kinematics of spur gearing, only a brief review of this topic will be given.
Although many good texts are available, reference [1] is suggested for those
who wish a more detailed review.
Spur gears are used to transmit power and rotary motion between parallel
shafts. As can be seen in Figure 10-1, the teeth are cut parallel to the axis of
the shaft on which the gears. are mounted. The smaller of two gears in mesh is
called the pinion, and the larger is customarily designated as the gear. In most
applications the pinion is the driving element whereas the gear is the driven
element. There are some applications, as in epicyclic gear trains, Figure 10-2,
for which gears with teeth cut on the inside of the rim are needed. Such a gear
is known as an internal gear.
The important definitions and geometric properties of external and internal
spur gears are presented in Figures 10-3 and 10-4.
The circular pitch, p, of a spur gear is defined as the distance, on the pitch
circle, from a point on a tooth to the corresponding point of the adjacent
tooth. The diametral pitch is defined as the number ofteeth ofthe gear divided
by the pitch circle diameter.
From these definitions it is clear that the following equations hold:

age
Ne 10-1
(10-1)

Figure 10-1 The spur gears shown

have their teeth cut parallel to the axis
of the shaft on which the gears are
mounted. The smaller of the two meshing
gears is called the pinion; the larger is
simply designated as the gear. [Courtesy
Illinois Gear Division, Wallace Murray
Corporation. ]
 ay
Section 10-1: Spur Gear Terminology

Figure 10-2 The large gear, in the epicyclic gear set shown, has its teeth cut on
the inside and is known as an internal gear. [Courtesy Fairfield Manufacturing,
La Fayette, Ind.]

where p = circular pitchin inches; d = diameter of pitch circlein inches;

and N, = number ofteeth ofthe gear.

P==oNs 22)
(10-2

where P = diametral pitch, and

Ph=t (10-3)
The diametral pitch determines the relative size of gear teeth as can be seen
from Figure 10-5. In order for two gears to mesh, they must have the same
pitch.
When two gears are in mesh, the distance between the centers of the two
gears, is equal to one half the sum oftheir pitch diameters. In equation form
pee
(10-4)
=a)
In Figure 10-3, the center distance, c, is shown.
 Pinion

Line-of
action

Pitch circle

Pressure
Angl ,
ngle (9) Tooth profile

Pitch circle Genter

Whole depth digtance
Addendum (c)

Root
; (tooth)
f
cee) Eee fillet
Clearance centers op
Base circle
Dedendum

Circular tooth
thickness (t)

Chordal tooth
thickness

Circular pitch (p,)

Figure 10-3 The basic geometry and nomenclature of a meshed spur

gear set is shown. Some of the important features of gear design that are
shown include center distance, pitch circle diameters, tooth size, number
of teeth, and pressure angle. [Courtesy Designatronics, Mineola, N.Y.]

522
 ; Anas Clearance
Line of action : d
Pitch point

<< Pressure angle

Circular
pitch
— a

Circular
thickness Base circle diameter

|
Pitch diameter
i
Outside diameter

Base circle of Center

internal gear distance

&
Internal diameter
Root diameter
\A

fe
Internal Gear NI SS a Pinion
20 teeth T Ate 12 teeth
4/5 diametral pitch 20° poe 4/5 diametral pitch 20°

Figure 10-4 The basic geometry and nomenclature of a meshed internal gear
and pinion are shown. [Courtesy Fellows Corporation. ]

Figure 10-5 The effect of the value of

the diametral pitch in the relative size of
gear teeth is shown. Clearly, for a given
pitch diameter, the smaller the diametral
pitch the larger the gear tooth.

523
 524
Chapter 10: Spur Gears

SECTION 10-2
Backlash
The space between teeth must be made larger than the gear teeth width as
measured on the pitch circle. If this were not the case, the gears could not
mesh without jamming. The difference between toothspace and tooth width
is known as backlash. Figure 10-6 shows the backlash present between two
meshing spur gears. Backlash will be discussed in greater detail in Section
10-19 of this chapter.

ircle Ee
Base c Pinion
Line of
action
Linear backlash = B=7,—T,

Angular backlash of
Gear = 6, =B/R
Pinion = 0, = B/r
where 7, = tooth space
Tp = tooth width
Base of pinion
circle Ty = tooth width
of gear,
all measured on pitch circle

(a) Angular backlash

Backlash, along line-

of-action = B, ,=B cos¢

Line of
action

Base circle

(b) Linear backlash

Figure 10-6 Backlash, the amount by which the width of a tooth space
exceeds the thickness of the engaging tooth on the pitch circles is shown.
Backlash can be designated as (a) angular, or (b) linear. [Courtesy
Designatronics, Mineola, N.Y.]
 oy2)
Section 10-4: Involute Gear Teeth

SECTION 10-3

Fundamental Law of Gearing and Velocity

Ratio
In order for two meshing gears to maintain a constant angular velocity
ratio, they must satisfy the fundamental law of gearing. This law may be
stated as follows : The shape of the teeth ofagear must be such that the common
normal at the point of contact between two teeth must always pass through a
fixed point on the line of centers. When two gears in mesh satisfy the funda-
mental law, the gears are said to produce conjugate action.
The velocity ratio mentioned in connection with the fundamental law is
defined as the ratio of the angular speed of the driven gear to the angular
speed of the driver gear. Or, stated another way, it is the ratio of the output
speed divided by the input speed. As is obvious from the above definition,
the velocity ratio is less than 1 when the pinion is the driver and greater than
1 when the gear is the driver. °
There are several useful equations for determining the velocity ratio as
indicated by equations (10-5).

r= = 2= HH (10-5)
where
r, = velocity ratio
@ = angular velocity, rad/sec
n = angular velocity, rpm
N, = number of teeth
d = pitch circle diameter, in.
subscript | refers to the driver and subscript 2 refers to the driven gear.

SECTION 10-4
Involute Gear Teeth
In order to obtain conjugate action, most gear profiles are cut to conform
to an involute curve. There are a few gears in existence that are cut wholly or
partly in the form of a cycloidal curve and are thus able to obtain conjugate
action. However, these gears are so few in number that we will consider only
the properties of the involute curve.
The involute curve may be constructed graphically by wrapping a string
around a cylinder and then tracing the path a point on the string makes as
the string is unwrapped from the cylinder. When the involute is applied to
gearing, the cylinder around which the string is wrapped is defined as the
base circle. Gear teeth are cut in the shape of an involute curve between the
base and addendum circles, whereas that part of the tooth between the base
and the dedendum circle is generally a radial line.
 526
Chapter 10: Spur Gears

SECTION 10-5
Gear Tooth Action
To clarify the action that takes place when two gears are in mesh, consider
Figure 10-7. Gear number | (pinion) is the driver and has its center at O,,
gear number 2 whose center is at O, is the driven gear. The pitch circle radii
Base circle

Follower

Base circle

Figure 10-7 Gear tooth action between two gears in mesh is shown.
 SAH
Section 10-5: Gear Tooth Action

arer, andr,, and they are mutually tangent along the line of centers, O,0,,
at the pitch point P.
Line DE is tangent to both base circles, passes through the pitch point, and
is normal to the teeth that are in contact because these are the properties of an
involute curve. Line DE is, therefore, the line along which all points of contact
of two teeth must lie, and also the line along which the normal load that
one tooth exerts on the other lies. For these reasons DE is commonly known
as the line of action or the pressure line. Line AB is perpendicular to the line
of centers at the pitch point. The angle ¢, between AB and DE is known as
the pressure angle. Most gears are cut to operate with standard pressure
angles of 20 or 25deg. A few gears are still manufactured with pressure
angles of 145 deg, but they are used primarily as replacements for gear
systems already in existence.
It is extremely important to locate the pressure line properly. The rule to
be used is as follows: The pressure line is properly located by rotating from
the line perpendicular to the line of centers at the pitch point through an
angle equal to the pressure angle ¢ and in a direction opposite to the direction
of rotation ofthe driver. To illustrate, refer again to Figure 10-7. The pressure
line, DE, was located by rotating AB through an angle @ in the counter-
clockwise direction, since the driver (gear number 1) was rotating in the
clockwise direction.
Finally, by considering triangles O, FP and O,GP, the following equations
hold
=
im
|=17,/ COS@

ry, = 1, CoS

or, in general
Lg = ahiCOS (10-6)

where r, is the base circle radius in inches.

Example 10-1. A gear has 45 teeth, a diametral pitch of 3, and is cut with 143 deg
full depth teeth. (Table 10-1 is to be used for tooth proportions.)
Determine: (a) The diameter of the base circle and outside diameter, (b)
circular pitch, and (c) the clearance as determined by calculating the addendum
and dedendum.

Solution:

@= =P Hsin
=
|
Part a:

Now
d, = dcos @ = 15cos 144 deg
= 15 x 0.968 = 14.52 in.
 528
Chapter 10: Spur Gears

Table 10-1 Tooth System

25-de:

Per eh acorn hae ee ree ee

radeon — ; Dae eee ee :
Medenduni Li? List Ls $ ee 2 + 0.002 in. 1

Clearance 9137. O51 oz o2 O25 92 +0.002in, 22°

a el ta a Hoag hee :
ane 2/31 PAN ES 238 ‘8 22 42 + 0.002 in. 228

and
dy =d+ 2a
From table 10-1 a = 1/P = 0.333. Then
doi Ete <a 01355)— a O67 iii
Part b:

ie era
[P 5
OL
Partc:
clearance = d, — a
where d, = dedendum; and a = addendum.
From Table 10-1 d, = 1.157/P = 1.157/3 = 0.386 in.
c = 0.386 — 0.333 = 0.053 in.
As a check, From table 10-1

clearance = eae = = 0.052in. @

Example 10-2. A gear having a diametral pitch of 4 is to drive another gear at 600 rpm.
The velocity ratio is to be and the center distance between the shafts must
be 8 in. (a) Determine the number of teeth of each gear. (b) Find the value of
the pitch line velocity.
Solution:
Part a:
ay il
i= rn=

therefore

add,
Because

mie as
Caan ae aa
 529
Section 10-6: Contact Length and Contact Ratio

Therefore
d, = 4in.
and
iy = Si = Bx Ae Diva
Now
N,, =d,P =4 x 4= 16
teeth
N,, = d,P = 12 x 4 = 48 teeth
Part b:

v=o. = 2 oo.

=> x a x = = 31.416 ft/sec @

SECTION 10-6
Contact Length and Contact Ratio
When two gear teeth come into mesh, the initial point of contact occurs
when the flank of the driver comes into contact with the tip of the driven gear.
The contact ends when the tip of the driver tooth comes into contact with
the flank of the driven tooth.
Because the tips of gear teeth lie on the addendum circle, contact between
two gear teeth starts when the addendum circle of the driven gear intersects
the pressure line and ends when the addendum circle of the driver intersects
the pressure line.
Figure 10-8 shows the important points in the previous discussion. Contact
occurs at point A and ends at point B. The length of contact AB, which can
be derived geometrically, is given by equation (10-7)

AB = \/(r, + a)? — r3. cos? 6 — r, sing

(10-7)
ay Jr, + a,)* — 1? cos? d —r, sing

where r = pitch circle radius in inches; a = length of addendum in inches;

and ¢ = pressure angle, degrees.
When two gears are in mesh, it is desirable to have at least one pair of teeth
in contact at all times. The method usually employed to indicate how many
teeth are in contact is the contact ratio. This ratio is defined as the length of
contact divided by the base pitch, where the base pitch is defined as the
distance measured on the base circle between corresponding points of
adjacent teeth. The relationship between the circular pitch and the base pitch
is

Py = pcos (10-8)
 530
Chapter 10: Spur Gears

Follower

Driver

Figure 10-8 The contact length for two gears in mesh is shown.
Contact starts when the flank of the driver touches the tip of the driven
gear and ends when the tip of the driver tooth touches the flank ofthe
driven tooth.

The contact ratio is therefore given by equation (10-9).

Vf, base 120s) — 2, sin @
contact ratio =
Pp
(10-9)
i Jr; + a,)? —ricos? @ — r, sing
Pp

Most gears are designed with contact ratios between 1.2 and 1.6. For
example, a ratio of 1.4 means that one pair of teeth is always in contact, and
a second pair is in contact 40 % of the time.

Example 10-3. How many teeth will be in contact for a gear set in which the driver
is a 24-tooth 4-diametral pitch 25-deg full depth spur gear and the driven
gear has 48 teeth.
Solution:

a + ax)’ — r3cos?d — r, sing

contact ratio =
Pp

o
ve + a,)? —r?cos? @ —r, sing
Pp
 Dol
Section 10-7: Interference

Now
DaedNe ae gee
a iy oer Fe
iy Ny 8
CUS Sy
en ar
Frome Table 10a a3. P = 1/4 02s ine ay =P = 1/4 = 0250:
p = n/P = 3.1416/4 = 0.7854 in.; and p, = pcos @ = 0.7854 x cos 25 deg =
0.7854 x 0.906 = 0.712. Therefore

_ _ /(6 + 0.25)? — 670.906)? — 6 x 0.423

contact ratlo = 0.712

VG + 0.25)? — 37(0.906)? — 3 x 0.423

0.712

DN O92 129 206 = A 1

0.712

Sie See Seo ted

0.712 ne

= 1.54teeth in contact @

SECTION 10-7
Interference
Because the portion of a gear tooth below the base circle is cut as a radial
line and not an involute curve, if contact should occur below the base circle,
nonconjugate action would result (fundamental law of toothed gearing
would not hold).
The condition for which interference will occur can be demonstrated
graphically by considering Figure 10-9. Points A and B, the points of tangency
of the base circles with the pressure line, are known as the interference points.
If contact occurs outside ofthese points (if the addendum circle intersects the
line of action outside these points), interference will occur. The gears shown
in Figure 10-9 will have interference (nonconjugate action) because the
points of intersection of the addendum circles C and D lie outside the inter-
ference points.
Equation (10-10), derivable from a geometric consideration of Figure 10-9,
can be used to determine whether interference will occur.

r, = /r’ cos? @ + c? sin? ¢

(10-10)
= ./ri +c’ sin’?
 D2
Chapter 10: Spur Gears

"b2

Figure 10-9 When contact between two gears in mesh occurs

below the base circle, nonconjugate action results; the resulting
condition is called interference. In the figure, interference will
occur if contact takes place outside of points A or B. Since
contact for the gears shown occurs at C and D, interference will
result.

where

r, = radius of addendum circle, in.

c = center distance, in.
od I= pressure angle, deg
r, = base circle radius, in.

. No interference will occur if the

actual is equal to or less than the calculated value.

Example 10-4. Determine whether or not interference will occur for the following
gear set. Both driver and driven gear are 2-diametral pitch 14} deg full depth,
with 16 teeth.
Solution:

pre Gig Sens

aay.
 2) fe)
Section 10-8: Standard Gear Systems

I, = Tr, = d5 008. = 4 x cos 14} deg = 4 x 0.968 = 3.9 in,

teeta tej eS i= oy a 8 (0.25)

= ./15.2 + 4 = 4.38 in.

the maximum allowable value for radius of the addendum circle. From
Table 10-1, a, = 1/P = 4 = 0.5 in. Then the actual
PRPs ge ALES
A
Therefore, interference will occur because the actual radius of the addendum
circle is greater than the maximum allowable radius. @

There are several methods available to prevent interference.

1. Undercutting is a procedure in which the portion ofthe tooth below the
base circle causing interference is cut away. Contact therefore does not occur
on noninvolute portions of the tooth and, of course, there is no interference.
This method has two distinct disadvantages. First, the contact ratio is reduced,
resulting in noisier and rougher gear action. Second, the section modulus is
reduced at the base of the tooth (increasing the stress), which is already the
weakest portion of the tooth.
2. The tooth may be stubbed or, in other words, have a portion of the tip
removed. Again, although interference is prevented, the resulting reduced
contact ratio produces noisier and rougher gear action.
3. Increasing the pressure angle will decrease the base circle. This in turn
increases the involute portion of the tooth profile and thus eliminates inter-
ference. However, the increased pressure angle increases the separating force
(to be discussed in Section 10-12). Again, the result is poorer gear action.
4. Gears may becut with long and short addendum gear teeth. Forexample,
the driver may be cut with an increased addendum (the dedendum is decreased
proportionately) while the driven gear is cut with a decreased addendum.
Obviously, such gears are not standard, therefore not interchangeable, and
thus more expensive.
To sum up this discussion, interference must be eliminated, but the method
to be used depends largely upon the application and the designer’s experience.

SECTION 10-8
Standard Gear Systems
Most gears are manufactured to conform to the standard gear systems set
up by national associations such as the American Gear Manufacturers
Association (AGMA), the American Standards Association (ASA), and the
American Society of Mechanical Engineers (ASME). Obviously, such
standardized gears are not only interchangeable but can also be manufactured
much more economically. Modern gears are usually manufactured with 20
 534
Chapter 10: Spur Gears

or 25 deg pressure angles. However, because some gear systems in existence

use 144 deg gears, this type ofgear is also manufactured. Table 10-1 lists the
proportions for some of the more important standard gear systems. The
composite system mentioned in the table has a tooth profile that is an
involute for a portion of the tooth and a cycloidal profile for the remainder.
As can be seen in Table 10-1, gears are classified according to type ofpitch.
Although no standards exist regarding pitch the following list is a good rule
of thumb to use
Coarse pitch ae 0
Medium coarse DezpPa—als
Fine DY) <a IE xe
Ultra fine 150 < P < 200

SECTION 10-9
Common Gear Manufacturing Methods

In this section we will discuss some of the more common methods of

manufacturing gears.

Milling cutter
In this procedure, a milling cutter is machined so that it will cut the desired
gear tooth space. The cutter is then passed across the gear blank to cut out a
space. Then the gear blank is rotated to the proper position and the next
space is cut. This procedure is repeated until all of the tooth spaces have been
cut. In Figure 10-10 is shown a typical form milling cutter.
There is a disadvantage to this method of manufacture that is sufficiently
important as to limit severely the uses of gears cut by this method. Clearly,
the cutter can only be used to cut a gear ofacertain pitch and acertain number
of teeth. In other words, a separate cutter must be used not only for every
pitch but for every number ofteeth. If the gear manufacturer did this, the cost
of a gear would be prohibitive. The usual procedure is to use the same cutter
for eight to ten different numbers of teeth at a given pitch. This enables the
gears to be manufactured at a more reasonable cost, but obviously only one
out of eight or ten gears will be properly cut. As a result the milling cutter
method is seldom used, and then only for those applications where accuracy
is relatively unimportant.

Generating rack-cutter
In this method, a gear rack is reciprocated across the gear blank-face.
After each cut, the rack-cutter is simultaneously rotated through a small
angle, and the next cut is made. This procedure is followed until the end of
the rack is reached, at which time the rack and blank are repositioned.
Figure 10-11 shows a rack-cutter and the gear blank upon which the teeth
are being cut.
 Jey)
Section 10-9: Common Gear Manufacturing Methods

Figure 10-10 The formed disc cutter shown is usually used to manufacture
gears for rough, heavy duty, slow speed applications. The process, which cuts
teeth one at a time, is quite economical. [Courtesy Horseburgh-Scott. ]

The disadvantages of the rack-cutter are the length of time needed to cut
the gear and the inaccuracy that results from the need to reposition the rack.
Despite the disadvantages previously mentioned, the rack-cutter is very
often used for cutting large gears. This is primarily the case because of the
inefficiency of other cutting methods when used to cut very large gears.

Generating gear shaper cutter

This cutting procedure is very similar to the rack-cutter process previously
discussed. The primary difference is that the cutting tool is in the form of a
gear. The advantages of this type cutter are that it is a true generating pro-
cedure because there is no need to reposition the cutter during the cutting
operation and the shaping operation permits the generation ofinternal gears.
Figure 10-12 illustrates how an external spur gear is generated with a gear
shaper cutter.
The primary disadvantage of this method is that the tooth shape of the
cutter is transferred directly to the gear blank, and any error in profile of even
one tooth ofthe cutter will result in errors in the corresponding teeth cut on
the blank. However, with care in manufacturing procedures, the generating
shaper method is probably the best of the methods discussed to this point.
 536
Chapter 10: Spur Gears

Figure 10-11 The rack-shaped cutter in

the figure is generating an external gear.
[Courtesy Horseburgh-Scott.]

Generating hob
Hob generating is another method similar to the rack generation and
shaper generation processes previously discussed. The cutting tool is a hob
or worm. The hob is drawn across the gear blank, while both hob and blank
are being synchronously rotated. The essentials of the hobbing operation are
illustrated in Figure 10-13. The primary advantage of the hobbing process is
that no repositioning of the hob is required, as was the case with the rack
generation. Also, because each gear blank tooth is cut by several hob teeth,
the effect of any one hob tooth profile error can be minimized. Hobbing is
probably the most popular of the gear cutting methods.

SECTION 10-10
Other Gear Production Methods
The methods to be discussed in this section produce gears that are low in
cost, can be produced in large volumes, but are poor in quality.
Die Casting. Gears are die cast by forcing molten metal under pressure into
a form die. As is true of all casting processes, the changes due to shrinkage,
 337
Section 10-10: Other Gear Production Methods

16 roughing teeth

Gap

Rough-cut gear
16 Teeth

Gear blank
Series gap-type

Generating cutter

i ONE AiG l
hori | |
nit ty kf |
A At at A
|
Wl
“ en ey anSays
Figure 10-12 The “conjugate” method of generating with a reciprocating
gear shaper cutter is illustrated. The cutter shown has a gap separating the
roughing and finishing teeth. The advantages of this “‘interrupted”’ generation
method are greater accuracy, higher production, and lower costs. [Courtesy
Fellows Corporation. }

the accuracy with which the mold is constructed, and so on, result in a cast
gear that is of rather low quality. In order to improve the precision of the
gear, additional finishing operations of the type to be discussed in Section
10-11 are required. However, the added expense of the finishing operation
negates the economical advantage gained from casting. As a result, casting is
generally used only as a means of producing large numbers of relatively low
quality gears.
Drawing. In the cold drawing process, the metal is drawn through several
dies and emerges as a long piece of gear from which gears of smaller widths
can besliced. The advantages and disadvantages are similar to those discussed
in the casting process.
Extruding. A process in which the metal is pushed through rather than
being drawn through the dies. This method is similar to the drawing process,
except that the material used is usually hot.
 538
Chapter 10: Spur Gears

Figure 10-13 The hobbing process, when used to cut gears, is very fast for
coarse pitches, better for higher hardnesses, and has the advantages of a generated
gear. The hobbing process is usually used to manufacture spur gears which must
operate smoothly at higher speeds. [Courtesy Horseburgh-Scott. ]

Sintered Powder Gears. The sintering process consists of applying pressure

and heat to a metallic powder in order to form the gear. The precision
obtained by this process is better than with the others discussed but, because
of the high cost of the equipment needed, production must be high in order
to make the process feasible.
Stamping. In this process, a press and a die are used to cut out the gear
shape. The gear that results is of necessity quite thin because the stamping
process is limited to rather thin materials. In order to obtain gears of normal
widths, some form of lamination is required. Press fitting, welding, or similar
procedures can be used to fasten together the number of stamped gears
needed to produce the required width.
 539
Section 10-11: Gear Finishing Methods

Injection Molding. With the advent of plastic materials, it became possible

to produce nonmetallic gears by the injection molding process. Large
quantities of gears can be produced by this process, but their quality is
relatively poor.

SECTION 10-11
Gear Finishing Methods
Grinding. A process whereby the finish on gears is improved and the
tolerance of the dimensions to which the gear is to be cut is enhanced. This
is accomplished by the use of some form of abrasive grinding wheel, Figure
10-14.
Shaving. A machining operation that removes small amounts of material
to accomplish the same result as grinding.
Burnishing. A process whereby the gear to be smoothed is rolled with a
specially hardened gear.
Lapping. A method whereby a gear is run with another gear that has some
abrasive material embedded in it.
Honing. A process that employs a tool known as a hone to drive the gear
to be finished. The hone, which can be abrasive impregnated or coated, drives
the gear in both directions.
Of the refining methods discussed, grinding and shaving are used to finish
precision gears, while the others are used mostly for low quality gears.

Figure 10-14 Gear grinding is a tooth

finishing process that greatly enhances the
gear tooth accuracy. The grinding operation
results in gear teeth whose tooth profile,
tooth spacing, and backlash can be well
controlled. [Courtesy Horseburgh-Scott. ]
 540
Chapter 10: Spur Gears

SECTION 10-12

Tooth Loads
In designing gears, the power to be transmitted and the angular speeds
involved are usually known. From this information the torque to be trans-
mitted from one gear to another can be calculated from equation (10-11).

ms Tn
hp ~ 63,000 (10-11)

where hp = input horsepower, n = revolutions per minute, and T = torque

in inch-pounds.
The force that the tooth of one gear exerts on the meshing tooth of the
other gear is normal to the tooth surface, and therefore acts along the
pressure line. Figure 10-15 shows two gears having their teeth in contact at
the pitch point, P. The normal force F,, is the force exerted by the follower
gear tooth on the driver gear tooth. Obviously, the driver tooth would exert
a force equal to F, in magnitude, but acting in the opposite direction. The

Forces Acting on Driver Tooth

oh

Z| Follower

a> Pitch circle

2 Dedendum circle

Addendum circle

Pitch circle
Addendum circle

Dedendum circle

Driver

E
Figure 10-15 Two gear teeth in contact at the pitch point. The
normal force and its components, the radial and tangential force, are
shown.
 54]
Section 10-12: Tooth Loads

normal force can be resolved into two components F, (tangential force) and
F, (radial force) related by equations (10-12) and (10-13).

F, = F,cos@ (10-12)

and

F =F, sng = F tang (10-13)

where @ = pressure angle, degrees.

Straight Spur Gears

Figure 10-16 The loads acting on the bearings supporting the shafts upon
which two meshing gears are mounted are shown. [Courtesy New Departure—
Hyatt Bearings Division, General Motors Corporation. ]
 542
Chapter 10: Spur Gears

In design it is assumed that the tangential force remains constant as the

contact between two teeth moves from the top of the tooth to the bottom of
the tooth. The torque that the normal force produces with respect to the
center of the gears is
d, pis
(Pils naoe 08 Oe F5 — -14
(10-14)

where d, is the pitch circle diameter in inches.

The pitch line velocity is given by
m dn
VSS
pen (10-15 )

where V, = pitch line velocity in feet per minute.

Substituting these values in equation (10-11) results in
h Th _ F(d/2)((12V,)(nd)]
P* 63,000 63,000
Therefore

pe hp)(33,000
ELERRECD aan
P

An important consideration in designing systems that make use of gears

to transmit power is the proper design of the bearings supporting the shafts
upon which the gears are mounted. This basically means een the
resultant bearing loads.
Figure 10-16 shows two straight spur gears in mesh, with the tooth loads
and bearing loads represented vectorially.

Example 10-5. Determine the reactions on the bearings supporting the shaft upon
which gears 2 and 3 are mounted. Gear number 1, the driver, is a 20-tooth
4-pitch 20-deg full depth gear, rotating at 2000 rpm clockwise and transmitting
100 hp. Gear number 2 has 40 teeth, and gear number 3 is a 5-pitch, 25-deg
pressure angle with 30 teeth. Gear number 4 has 45 teeth.
 543
Section 10- 12: Tooth Loads

Solution:
For gear number 1:

dy = ~ = = = Di

V,, = oe ages > £000 = 2620 ft/min

= eee = — 100 = 1260 Ib
Pi

F,, = F,, tang@ = 1260 x tan 20 deg

lI 1260 x 0.364 = 458 Ib

For gear number 3:

N, 20
iy = ava = 2000 x Ans 1000 rpm
(2

New eo)
d, Pp = ie; =6in

md3n; mx 6x 1000 :
Don oP) 1D = 1573 ft/min

_ (33,000)(hp) — 33,000 x 100 _

jae V = 1573 = 2100 lb
P3

F,, = F,,tan@ = 2100 x tan 25 deg

= 2100 x 0.466 = 980 Ib

The tooth loads on the gears.of the shaft in question are shown in the
following figure:
 544
Chapter 10: Spur Gears

In order to determine the bearing reactions we consider the front view and
the top view as shown in the following figure.
Top View

2 ins 12 in. 4 in.

1260 2100

ME = OF 12602
00 1 e Be
2520 + 29,500 32,020
Be 18 Lar See 1780 Ib

=F, = O = 1260 + 2100 — 1780 — A,

A, = 1580 Ib
Front View

XM, = O = 458 x 2 — 980 x 14 + 18B,

—916 + 13,700 _ 12,784
= = = HO
%, 18 18 Pu
EF, = O = —A, + 458 — 980
+ 710
A, = 188lb @
When a combination of several gears is used in a system, as in a gear train,
the problem is a little more complex. Often the shafts supporting the gears lie
in different planes. For this situation, the tangential and radial components
of force of one gear must be further resolved into components in the same
plane as the components of the meshing gear, thus permitting the algebraic
addition of forces along two mutually perpendicular directions. Figure 10-17
shows such a system (also see Chapter 7).

SECTION 10-13
Beam Strength of Spur Gear Teeth
Designing gears presents an extremely difficult problem because it is
primarily a trial and error procedure. However, there are several methods that
can be used to develop a design. We will follow the procedure of getting a
first design by the simplest method (Lewis equation) and then analyze and
 545
Section 10-13: Beam Strength of Spur Gear Teeth

Gear 1

iti
ZT BIBIBIGIBIRINIII
TT
CTT TT
PEAT

et

UT
TT
UT

CTI
TT
TTT

Figure 10-17 The forces acting on various gears making up a gear

train. [Courtesy New Departure—Hyatt Bearings Division, General
Motors Corporation. ]

modify this design by the AGMA method. This approach has the advantage
of showing students the fundamental and professional methods that are
available. At a later date, when they have gained some experience, they will
be in a position to make judgments as to which approach will give better
results in a given problem.
Gears generally fail because the actual loads applied to the teeth are greater
than the allowable loads based upon either the beam strength of the tooth
(tooth fracture) or its wear strength (surface failure). We will first consider
how the designer may reasonably determine the beam strength of a gear
tooth.
Wilfred Lewis, in a paper published in 1892 [2], derived an equation for
determining the stress in a gear tooth by treating the tooth as a cantilever
beam. This so-called Lewis equation is still used for preliminary design
calculations or in those instances where a high degree of accuracy is not
necessary.
Figure 10-18 shows a gear tooth with the force acting at the tip of the tooth.
This normal force F,,, is resolved into components F, and F, acting at point
A, the intersection ofthe line of action of the normal tooth load and the center
of the tooth. The usual assumption made at this point is that the load is
uniformly distributed across the width of the tooth, b.
 546
Chapter 10: Spur Gears

Figure 10-18 The geometric factors that

are used to determine the Lewis form factor.

This assumption is wrong because the gear as well as its mountings are
elastic, and we would therefore expect a nonuniform load distribution. The
basic Lewis equation can be modified to include the effect of nonuniformity
of load distribution. However, in deriving the simplified Lewis equation, we
will accept the assumption of uniformity. To minimize the difficulties resulting
from this assumption, most designers limit the value of gear tooth face width
compared to tooth thickness. This is done by limiting the ratio of tooth
width to circular pitch to a maximum value of 4 or S. It should be emphasized
that the above limits are merely suggested ones, and many exceptions do
occur. However, the designer would do well to investigate the possible effects
of nonuniformity of load distribution whenever the recommended limits are
exceeded.
The radial component F, will produce a uniform, direct compressive stress
over the tooth cross section, alle the ne Oe Fe sl produce a

ig 1in determini 00th To furthen jagtifptie as

assump-
tion we are making, it is clear that including the direct compressive stress will
increase the bending stress on the compression side of the tooth and decrease
the resultant stress on the tensile side. For the many gear materials that are
stronger in compression than in tension, the assumption made will result in
a stronger tooth design. The final and clinching argument for ignoring the
direct compressive stress is as follows. Because gear teeth are subjected to
fatigue failures and the fatigue failures start on the tension side of the tooth,
the direct compressive stress reduces the resultant tensile stress and thus
strengthens the tooth.
 547
Section 10-13: Beam Strength of Spur Gear Teeth

If we now treat the tooth as a cantilever beam having its fixed end at BD,
the stress may be obtained from equation (10-17).
MCSE IN t/2) Oli
I be/i2 be?
abt?
a (10-17)
Ina well designed beam, the stress should be uniform in every section of the
beam. In other words, in a gear tooth of uniform strength the stress is a
constant, and since the gear width and gear load are also constants, equation
(10-17) can be rewritten as

ob
L = ——t* = constant x t?
6F t

Clearly this is the equation of a parabola. Referring to Figure 10-18, the

weakest section of the tooth, BED, can be determined by inscribing the
parabola through point A, and locating the points at which the parabola is
tangent to the tooth profile, B and D. Equation (10-17) was therefore derived
for the section having the maximum stress.
Triangle ABE is similar to triangle BCE and thus
x t/2 t*
ke ps je ee
Dia SE wa x
abt? 4x

multiplying and dividing by the circular pitch p, gives

4
F.=ob—x?® (10-19)
0: pit

since x and p are geometric properties depending upon the size and shape of
the tooth, it is possible to define a factor

Bak
y
1h
y is the so-called Lewis form factor and therefore permits us to write the
Lewis equation as
F. = obyp
1
(10-20)
Because the diametral pitch, rather than circular pitch, is usually used to
designate gears, the following substitution may be made: p = x/P and
Y = ay
Yu Y
EF top
ob 0b =
obs 10-2
(10-21)
 a1qeL Z-OL sanjeA
JOY SIMA] WIOY 10984

peoq
18 sdij peoy 1eaN A[PPIW

548
02 30p “da OZ ap qnis %pI 8ap Gd 07 32p Cd
x & x & A a &
i
« hi es

A
8€70
| 9L0°0
céI0
| 190°0 97770
| 7L0'0
| 6870
| c60'0 6S7'0
| c80°0
OUEW
|) EOD S70
| 82400 II€0}
| 6600 LL70
| 8800 Sse0

€772°0
| ILO°0 +9770
| €80'0
| vceO
| €OlO £670
| £60°0 LES

9€70
| S400 9170
| 880°0
| 6€€0
| 8010 LOE'0
| 860°0 66€°0

S70
| 8400 6870
| c60'0
| 6ve0 O@E0
| COTO Slv0

ssz0
| 180°0 S6z'0
| F600
| 09€0
| SITO CEEO
| 9010 Ofr'0
£970
| v800 ZOE'0
| 960°0
| 89€0
| LITO creo
| 6010 9bP'0

OL70
| 980°0 80¢0
| 8600
| LLEO
| OcI'0 @Se0
| CILO 6SP'0

LL70
| 8800 bIE'O
| OOI'0
| 98€°0
| €cl'0 I9¢0 SILO} ILv'0
€870
| 0600 WED
| COW
|| SEED
|) SAD 69€°0
| 8IT'0 180
6870
| 760'0 97¢E0
| POl0
| 66€0
| L710 LLE0
| OcI'0 06r'0
7670
| €60°0 O€E'O
| SOl‘0
| POO
| 6cl0 ¥8E'0
| cCl'0 96P'0

(a6
9670
| vo00 EEEO
| 9010
| 800
| O€l'O cOS'0

€C
ve 7OE0
| 960°0 LEEo
| LOO
| IIv0 ce0'I 96€°0
| 9CT'0 60S'0
 a19e1 c-OL
(penunuos)

peoy
18 Sdip
peroT Jean SIPPIW
OZ Bap qnig SZ Bap %pI 3p dd 02 3p Ga

Sc SOE0 L600
| Ove'0
97 0 80€ 8600
| bre
LZ ITEO 6600
| 8vE0

0¢€ si¢o
| 1010
| 8S€0
I€ ozE'
|o Iolo
| I9€°O
cE zz¢-0
| Lolo
| v9E'O

549
 aiqey 2-O1 (Panunuos)

550
peoy
qv sdij peo] JE2N SIPPIW

0 vl.v 880

08r'0 96S°0

0 sr £€09°0

88Pr0 L090

0 6b 90 Ol

96F'0. £19°0

66r'0 $190

0» €OS 619°0

90S°0 GED

810 Se9'0

0v9'0

0S9°0

099°0
 Dol
Section 10-13: Beam Strength of Spur Gear Teeth

Values for the Lewis form factor have been computed for standard gear
systems and are readily available (see Table 10-2).
An examination of equation (10-21) shows that the maximum allowable
tangential or transmitted load can now be determined if the allowable stress
for the gear material is used. In order to avoid confusion, the usual procedure
is to designate the allowable load based on bending as F,. Therefore, we will
write the Lewis equation as
Y
F, = Sbyp = Sb= (10-22)
At this point it would be well to consider the SUSE of another SSSI
made in deriving the Lewis Soon [ was assumed the i e
oad F, actec he tip of the tooth. Because most gears are designed with a
contact ratio belive 1.2 and 1.6, itis clear that when the load acts on the tip
of one tooth, another tooth is still in contact, and the full load does not act on
the number one tooth.
In Figure 10-19, the load has been moved away from the tooth tip to a
point near the middle ofthe tooth (the second tooth has gone out of contact,
and the full load does act on the tooth shown). As can be seen from the
diagram, the derivation of the Lewis equation would follow exactly as the
previous one. The only difference would be in the values for the Lewis form
factor. Pa C ste y c f L : rm acto} PN 1 9

Ts nn ortie eon center will reduce the size and weight of the gears
because a smaller actual stress is being used. However, the equation is only
used for those designs where weight and size reduction is of major importance.

Figure 10-19 The load acts near the center of

the tooth, the other pair of teeth, which were
originally in contact, have separated, and the total
load being transmitted acts on the tooth shown.
 DZ,
Chapter 10: Spur Gears

At this point in our discussion, it is possible to design a gear, by using

equation (10-22), that will have the required strength, providing the proper
allowable stress for the gear material is used. The allowable stresses presented
in Table 10-3 are based upon investigations of the performance of gears in
actual use. Therefore, gears to be used for ordinary applications will give
satisfactory service. However, a much better design procedure is to use
equation (10-22) and Table 10-3 to come up with a preliminary design and
then to check the gear for the items to be discussed in succeeding sections.

Table 10-3 Safe Static Stresses for Use in The Lewis Equation

Material _ So, psi BHN

Gray cast iron

ASTM 25 8,000 174
ASTM 35 12,000 212
ASTM 50 15,000 223
Cast steel (low carbon)
0.20% C not heat treated 20,000 180
0.20% C WQT 25,000 250

Forged carbon steel

SAE 1020 case hardened and WQT 18,000 156
SAE 1030 not heat treated 20,000 180
SAE 1035 not heat treated 23,000 190
SAE 1040 not heat treated 25,000 202
SAE 1045 not heat treated 30,000 215
SAE 1045 hardened by WQT 32,000 205
SAE 1050 hardened by OQT 35,000 223

Alloy Steels
SAE 2320 case hardened and WQT $0,000 225
SAE 2345 hardened by OQT $0,000 475
SAE 3115 case hardened and OQT 37,000 212
SAE 3145 hardened by OQT $3,000 475
SAE 3245 hardened by OQT 65,000 475
SAE 4340 hardened by OQT 65,000 475
SAE 4640 hardened by OQT $5,000 475
SAE 6145 hardened by OQT 67,500 475

Copper base materials

SAE 43 (ASTM B147-52, 8A) (manganese
bronze) 20,000 100
SAE 62 (ASTM B143-52, 1A) (gun metal) 10,000 80
SAE 65 (ASTM B144-52, 3C) (phosphor
bronze) 12,000 100
SAE 68 (ASTM B148-52, 98) (aluminum
bronze, heat treated) 22,000 180

Nonmetals
Bakelite, Micarta, Celeron 8,000
SSS
 JS
Section 10-14: Stress Concentration

SECTION 10-14

Stress Concentration
Another important factor affecting the stress in a gear tooth, and not
included in the simplified Lewis equation, is the stress concentration existing
at the root of the tooth. It is extremely difficult to determine the theoretical
values of stress concentration for the rather complex shape of a gear tooth.
Fortunately, photoelastic investigations have resulted in empirical equations
giving reasonable values for stress concentration. Dolan and Broghamer [3]
established equations (10-23a-c).
0.2 t 0.4
if
K, = 0.22 + r (3 for 145-deg pressure angle _—(10-23a)

t 0.15 t 0.45
K, = 0.18 + ( for 20-deg pressure angle (10-23b)
r

t 0.11 t 0.5
K, = 0.14 + (4 (7 for 25-deg pressure angle —(10-23c)

where K, = stress concentration factor; t = thickness of tooth at base of

tooth (weakest section); r = radius of fillet at root of tooth; L = height of
load above weakest section.
Figure 10-20 shows the stress concentration present in gear teeth.
Since the stress concentration as obtained by the above equation is for
static stress and a gear tooth is subjected to fatigue stress, the factors obtained
by equations (10-23) should be modified by the notch sensitivity factors, q
discussed in Chapter 3.

Equation (10-22) can now be modified to include the effect of stress

concentration, and can be used for final design. ac |
)Vi
/1 — y

hae tO (10-24)

Figure 10-20 Stress concentration in

gear tooth. [Courtesy T. J. Dolan,
Univ. of Illinois.]
 554
Chapter 10: Spur Gears

SECTION 10-15
AGMA Equation!
The final strength equation to be presented is the AGMA modification of
the Lewis equation. This equation is particularly useful to the designer because
it applies correction factors to the original Lewis equation that compensate
for some of the erroneous assumptions made in the derivation as well as for
important factors not originally considered. Furthermore, since most of the
factors are obtained empirically, the equation can be kept up to date by
merely changing the values of the factors as more information about gear
behavior is obtained.
The equation is written as follows:

6, = desea (10-25)
K,,bJ
where
o, = calculated stress at root of tooth, psi
F, = transmitted load, lb
K,, = overload correction factor
P = diametral pitch
K, = size correction factor
K,,, = load distribution correction
K,, = dynamic factor
b = face width, in.
J = geometry factor
As an aid in understanding and using the equation, the following discussion
of the correction factors is presented.
The overload correction factor, K,, accounts for the fact that, while F, is
an average value for the transmitted load, the actual maximum load may be
as much as two times as great due to shock loading in either the driven or the
driving system. Table 10-4 gives some suggested values for K,.
The size factor K, is designed to account for nonuniformity of material
properties. It is a function of tooth size; diameter of parts; ratio of tooth
size to diameter ofparts ; face width ; area of stress pattern ; ratio of case depth
to tooth size; quality, hardenability, and heat treatment of materials;
magnitude and gradient direction of residual stresses. For spur gear applica-
tions the size factor is usually taken as unity.
The load distribution factor K,, depends upon the combined effects of
misalignment of axes of rotation due to machining error and bearing
clearances ; load deviations ; elastic deflection ofshafts, bearings, and housing
‘The material in this section is extracted from AGMA information sheet, Strength of Spur,
Helical, Herringbone, and Bevel Gear Teeth (AGMA 225.01) with the permission ofthe publisher,
the American Gear Manufacturers Association, 1330 Massachusetts Avenue, N.W., Washington,
D.C. 2000S.
 2p))
Section 10-15: AGMA Equation

Table 10-4 Overload Factor, K, (for speed

increasing and decreasing drives)*
SS
SS EN a oe ey

Load on Driven Machine

Power
Source Moderate Heavy
Uniform Shock Shock

Uniform 1.00 1:25 1.75 or

higher

Light shock 2S) 1.50 2.00 or

higher

Medium shock 1.50 eS 2.25 or

higher

“For speed increasing drives of spur and bevel gears (but not helical and herringbone gears),
add 0.01 (nginpy)? to the factors in Table 10-4, where Nip = number of teeth in pinion;

Ntg = number of teeth in pinion;

SOURCE: AGMA

due to load. The AGMA information sheet presents tables and figures that
give values for K,, when misalignment information is available. However,
when estimated or actual misalignment is not known, the K,,, factor for spur,
helical, and herringbone gears as given in Table 10-5 may be used.
The dynamic factor K,, depends on the effect of tooth spacing and profile
errors; effect of pitch line speed and revolutions per minute; inertia and
stiffness of all rotating elements; transmitted load per inch of face; tooth
stiffness. Figure 10-21 shows three of the commonly used dynamic factors.

Table 10-5 Load Distribution Factor, |B

Face Width

2-in. Face
Condition of Support and

Accurate mounting, low bearing

clearances, minimum elastic
deflection, precision gears

Less rigid mountings, less

accurate gears, contact across
full face

Accuracy and mounting such

Over 2.0
that less than full face
contact exists

SOURCE: AGMA
 556
Chapter 10: Spur Gears

1.00

0.80

2 2)

Ky
factor,
Dynamic
2 + rs)

0.20 oat j

i
0 1000 2000 3000 4000 5000 6000 7000 8000
Pitch line velocity, V,, fom

Figure 10-21 Dynamic factor, K,. [Courtesy American Gear Manufacturers’

Association. ]

Curve 1 is used for (1) high precision helical and shaved or ground spur
gears where the effects previously mentioned do not cause an appreciable
dynamic load to be developed and (2) generated bevel gears having the
preferred pattern of tooth contact and accurate tooth spacing and concen-
tricity. Curve 2 is used for (1) high precision helical and shaved or ground
spur gears where the effects previously mentioned can develop an appreciable
dynamic load, (2) commercial helical gears, and (3) large planed spiral bevel
gears. Curve 3 is used for (1) spur gears finished by hobbing or shaping and
(2) large planed straight bevel gears.
Lower dynamic factors than those shown in the figure must be used when
milling cutters are used to cut the teeth or inaccurate teeth are generated.
The geometry factor J accounts for the effect of the shape of the tooth, the
position at which the most damaging load is applied, the stress concentration,
and the sharing of load between one or more pairs of teeth.
Tooth shape depends upon the geometry of the tooth system. In other
words, such factors as pressure angle, number of teeth, whether full depth,
stubbed, and so on.
The position at which the most damaging load is applied depends upon
how accurately the gears have been cut. For accurately cut teeth, the greatest
stress will occur when the load acts at the greatest height for which only one
pair of teeth are in contact. In other words, when the load acts at the tip of
 yA
Section 10-15: AGMA Equation

Pa
&
lo}
She TS SSS = Pinion addendum 1.000 = 8
iS
3 Gear addendum 1.000 cs
0.60 5 z & 8
s iat cee
WES 4 s:
Seo as
re
1000 8 &
Nee ZO Rae
= 85 Q &
0.50 50
35 8 6
25 aol
ne
0.45 17
Generating rack 1 pitch
ayes Number of teeth
: in mating gear
J
factor,
Geometry
0.35

0.30 0.30

0.25 Load applied at tip of tooth 4 0.25

0.20 — 0.20

12
SEE15 «tas
17
wala 24
20
Kn35 aT
30
aed
40455060 80
een La
125 275 oo
=|
Number of teeth for which geometry factor is desired

Figure 10-22 Geometry factors, 20 deg spur, standard addendum.

[Courtesy American Gear Manufacturers’ Association. ]

€ SS Se Pinion addendum 1.000 of

= Gear addendum 1.000 ra

Z é as
aes a <I
5

0.65 -— ‘3 ae
eg 22
5 € 0.65
= 1000 © oO
= Wo. eh
0.60 6 = 2 0.60

ee
50 a2
as

4) * me}
Se Generating rack 1 pitch 7 bs 2 Oo
mal)

0.50 0.50
im Number of teeth
in mating gear
0.45}—
factor,
Geometry
J
S 0.45

0.40 0.40

0.35t 0.35
Load applied at tip of tooth
0.30 0.30

0.25 I a | 0.25
12 15 7 20 24 30 35 40455060 80 125 275 ©
Number of teeth for which geometry factor is desired

Figure 10-23 Geometry factors, 25 deg spur, standard addendum.

(Courtesy American Gear Manufacturers’ Association. ]
 558
Chapter 10: Spur Gears

Table 10-6 Guide to Limiting Error in Action for Steel Spur Gear
(variation in base pitch)

Allowable Error When Amount of Error When

Number Teeth Share Load Teeth Fail to Share Load
of

wo |co [om |wo |e |

Pinion See per in, of Face, lb Load per in. of Face, lb
Teeth

15 0.0004 | 0.0007 |} 0.0014 | 0.0024 | 0.0042 | 0.0006 } 0.0011} 0.0023 |} 0.0039 0.0064
20 0.0003 | 0.0006 | 0.0011 | 0.0020 | 0.0036 | 0.0006 } 0.0011} 0.0023 |} 0.0039
25 and higher 0.0002 0.0005 | 0.0009 | 0.0017 | 0.0030 | 0.0006 | 0.0011} 0.0023 | 0.0039
SOURCE: AGMA

the tooth another pair of teeth are in contact and are sharing the load and
thus the greatest stress does not occur at this position. For less accurately cut
gears errors may prevent two pair of teeth from sharing the load, and the
greatest stress probably will occur when the load acts at the tip of the tooth.
The geometry factor also includes the effect of stress concentration as
determined by the Dolan-Broghamer equations discussed in Section 10-14,
as well as tangential (bending stress) load and radial (compressive stress)
load.
Figure 10-22 may be used to determine the geometry factor for 20-deg spur
gears, and Figure 10-23 may be used to determine the geometry factor for
25-deg spur gears. Both of these figures are presented by the AGMA and
assume that the theoretical stress concentration factor is not greatly affected
by such factors as surface finish, plasticity, residual stresses, or other factors.
As can be seen from Figures 10-22 and 10-23, the upper set of curves are
used when the load is applied at the highest point of single tooth contact,
while the lowest curve in each figure is used when the load acts at the tip
of the tooth.
Table 10-6 lists the allowable variations in base pitch between gear and
pinion that can be used to determine whether load sharing exists or not. For
errors less than those shown in the left hand half of Table 10-6, the upper
curves in Figures 10-22 and 10-23 may be used. If the errors are greater than
those shown in the right hand half of the table, the lower curve in both
Figures 10-22 and 10-23 is to be used. Obviously, there will be cases where
the error lies between the two extremes shown in the table. In this situation,
the designer must exercise judgment in the geometry factor by considering
such things as material hardness, surface finish, wearing-in, and lubrication.
All the information needed to calculate the actual bending stress by the
AGMA formula is now available. All that remains is to compare this stress
with the maximum allowable design stress. The AGMA equation for maxi-
mum allowable design stress is

Sa => (10-26)
Table 10-7 Strength of Spur, Helical, Herringbone, and Bevel Gear Teeth

Min.
Sats PSi
; Heat Material Hardness
Material or Min. Tensile
Treatment Strength Spur, Helical, and Herringbone | Bevel

Steel Normalized 140 BHN 19-25 ,000 11,000

Quenched and tempered 180 BHN 25-33,000 14,000
Quenched and tempered 300 BHN 36-47,000 19,000
Quenched and tempered 450 BHN 44-59,000 25,000

Case carburized Bs) [K 55-65 ,000 27,500

Case carburized 60 RK, 60-70,000 30,000
|
oe

Induction or flame
Hardened, hardness

Hoa usos
Footnote |
54 RK 45-55,000"
Hardness Pattern 54 R, at
of Footnote 1 Hardened | 22,000 13,500

mle 53 R, case -42

* 2 0,000
Nitrided AISI 4140 300 BHN core 37-42,000

Cast Iron
AGMA Grade 20 5,000 2,700
AGMA Grade 30 175 BHN 8,500 4,600
AGMA Grade 40 200 BHN 13,000 7,000

Nodular Iron
ASTM Grade 60-40-18 Annealed 15,000 8,000
ASTM Grade 80-55-06 20,000 11,000

ASTM Grade 100-70-03 | Normalized 26,000 14,000

ASTM Grade 120-90-02 Quenched and tempered 30,000 18,500
all
Bronze
AGMA 2c (10%-12% Tin) 40,000 psi 5,700 3,000

Aluminum Bronze
ASTM B-148-52
Alloy 9C-H.T.

* Values for teeth 6 DP and finer

Footnote (1)

Hardened
TPor heavy gears these hardnesses will be lower,
section hence, lower values of allowable stress should
be used,

Pattern B

Pattern A

SOURCE: AGMA

559
 560
Chapter 10: Spur Gears

where

S.q = Maximum allowable design stress, psi

n I allowable stress for material, psi
K, = life factor
K, = temperature factor
K, = factor of safety (reliability factor)
The allowable stresses for standard gear materials will vary with such
factors as material quality, heat treatment, forging or casting practices, and
material composition. Table 10-7 gives the values of the allowable fatigue
bending stress (S,,) as recommended by the AGMA. Note that R, refers to
case carburized Rockwell hardness numbers.

— — — Quenched and tempered

Annealed or normalized

Allowable
yield
Sy
kpsi,
strength,

Brinell hardness number

Figure 10-24 Allowable yield strength, S,,. [Courtesy American Gear

Manufacturers’ Association. ]
 561
Section 10-15: AGMA Equation

Table 10-8 Life Factor, K,

Spur, Helical and Herringbone Bevel Gears

Number
f
anes 160 250 450 Case Case
BHN *BHN BHN Carburizedt } Carburizedt

Up to 1,000

10,000 all

100,000 Ps)

1 million 1.4

10 million 1.0

100 million
and over

*The life factor for over 100 million cycles is sometimes handled by service factor, the factor of safety, or by
reduced allowable stresses, For reference purposees, see Standard AGMA 151.02, Application classification
for helical, herringbone and spiral bevel gear speed reducers, and Standard AGMA 411.02, Design procedure
for aircraft engine and power take-off spur and helical gears.

¥Case carburized $5-63R>.

SOURCE: AGMA

As can be seen from Table 10-7 some ofthe allowable stresses are given as a
range of values. The lower range values should be used for general design,
whereas the upper range values are suggested when high quality material is
used, heat treatment is sure to be effective, and good quality control is
possible. The AGMA also suggests that 70% of the value given in Table 10-7
be used for designing idler gears and other gears subjected to reversed
loading.
For gears that are subjected to infrequent momentary high overloads, the
AGMA suggests using the allowable yield properties to determine the maxi-
mum allowable design stress, rather than the fatigue strength of the material.
Figure 10-24 can be used to obtain allowable yield strengths for through-
hardened steels. :
The life factor K, corrects the allowable stress for the required number of
stress cycles. Table 10-8 lists the recommended values for the life factor.
The temperature factor K, is an attempt to adjust the allowable stress for
the effect of operating temperature. For oil or gear blank temperatures less
than 250°F, K is usually taken as unity.
An alternative correction factor may be obtained by using equation
(10-27)

K
_ 460 + T, (10-27)
4 620
where 7, = peak operating oil temperatures in degrees Fahrenheit. This
equation is used for case carburized gears at temperatures above 160°F.
 GQ62
Chapter 10: Spur Gears

Table 10-9 Factors of Safety, K,

(fatigue strength)

Requirements of Application Kr
SSS qQQQ';_QuuN
High reliability 1.50 or higher

Fewer than | failure in 100 1.00

Fewer than | failure in 3 0.70

SOURCE: AGMA

Table 10-10 Factors of Safety, K,

(yield strength)

Requirements of Application Kp

High Reliability 3.00 or higher

Normal Design 1.33

SOURCE: AGMA

The factor of safety Kp, sometimes called the reliability factor, is introduced
into the equation in order to ensure high reliability or, in some cases, to allow
for designing for calculated risks. Table 10-9 lists some typical Kz values
that are to be applied to the fatigue strength of the material. It should be
pointed out that, although these factors are smaller than one would normally
expect in machine design, they are applied to the fatigue strength of the
material; thus, a failure simply means that the gear will have a shorter life
than the minimum designed for.
Table 10-10 lists safety factors to be used for noncarburized gears. These
factors are applied to the yield strength of the material, and the maximum
load to which the gears are subjected.
To sum up the AGMA method for designing spur gears for strength, the
calculated stress o, of equation (10-25) must always be less than or equal to
the maximum allowable design stress as determined by equation (10-26).

Example 10-6. A conveyor system is to be driven by an electric motor turning at

1200 rpm. The velocity ratio of the gears connecting the motor and conveyor
is to be 1:3. The pinion is a 6-diametral pitch, 18-tooth 20-deg full depth
140-BHN steel gear. Both gears are of the same material and have a face
width of 2in. What is the maximum horsepower than can be transmitted,
based upon strength only and using the AGMA method?
Solution: By equation (10-26)

- SaKr
ad Koko
 563
Section 10-15: AGMA Equation

From Table 10-7, the lower value of 19,000 psi will be used because it falls
in the category of general design. The life factor, from Table 10-8, will be
taken as 1, based upon the assumption that we are designing the system for
indefinite life. We may assume that the oil temperature is less than 160°F;
therefore, K; may be taken as unity. Kp will be taken from Table 10-9 as 1.2.
For this application this value will give reasonable reliability.
Before proceeding with the solution, it should be emphasized that this type
of decision making (choosing reasonable values) is common in all design
work. The student will doubtless experience considerable difficulty in making
these decisions but, as a designer gains more experience, decisions will be
more easily reached.
Substituting the chosen values into equation (10-26), we get
_ 19,000 x 1
= 15,800 psi
oo Tia
This value is now used for a, in equation (10-25), and that equation is solved
for the maximum allowable transmitted load. In other words
15,800 x K,bJ
I ere
aeKe Keke a (a)
The pitch diameter
N, 18 :
d, = P
—2 = F
—_ = 3 in.

711] ne ae ea WAU 0)
a
V. = 1D
P = 1D 942 ft/min
i

From curve 3 of Figure 10-21 K, = 0.62.

The geometry factor J is obtained from Figure 10-22, with the upper
curves used; in other words the load is assumed to act at the highest point
of single tooth contact. In practice, the determination as to whether to use
upper or lower curves would be made by using Table 10-6 and having some
information regarding the error in action for the gear.

N= ty N,, xo = 18 x 3.=54 teeth;

t
therefore from Figure 10-22,
= O32.
The overload factor can be obtained from Table 10-4.
The power source (electric motor) is considered a uniform source of power,
the conveyor is assumed to have moderate shock. Therefore K, = 1.25.
The size factor for spur gears is usually taken as unity, that is, K, = 1.
Because no information concerning misalignment is available, Table 10-5
is used to obtain K,, = 1.6.
We are now in a position to calculate the allowable F, from equation (a)
Nts 80032062) 280.32
HT OS se] x ilse 16
522 lb
FV 226942
hp allowable =
33,000 33,000
=15hp @
 564
Chapter 10: Spur Gears

SECTION 10-16

Surface Durability of Spur Gears

Having discussed the first of the two general causes of tooth failure,
breakage based on the strength of the gear tooth, it is now appropriate to
consider in detail the second category, namely, surface destruction. The
following types of surface destruction are usually lumped together under the
general term wear:

1. Abrasive wear. A failure due to the presence of foreign material in the

lubricant that can scratch the tooth surface.
2. Corrosive wear. A failure due to chemical reaction on the tooth surface.
ies). Pitting. A fatigue failure due to repeated application of stress cycles.

4. Scoring. A failure due to metal-to-metal contact attributable to

lubricant failure.

Figure 10-25 The two cylinders shown are kept in rolling contact
by the forces F.
 565
Section 10-16: Surface Durability of Spur Gears

Abrasion, wear, and scoring are clearly failures that are the results of
improper lubrication and gear enclosure. Pitting, on the other hand, is
primarily a fatigue stress failure and is the type of failure the designer protects
against in determining the so-called allowable wear load, to be discussed
shortly. Tests have shown that pitting usually occurs on those portions of a
gear tooth that have relatively little sliding motion compared to rolling
motion. Clearly then spur gears will have pitting occurring near the pitch
line where the motion is almost all of the rolling form.
In designing a gear, the material chosen must have an endurance limit
large enough to resist the repeated dynamic load gear teeth are subjected to.
As will be seen shortly, the basic tooth proportions also are entailed in the
wear load calculation.
In order to obtain a formula for the actual compressive surface stress that
exists between two meshing gear teeth, the Hertz equation for the surface
stress between two cylinders in rolling contact is used.
Two similar equations are commonly used to determine the surface wear
strength of gears. The first of these is the so-called Buckingham equation and
the second is the AGMA formula.
Figure 10-25 shows the two cylinders and the dimensions used to obtain
the Hertz stress equation

oO = ST
i
(l—p?)
ys to
(1-45)
(10-28)

nt
| E, sf E, |

where

o = surface stress, psi

r, = radius of smaller cylinder, in.

r, = radius of larger cylinder, in.

L = length of contact of cylinders

ut = Poisson’s ratio

E = modulus of elasticity

In applying the Hertz equation to spur gears, the contact between the
teeth is taken at the pitch point, for the reasons mentioned previously. The
force F is replaced by the allowable wear load F,,, r, and r, are replaced by
r,, and r,, the radii of curvature of the respective gear teeth at the point of
contact, and oa is replaced by S,, the allowable surface endurance strength.
 566
Chapter 10: Spur Gears

Figure 10-26 The relationship between the radius of curvature

of a gear tooth profile and the pitch circle radius.

As can be seen in Figure 10-26, the radii of curvature are related to the pitch
circle radii by equation (10-29).
r,, =r,sin d
(10-29)
rt, =1, sing
1 1

» eget
s.= | — oe _p: (10-30)
no wie He
E, E,

We can simplify the above equation by considering the term

1 1
bs faba wee! 1 tS) a! (tee)
sing sing sing ee ~ sing Ete
now since the pitch circle diameter d = 2r we have

Cae
1 2 DD 1 | 21d.) |) eid ed)
sing d,d, 5 sing| d,d, ~ sing d,d,
 567
Section 10-16: Surface Durability of Spur Gears

Assuming that the gears are made of materials having the same value for
the Poisson’s ratio (u = 0.3 is a reasonable value for most metallic gears), the
denominator of equation (10-30) can be written as

no(ajayage es He ie
Be Pee = (013)71 |=e
EF E, 1B E,

jae
5 ON 6b pee
E, aE E,

If we perform the substitutions just discussed into equation (10-30) we get

oF (dtd)
“sin @dd,
1 F 1
2.86b
EeDp ee g

This equation is then solved for the allowable wear load F,,
1 1
S2b sin | — +
Fe cee amet a cl ea 12 (10-31)

Table 10-11 Wear Load Factor, K, and Surface Endurance Limit

Surface
Materials in Endurance
Pinion and Gear Limit,
@ = 25 deg.
o= 145,deg. @ = 20 deg.
So, psi

Both gears steel, with average brinell

hardness number of pinion and gear

150 50,000 51
Wis 60,000 V2
200 70,000 98
225) 80,000 127
250 90,000 162
275 100,000 200
300 110,000 242
35 120,000 288
350 130,000 333
S15) 140,000 384
400 150,000

Steel (BHN 150) and cast iron 50,000

Steel (BHN 200) and cast iron 70,000 147
Steel (BHN 250) and cast iron 90,000 242
Steel (BHN 150) and phosphor bronze 59,000 Wil
Steel (BHN 200) and phosphor bronze 65,000 123
Steel (BHN 250) and phosphor bronze 85,000 228
Cast iron and cast iron 90,000 B27
Cast iron and phosphor bronze 83,000
 568
Chapter 10: Spur Gears

Values for the allowable surface endurance strength are given in Table
10-11.
For steel gears, the following formula may be used to obtain approximate
values for S$, when more exact values are not available:
S. = (400)(BHN) — 10,000 psi
where BHN is the Brinell hardness number.
The allowable wear load as obtained by the above equation is clearly a
normal force (see Figures 10-25 and 10-26). Because Buckingham compared
the allowable wear load to the dynamic load (a function of the transmitted
load, F.), some further manipulation of equation (10-31) is necessary. Letting
S? sing
Ke 10-31a)
1.4 5 a | (
and
0 2d, aN,

where N, = number of teeth. The wear load can be written as

F, = 4,bOK (10-32)
Values of S, and K are listed in Table 10-11 for various combinations of
materials for the pinion and the gear and different pressure angles. Se
(10-32) is the Buckingham Equation.
SECTION 10-17
AGMA Wear Equation
An alternate method ofdetermining gear safety as far as wear is concerned
is the AGMA? method. The fundamental wear equation is given by

EOFG Coe
¢=C LO Seer ea
10-33
oPmes Ge dha! ioe)

o. ° = calculated contact stress number

C, = coefficient depending on elastic properties of materials
F, = transmitted tangential load, lb
C, = overload factor
C,, = dynamic factor
d = pinion operating pitch diameter, in.

? Extracted from AGMA information sheet, Surface Durability (Pitting of spur, helical, herring-
bone, and bevel gear teeth), (AGMA 215.01) with the permission ofthe publisher, the American
Gear Manufacturers Association, 1330 Massachusetts Ave., N.W., Washington, D.C. 20005.
 569
Section 10-17: AGMA Wear Equation

b = net face width of the narrowest of the mating gears, in.

C. = size factor
C,, = load distribution factor
I = geometry factor
C, = surface condition factor

The elastic coefficient, as the name implies, is determined by the elastic

properties of the pinion and gear materials. It can be obtained from equation
(10-34).
k
= | Ty Cw
a ee
(10-34)

p g

where yu, and u, = Poisson’s ratio for pinion and gear, respectively, and E,,
and E, = modulus of elasticity for pinion and gear respectively ;k = 1 (for
most spur, helical, and herringbone gears) and k = 3 for most bevel gears.
Table 10-12 lists the values for C, for spur, helical, and herringbone gears.
The overload factor, C,, is meant to adjust for overloads due to the
operating characteristic of both the driving and driven apparatus, as well
as momentary overloads due to momentary conditions of operation such
as starting. The designer should usually draw upon his experience in a
particular field of application in determining proper numerical value for C,.
If no specific information is available, the factors given in Table 10-4 may
be used.
The dynamic factor, C,, is primarily a function of the interaction of the
teeth in mesh. The magnitudes used for the dynamic factor depend upon such
factors as tooth spacing accuracy, profile error, pitch line velocity, angular
speed, the inertia and stiffness of the rotating masses, the transmitted load,

Table 10-12 Elastic Coefficient, C,,

Non-localized Contact

Gear Material and Modulus of Elasticity

Pinion Material

and Modulus of Tin Bronze

Elasticity, E
16X 10°

Cast iron

Aluminum

Tin bronze

SOURCE: AGMA,
Poisson’s ratio = 0.30
NOTE: When more exact values ofE are obtained from roller contact tests, they can be used.
 570
Chapter 1 0: Spur Gears

ar
1.00

0.80

S Q©

C,
factor,
Dynamic
o i°

0.20

Si hae = i —ESE———EEE— SE eS ‘
0 1000 2000 3000 4000 5000 6000 7000 8000
Pitch line velocity, Vor fpm

Figure 10-27 Dynamic factor, C,. (Courtesy American Gear Manufacturers’ Association. }

viscosity of the lubricant, and tooth stiffness. If the dynamic loads can be
determined, either computed or measured, they can be used instead of the
transmitted load ;then there would be no need for the dynamic factor.
The curves shown in Figure 10-27 may be used to obtain reasonable
values for the dynamic factor. The AGMA recommends that the proper
curve to be used, of the four shown in the figure, is to be determined by the
following general considerations:
Curve | is used for: (1) Shaved or ground spur gears, when a relatively
small dynamic load exists. (2) High precision helical gears, also with a
relatively small dynamic load. (3) Accurately generated bevel gears.
Curve 2 is used for (1) Shaved or ground spur gears when a light dynamic
load can develop. (2) High precision helical gears with a light dynamic load.
(3) Large planed spiral bevel gears, again with a light dynamic load.
Curve 3 is used for (1) Commercial helical gears. (2) High precision helical
gears with a moderate dynamic load.
Curve 4 is used for (1) Shaved or ground spur gears when a moderate
dynamic load is expected. (2) Commercial spur gears.
The size factor C, takes into account the effect of gear size, gear-tooth size,
area of tooth-contact pattern, hardness and efficiency of heat treatment. If
gears have been properly proportioned, the proper type of steel chosen, and
effective heat treatment used, the size factor is taken as unity. However,
since fatigue tests indicate that the allowable stress for a given fatigue life
decreases as the size of the gear increases, it may be necessary for the designer
to use a value for C, as high as 1.25.
 7
Section 10-17: AGMA Wear Equation

The AGMA suggests that the effective case depth be used to determine
whether a factor greater than unity should be employed. Figure 10-28 is a
' plot of effective case depth, at pitch line, versus normal diametral pitch. The
curves may be used for spur, helical, and herringbone gears. If the effective
case depth agrees with the value obtained from Figure 10-28, unity is used
for C,. If the effective depth does not agree with values from Figure 10-28, a
value larger than unity should be used. Although there are no established
standard values for C,, a maximum value of 1.25 seems appropriate. In
Figure 10-29 the effective depth of the case hardening can be seen as the dark
strip around the surface of the tooth.
The load distribution factor, C,,, as the name clearly implies, is an attempt
to take into account the nonuniformity of load distribution on the gear tooth.
The AGMA lists the following factors upon which the magnitude of C,,
depends: cutting errors; error in rotating axis in mounting due to bore
tolerances ; internal bearing clearance; paralellism of shafts carrying each
gear (includes runout); tooth, blank, shaft, and housing stiffness; bearing
and Hertz deflection ; and thermal expansion and distortion due to operating
temperatures.

Effective case depth is defined as depth of case

which has a minimum hardness of 50 R,.

Total case depth to core carbon is approximately

1.5 X effective case depth.

S\ eee se
NN mer
pitch
diametral
Normal

Ley
1Z J)
2 Se omon cid
0.010
Minimum effective case depth, in.

Figure 10-28 Depth of effective case at pitch line for spur, helical, and
herringbone gears. The values and ranges shown on the case depth curves are
to be used as guides. For gearing in which maximum performance is required,
detailed studies must be made ofthe application, loading, and manufacturing
procedures to obtain desirable gradients of both hardness and internal stress.
Furthermore, the method of measuring the case as well as the allowable
tolerance in case depth should be a matter of agreement between the customer
and the manufacturer. [Courtesy American Gear Manufacturers’ Association. ]
 D7
Chapter 10: Spur Gears

Figure 10-29 The effective depth of the case hardening of the gear tooth shown
can be determined by the dark strip around the surface of the tooth. [Courtesy
Arrow Gear Company. ]

Because of the difficulty of evaluating the stiffness of gear mounting, most

manufacturers run deflection tests on prototypes before producing the gear
systems. This is especially true in the aircraft and automotive industries. The
tests also allow the tooth contact under service loads to be studied.
Figure 10-30 shows curves that illustrate the effect of different rates of
misalignment. In the figure, b,, represents the face width having just 100%
contact for a given tangential load and alignment error.
For wide-face gears, Table 10-13 gives reasonable values for C,,,.
Figure 10-31 shows values for C,, that can be used for spur and helical
gears similar in quality to those used in commercial gear units. When the
b/d ratio is greater than 2, a more detailed analysis is suggested.
Since the gear hardening process results in distortion, it is suggested that,
for gears which have been hardened without having a finishing operation (as
for example, grinding), the C,, factor obtained from Figure 10-31 should be
multiplied by 1.05 if one element is hardened and by 1.10 if both elements are
hardened after cutting.
The geometry factor I, is a function of the following factors: pressure
angle, gear ratio, load sharing ratio, length of the line of contact, base pitch,
 73
Section 10-17: AGMA Wear Equation

and length of action. Figure 10-32 presents values of the geometry factor for
standard spur gear systems. For gears having pressure angles other than
those shown in Figure 10-32, the AGMA in Appendix B of the information
sheet Surface Durability (see footnote 2), presents a procedure that can be
used to determine reasonable values of the geometry factor.

W, = tangential load, Ib
e = alignment error, in./in.

5.0

4.0 ess

c& 3.0 eee

Force
C,
coefficient,
error
—

2.0 =|

1.0
0 1.0 2.0

Om
b
b = face width
(inches)

10.30

Figure 10-30 Spur gear load distribution factor, C,,. [Courtesy American Gear
Manufacturers’ Association. ]
 574
Chapter 10: Spur Gears

Table 10-13 Load Distribution Factor for Spur, Helical, and Herringbone
Gears, C,,

Ratio of 8 Contact Cm

95% face width contact obtained at one-third torque 1.4 at 7 torque

95% face width contact obtained at full torque 1.1 at full torque

75% face width contact obtained at one-third torque 1.8 at 4 torque

95% face width contact obtained at full torque 1,3 at full torque

: ]
35% face width contact obtained at one-third torque PE Seat 3 torque
1.0 or less | 95% face width contact obtained at full torque 1.9 at full torque

20% face width contact obtained at one-third torque 4.0 at - torque

75% face width contact obtained at full torque 2.5 at full torque

Teeth are crowned

2.5 at Ftorque
35% face width contact obtained at one-third torque
85% face width contact obtained at full torque 1.7 at full torque

Calculated combined twist and bending of pinion not over

0.001” over entire face
Pinion not over 250 BHN hardness
75% contact obtained at one-third torque 2.0 at 3 torque
95% contact obtained at full torque 1.4 at full torque

Calculated combined twist and bending of pinion not over

0.0007” over entire face
Pinion not over 350 BHN hardness
Over | 75% contact obtained at one-third torque 2.0 at 3 torque
but 95% contact obtained at full torque 1.4 at full torque
less
than 2
30% contact obtained at one-third torque 4.0 at 4torque
75% contact obtained at full torque 3.0 at full torque

Calculate effects of
deflection and either
adjust helix angle to
Twist and bending exceeds 0.001” over entire face
compensate for deflection
or increase C,,, to allow
for both alignment errors
and deflection

SOURCE: AGMA

The surface condition factor C, is an attempt to take into account such

considerations as surface finish, residual stress, and plasticity effects. C, is
usually taken as unity when the surfaces have a good finish, resulting either
from a finishing operation or the running-in process. When rough finishes
are present or when the possibility of high residual stresses exists, a value of
1.25 is reasonable. If both rough finish and residual stresses exist, 1.5 is the
suggested value.
 D6)
Section 10-17: AGMA Wear Equation

b
Cr = 0.456 + 2.0
am Tr ii
For fourth and additional reductions
use third reduction values

b
om = 9596 +23
b
Cm = 0.555 +25

Use above
C,,
factor,
distribution
Load formulas when
face widths
exceed 18 in.

2 3 4 5 6 i 8 9 10 11 12 13: 14 (6) 16 if 48

Face width, in.

Figure 10-31 Spur and helical gear load distribution factor, C,,. [Courtesy
American Gear Manufacturers’ Association. ]

Thus, when reasonable values of the factors discussed are chosen, based
either on the values suggested by the AGMA or upon personal experience
or tests run on prototypes, the designer can calculate the contact stress
number from equation (10-33). Clearly, this calculated or actual stress number
must be compared to an allowable value in order to determine whether or
not the gears will be satisfactory as far as wear is concerned.
The AGMA specifies that the calculated stress number must be less than
or equal to an allowable contact stress number, which has been modified
by several correction factors. In equation form, the relationship that must be
satisfied is as follows:
@
o.< 5. EACH (10-35)

where
Sac = allowable contact stress number
Gy = lie factor
Cy = hardness ratio factor
C, = temperature factor
C,=lactor of safety

We will present a brief discussion of the factors indicated above. It should

be noted that both the pinion and gear must be checked because of the
difference in material properties and number of tooth contact cycles under
load.
 576
Chapter 10: Spur Gears

Gear ratio
(a)

/Geometry
factor,
Geometry
/
factor,

/
factor,
Geometry

Gear ratio
(c)

Figure 10-32 External spur pinion geometry factor, I, for

standard center distances. All curves are for the lowest point of
single tooth contact on the pinion. (a) 143 deg pressure angle full
depth teeth, standard addendum = 1/P. (b) 20 deg pressure angle
full depth teeth, standard addendum = 1/P. (c) 20 deg pressure
angle stub teeth, standard addendum = 0.8/P. [Courtesy American
Gear Manufacturers’ Association. ]

The allowable contact stress number S,.is a function ofsuch factors as the
material of the pinion and gear, the number of cycles of load application, the
size of the gears, the temperature, the type ofheat treatment or work hardening
to which the gears have been subjected, and the presence of residual stress.
The AGMA publishes many standards that give values for the contact
stress number. If the particular application the designer is working on does
not fall into one of the categories mentioned, Table 10-14 may be used to
obtain reasonable values for S,.. As can be seen, ranges ofvalues are given in
 S77
Section 10-17: AGMA Wear Equation

Table 10-14 Allowable Contact Stress Number, Se

Poe ealiraaey Surface Hardness,

Material Material =a

Through hardened Cast iron

AGMA grade 20 a 50-60,000
Ue Bhn 85-95 ,000 AGMA grade 30 175 Bhn 65-75,000
240 Bhn 105-115,000 ! AGMA gerade 40 200 Bhn 75-85 ,000
300 Bhn 120-135,000
360 Bhn 145-160,000 |} Nodular iron 90-100% of
44 0 Bhn 170-190,000]]
-19 sated vere the Sye_
value of
Normalized 210 Bhn steel with
Steel the same
Case carburized Oil quench and temper 255 Bhn hardness
55 R, 180-200,000 Tensile Strength
60 RK, 200-225 ,000 Bronze ae
psi (min)

Tin bronze
AGMA 2C (10- 40,000
Flame or induction 12% Tin)
hardened Aluminum bronze
ASTM B 148-52 90,000
SOR, 170-190,000 |} (Alloy 9C-H.T.)
SOURCE: AGMA.

the table. The upper range values should be used for problems where high
quality material is used, there is reason to expect maximum response to heat
treatment, and when adequate inspection is available to ensure proper
quality control. The lower range values are to be used for general design
purposes.
The life factor C; accounts for the expected life of the gear. In other words,
the fewer the number ofload cycles during the lifetime of the gear, the greater
can be the allowable contact stress number. While, at present, there is
insufficient information available to plot accurate fatigue curves, Figure
10-33 can be used to obtain reasonable values for C,. As can be seen from

5.0

4.0

3.0

2.0

factor,
Life
C,

104 10° 106 107 108 102

Figure 10-33 Life factor. [Courtesy American Gear Manufacturers’

Association. ]
 578
Chapter 10: Spur Gears

Table 10-15 Typical Gear and Pinion

Hardness Combinations

Gear BHN Pinion BHN

En ta RT RRR a A A

180 210

AIO) 245

ES 265

245 285

255 300
270 BS

285 335

300 350

SOURCE: AGMA,

the curve, for a required life of 10 x 10° cycles or more, the factor is taken as
unity, but a factor as large as 1.5 can be used if the required life is 10,000
cycles.
The hardness ratio factor C,,, although clearly a function of the hardness
of the two gears in mesh, is also dependent upon the gear ratio. Table 10-15
shows some typical gear pinion hardness combinations that have worked
well in applications. Figure 10-34 can be used to obtain reasonable values
[Om C 5.
Because the allowable stress of steel varies with temperature, a temperature
modifying factor C; is desirable. A value ofunity is usually used for C; when
the lubricant or gear blank temperature does not exceed 250°F. Values
greater than unity may be required for carburized gears operating at lubricant
temperatures above 180°F. Until more data 1s available, it is suggested that
the empirical equation for the strength temperature factor, K ;, be used when
values for C; greater than unity seem to be indicated. In other words, for
this situation, equation (10-36) may be used.
460
+ T,
ee S
(10-36)

where 7; = the peak operating oil temperature in degrees Fahrenheit.

The factor of safety Cp enables the designer to design for a calculated risk
or to design for high safety, which means high reliability. Table 10-16 lists
some suggested values for Cr. It should be noted that the values shown are
much smaller than those customarily used in machine design. These smaller
values are suggested because they are applied to the fatigue strength of the
material rather than its tensile strength, and if failure occurs it is not an
abrupt failure under the applied load but rather a shortening ofthe expected
minimum life.
 De
Section 10-17: AGMA Wear Equation

Brinell of pinion
K = —________ when K < 1.2 use C,, = 1.00
Brinell of gear

Hardness,
C,,
factor,
ratio

Single reduction gear ratio

Figure 10-34 Hardness, ratio factor, C,. [Courtesy American

Gear Manufacturers’ Association. ]

To sum up the discussion ofthe wear strength ofgears, a gear is considered

safe in wear if equation (10-35) is satisfied. Namely, if,

Feb
CrCp
where o, is determined by equation (10-33)

= @ TFto
Cae
~s Cue
wee Be

aR emNeGiecdby 1
 580
Chapter 10: Spur Gears

Table 10-16 Factor of Safety, C,

Requirements of Application Cr

High reliability 1.25 or higher

Fewer than one failure in 100 1.00

Fewer then one failure in three 0.80*

SOURCE: AGMA,
“At this value plastic profile deformation might occur before pitting.

A designer very often is interested in calculation of themaximum allowable

horsepower a gear set can transmit as based upon the surface durability of
the gears. Equation (10-37) may be used to determine this power in one step.

P,. =
nb Le te ee 2
(10-37)
#10126,000, CONCCo NCARGrae
where P,, = allowable power, in horsepower and n, = pinion speed, in
revolutions per minute.

Example 10-7. Check to see if Example 10-6 is safe in wear according to the AGMA
method.
Solution: Using equation (10-33)

From Table 10-12: C, = 2300; F, = 522 lb from Example 10-6; C, = 1.25

(same value as K, in Example 10-6); and V, = 942 ft/min from Example 10-6.
For curve4 from Figure 10-27: Cy = 0.62; C, = 1;d, = 3 in. from Example
10-6; b = 2 in. from Example 10-6.

2D)
7 aeais
From Figure 10-31: C,, = 1.3; C; = 1. And from Figure 10-32b, J = 0.1.
Thus

AR 500105
‘| at ia
oe ne. ae ae
= 2300,/2290 = 2300 x 47.8 = 110,000

Then from equation (10-35)

S
{EF
GG
LCH
 581
Section 10-18: Gear Design

From Table 10-14 use lowest value for 180 BHN to approximate 140 BHN
so that S,. = 85,000; C, = 1 from Figure 10-33; C, = 1 from Figure 10-34;
Cy = 1 and Cp = 1.15 from Table 10-16. Therefore
C.Cy il x< i
a facd 85,000 x ——— +; = 74,000
Cr. Cr
Oo.
>S ac a

and therefore, the gears are not safe as far as wear is concerned. @

SECTION 10-18
Gear Design
The first step in a discussion of spur gear design should be the selection of
the type of gear to be used. When the design problem involves nonparallel
shafts, one of the gear types to be discussed in Chapter 11 should be used.
For parallel shaft applications, either spur or helical gears may be used.
Since helical gears are usually more expensive than spur gears, the designer
is advised to use spur gears wherever feasible. A reasonable rule to follow is
to use spur gears for low speed applications and helical gears for high speed
applications. Clearly then, the first design step is to choose the proper type
of gear.
A reasonable criterion is to consider the application to be a high speed
one if the pitch line velocity exceeds 5000 ft/min or the rotational speed of
the pinion exceeds 3600 rpm. It should be pointed out that this criterion is
perfectly arbitrary, and spur gears may be used for applications exceeding
the limiting speed. This is particularly true if noise limitation is not important,
if end thrust load is to be avoided, or if the spur gears can be manufactured
with great precision.
At this point, the design procedure becomes one of trial and error because
the size, tooth form, and dimensions of the gear must be known before the
actual loads and stresses can be determined. Clearly then the simplest type
of analysis should be used to determine preliminary values, and then the
more accurate procedures used to finalize the design.
There are many possible design procedures that will give a satisfactory
result. It should be obvious that the more experienced a designer is, the
easier it will be for him to assume values which will work well. The following
procedure usually gives good results:
Step 1. The required horsepower, revolutions per minute, velocity ratio,
and center distance are used to obtain values for the gear diameters, tangential
force, pitch line velocity, and the class of gear. In many problems the center
distance and velocity ratio are either not specified or are not critical. In this
situation the designer must assume reasonable values for these quantities.
Step 2. The dynamic effect on gear teeth, due to such factors as in-
accurately cut teeth, improper tooth spacing, improper mounting, deflections
 582
Chapter 10: Spur Gears

due to elasticity of the gears and shaft, and irregular load requirements must
be taken in account in this preliminary design stage. The simplest method by
which to accomplish this is to use the appropriate equation of the following
three to calculate the so-called dynamic load.

Fe
_ 600 + V, Pe 10-38)
ari & ese
for 0 < V, < 2000 ft/min
= 1200 + oF (10-39)
1200
for 2000 < V, < 4000 ft/min
Ve
a= ai: (10-40)

for V, > 4000 ft/min where F,, > F, and F, > F;

Step 3. Choose values for pressure angle, tooth type, and gear material.
In choosing the material to be used, it should be kept in mind that softer
materials are usually less expensive.
Step 4. Since the wear load does not depend on the diametral pitch, the
dynamic load can be substituted into equation (10-32) and the required gear
width determined. Generally, the width should be greater than 9/P and less
than 13/P. These limits are used because a small width tends to cause align-
ment problems, whereas too large a width can produce twisting which results
in nonuniform load distribution.
Step 5. The diametral pitch is determined by substituting the dynamic load
into the simple Lewis equation (10-22). The Lewis form factor Y must be
estimated, and after determining P the number of teeth are calculated, which
enables Y to be obtained from the table. If there is a difference between the
assumed and calculated values of P, the calculation must be done again.
Clearly standard values of diametral pitch should be used. However, most
gear manufacturers have their own set of standard pitches, and these are not
necessarily all in agreement. Typical standard pitches would be 2, 23, 3, 4, 5,
6; 83,10) IDR 165205 245532548:
The minimum number of teeth for standard gears is usually taken as 18,
although for light loads as few as 14 may be permitted. Gears with fewer than
14 teeth can be used, but they must be specially modified.
Step 6. The gear dimensions have now been determined and the usual
procedure is to recheck the design using the AGMA method. In other words
by using equations (10-25), (10-26), (10-33), and (10-35).
It should be remembered that both the pinion and the gear must be checked.
Many designers have found through experience, that certain modifications
in this procedure are desirable. For example, designers often make a gear
 583
Section 10-18: Gear Design

weaker in wear than in bending because an impending wear failure gives

visible evidence in the form of pitting, whereas the first sign of a bending
failure is usually a broken tooth.
Finally, it should be stated once again that the procedure outlined above
is but one of many that can be and are used by designers.

Example 10-8. A pair of gears is to be designed to transmit 40 hp for a pinion speed

of 1000 rpm and a velocity ratio of 1 to 5.
Solution: Since the rotational speed of the pinion is less than 3600 rpm,
spur gears will be chosen for this problem.
No condition regarding center distance is given, so it will be assumed that
the center distance should be kept as small as possible in order to conserve
space. Let the pinion have the minimum number of teeth, N,, = 15. Then
the gear will have N,, = 75 teeth. Assume the teeth are to be 20-deg full depth.
The Lewis form factors from Table 10-2
Y,= 0.289 Y, = 0.433
We will next choose the gear materials so that their strengths are roughly
the same. Let us try as the pinion material gray cast iron ASTM3S, S, = 12,000,
BHN = 212 (Table 10-3) and for the gear gray cast iron ASTM25, S, = 8000,
BHN = 174.

S,Y for pinion = 12,000 x 0.289 = 3468 psi

S,Y for gear = 8000 x 0.433 = 3464 psi
Therefore both gears are of approximately the same strength.
Because
_ Tn
hp ~ 63,000
_ 40x 63,000
= 2520 in.-lb
~ 1000

The diametral pitch should be as large as possible in order to obtain the

best economy. Assume P = 5. Then
N, :
d, = cr = 3 1n.

md,n mt. x 3 x 1000 ;

& = 1D = D =] 85 ft/min

if 2520
imap
=—_=
e = 1800 lb

600 + V, 600 + 785

is
: 600 F; 600 Md
= 4160 lb
F, = d,bQK
De 7 US 150 | 167
157550
 584
Chapter 10: Spur Gears

From Table 10-11, K = 264

a F, 4160
“GOK 35 kOT x 264
== SLSR Shit
Now
9 9 ies NG:
—=-=1., — =— = 2.6
jo se JY 5
Because 3 > 2.6, P = Sis no good. Try P = 4
N, 15
dp = pipe aie

tdn mx 3.75 x 1000 5

= p = D = 982 ft/min

2520
F, =. 37572 13454 lb

600 + 982
F, = ——— 1345 = 3550 Ib
600
3550
=) loin, use 2 In.
~ 3.75 x 1.67 x 264
Now 9/P = 9/4 = 24: while 2 < 24, it is close enough to be acceptable.
je Sb a 12,000 S x 0.289 = 1734 1b

Therefore strength is unacceptable because 1734 < 3550. Either the material
can be rechosen or face width increased
_F,P_ 3550 x 4
Oe SY — 12,000 x 0.289

But 13/P = 13/4 = 3.25 is maximum; therefore, increasing face width is not
enough. Therefore, we must also choose new material for the pinion. Assume
b = 3; in. and ASTM S0 gray cast iron is used, S$, = 15,000 psi. Now
VV si 15,000 x 3%i x 0.289
ae Sb5 :
4
= 3522 ~ 3550 is satisfactory
The BHN of ASTM 50 is greater than that of ASTM35 cast iron and 33 is
greater than 3 in. width. Therefore the pinion is also safe in wear.
Check of gear is also necessary.
S,Y for pinion now = 15,000 x 0.289 = 4335
S,Y for gear = 8000 x 0.433 = 3464
Y 8000 x 33 x 0.433
Ee = Sob = 4

= 2820 < 3550 is unsatisfactory

 585
Section 10-18: Gear Design

Therefore choose gear material ASTM35 S, = 12,000

12,000 x 35 x 0.433
bgear = 4
= 4230 > 3550 is satisfactory

Final check of gear design using AGMA method for pinion.

peakK
K7Kr

ASTM 50 from Table 10-7 gives by extrapolation S,, = 17,500 psi. K, = 1

fOmIndelnites itcew heal Ke male aalinen
17,500 x 1
Sa = = 14,583 psi
1x 1.2
F.K,PK.Kp
Kees
On=

V,, = 982 ft/min. From Figure 10-21 curve 3, K, = 0.62. From Figure 10-22
N= 15, Nh = TS, d= OMS, Iai IO) mes iS, = LYS), JX = Il, ariel
Keele from Table 10-5. Then
Svs Se ilwey Se hse il Se LG
CG,=
Kye Bes) x YS
= 22,800 psi
22,800 > 14,583; is not safe
From Table 10-3, SAE 1040 has BHN 202; from Table 10-7 for BHN 202;
interpolation gives S,, = 27,000 psi.
27,000 x 1
Sea = a5 = 22,500
22,500 is close enough to 22,800 to be satisfactory.
Check of gear strength
SaKy
Sad ai
K;Kp
ASTM 35 (same as AGMA 35) has BHN = 212 from Table 10-3. Table 10-7
gives by interpolation S,, = 10,750 psi. Therefore

Se ee = 8,958 psi

V, = 982 ft/min; K, = 0.62; N,, = 75; N,, = 15; J =042; K, = 1.25;

I 3 GIS, = We
13455— 1255x417
eee (0.60 Si042
= 13,000 psi
Since S,4 < d,, gear is unsafe.
Try ASTM SO, S,, = 17,500 psi, then S,, = 14,583 > 13,000, gear OK.
Wear check by AGMA _ formula. The pinion: SAE 1040 BHN = 202; gear:
ASTM 50 BHN = 223.

OMEN,
sie on C, db 1
 586
Chapter 10: Spur Gears

For the pinion from Table 10-12, c, = 2000; F, = 13451b; C, = 1.25 (same
as K,); V, = 982 ft/min. From Figure 10-27 and cunve 4) G1 — 0:62. Ga—ai-
d, = 3.753in.; b = 34 in. From Figure 10-31, C,, = 1.33; C, = 1. From Figure
10-32b I = 0.103.

(345 0105 i 1335e1

ee 200| O6l* Sa58es05 0104
= 107,000.
Evaluate

CiCy
s.|
aes

From Table 10-14, by interpolation S,. = 98,000. From Figure 10-33, C, = 1.

From Figure 10-34, K = BHN,/BHN, = 202/212 < 1.2. Therefore Cy = 1
and C; = 1. From Table 10-16, Cp, = 1.15.
ilSe I
98 00s )=' 85,217
AS
Since 85,217 < 107,000, the pinion is unsafe in wear. Try a BHN of 300, that
has an S,. = 135,000. Then allowable is equal to
Sorell
135500( 8 )= 117,400.
iLaIlss
The pinion is now safe in wear since
117,400 > 107,000.
It is clear by inspection that the gear will also be safein wear. @

In conclusion, it should be emphasized once again that the above design

is by no means the only solution to the problem. In fact, the designer might
try several other approaches, giving different gears, and then pick the best of
these solutions.

SECTION 10-19
Backlash Control
As was stated in Section 10-2, a certain amount of backlash is necessary
when two gears are in mesh. In other words, if the space between teeth were
not made larger than the tooth width as measured on the pitch circle, the
gears could not mesh without jamming. However, any amount of backlash
greater than the minimum amount necessary to ensure satisfactory meshing
of gears can result in instability in dynamic situations and position errors in
gear trains. In fact, there are many applications such as instrument differential
gear trains and servomechanisms that require the complete elimination of
backlash in order to function properly. Figure 10-6 is reproduced here to
show graphically again the angular and linear backlash present in two
meshing gears.
 587
Section 10-19: Backlash Control
ase circle =r
B Pinion
Line of O
action
Linear backlash =B=7,—T,

Angular backlash of
Gear = 0, =B/R
Pinion = 05 = B/r
where 7, = tooth space
Tp = tooth width
of pinion
Base
circle Tg = tooth width
of gear,
all measured on pitch circle

(a) Angular backlash

Backlash, along line-

of-action = B, , = B cos @

Base
circle

Line of
action

Base circle
o'

(b) Linear backlash

Figure 10-6 Backlash, the amount by which the width of a tooth space
exceeds the thickness of the engaging tooth on the pitch circles is shown.
Backlash can be designated as (a) angular, or (b) linear. [Courtesy
Designatronics, Mineola, N.Y.]

Many schemes have been propounded to control the amount of backlash.

It is the purpose of this section to explore a few of the more commonly used
procedures. Backlash is generally considered to result from two general
areas: those due to (1) fixed magnitude sources such as gear tooth width
compared to mating gear tooth space, errors in actual center-to-center
distance compared to ideal distances, and radial clearance in gear shaft
bearings, and (2) those associated with gear rotation such as composite error,
rolling bearing element eccentricity, and eccentricity resulting from clearance
between gear bore and mounting shaft diameter.
One method used to eliminate much of the unnecessary backlash is the use
of adjustable center gears. Clearly this procedure cannot eliminate the
backlash due to gear rotation because the gears must be free to mesh before
rotation occurs. The method does have a number of other disadvantages,
 588
Chapter 10: Spur Gears

Figure 10-35 Typical spring loaded scissor gear. [Courtesy

Designatronics, Mineola, N.Y.]

Gear half fixed to shaft

namely, much skill is required to set up the gears, proper adjustment must
be maintained while the gears are in motion, and the provision for adjustable
centers results in increased costs.
Another procedure commonly used to eliminate backlash is the spring-
loaded scissor gear, shown in Figure 10-35. As can be seen in the figure, the
gear is made of two halves, one of which is fixed to the shaft and connected
to the free half by means of springs. In operation, these gear halves are forced
to move relative to each other by the spring force until the tooth space of
the mating gear is completely filled. The advantages of this procedure are
simplicity, standardization, interchangeability, and reliability. The dis-
advantage is that the springs must be given a torque larger than the load to
be driven, or else the springs will deflect, resulting in lost motion. As a result,
spring loaded gears are used for light duty torque applications. Torque values
would typically be measured in inch-ounces (in.-oz.).
For high torque applications it is possible to use fixed or locked spring-
loaded gears. In this procedure, the split gears are assembled and the halves
of each gear are permanently joined by means of rivets or screws. Thus, all
of the backlash except that due to rotation is eliminated. Clearly the trans-
mitted torque capacity is limited by the strength of the fastening element.
The primary disadvantage of this procedure is the fact that it is no longer
possible to interchange gears without removing the permanent connection
(rivets or screws).
Many schemes of a similar nature to those we have discussed are available.
However, generally speaking, some form of spring-loaded gear is probably
the best solution to the problem of backlash control and elimination.

SECTION 10-20
Gear Lubrication
With the advent of high load, high speed gears, the problem of proper
lubrication of such gear trains became extremely important. In general, the
purpose of the lubricant was to separate the tooth surfaces and remove heat
 589
Section 10-21: Gear Materials

from the gears. As can be demonstrated by a kinematic analysis of two

meshing gears, the relative motion between two teeth in contact is a combina-
tion of rolling and sliding motion. This combination of motions increases the
lubrication problem.
Splash lubrication and pump fed lubrication are the two methods by which
oil is usually applied to gear systems. Which of the two systems is used is
usually determined by the pitch-line velocity. For low velocities, under
3000 ft/min, the splash system is used, whereas for velocities above 5000 ft/min
pump fed lubrication is indicated.
A brief discussion of what type of lubricant should be used in gear
applications is presented in the following extract from reference [5].*
For many enclosed gear drives operating under light loads and normal
speeds, straight mineral oils provide satisfactory lubrication. Small gears
operating at speeds below 1200 rpm and lightly loaded should be lubricated
with an oil having a viscosity of 700 to 1000 SUS at 100°F. Large gears
operating at speeds up to 2000rpm and normally loaded should be
lubricated with a gear oil having a viscosity of 80 to 105SUS at 210°F.
This gear oil should be the mild extreme pressure type. Gears operating
at low speeds and heavily loaded should be lubricated with an E.P.
inactive-type multi-purpose gear oil of SAE 90 grade at normal ambient
temperatures and SAE 140 grade at high ambient temperatures.
Applications such as aircraft turbine engines, which are required to
operate under extreme temperature conditions (—60°F to +500°F), are
generally lubricated with synthetic oils. It is obvious that petroleum base
products could not operate under these extremes of temperature. Synthetic
lubricants are either chemical compounds or mixtures of compounds of
of the same chemical class produced to meet operating requirements
beyond the normal capabilities of petroleum lubricants. Among the most
important of these requirements are high viscosity index, high temperature
stability, low oil volatility, and foam resistance. Synthetic oils must pass
engine qualification tests to insure satisfactory performance. The military
is becoming concerned with the synthetic lubricants and their load carrying
ability. Currently investigations are under way to provide this type
lubricant with increased film strength and to give greater protection of
the gears under higher horsepower loadings encountered in high speed
aircraft.

An interesting new development in gear lubrication is the so-called self-

lubricating gear. It is a laminated gear made up ofalternating layers of metal
and dry lubricant. After the sandwich has been bonded it is machined to form
a standard gear. During rotation, the lubricant moves across the entire
surface of the meshing gear.

SECTION 10-21
Gear Materials
Gears are manufactured from a wide variety of materials, both metallic as
well as nonmetallic. This section will be devoted to a discussion of metallic
materials, and non-metallic materials will be considered in Section 10-22.
3 Courtesy of Gleason Works, Rochester, N.Y.
 590
Chapter 10: Spur Gears

As is the case with all materials used in design, the material chosen for a
particular gear should be the cheapest available that will ensure satisfactory
performance. Before a choice is made, the designer must decide which of
several criteria is most important to the problem at hand. If high strength is
the prime consideration, a steel should usually be chosen rather than cast
iron. If wear resistance is the most important consideration, a nonferrous
material is preferable to a ferrous one. As still another example of how a
choice can be made, for problems involving noise reduction, nonmetallic
materials perform better than metallic ones. However, as is true in most
design problems, the final choice of a material is usually a compromise. In
other words, the material chosen will conform reasonably well to all the
requirements mentioned previously, although it will not necessarily be the
best in any one area. To conclude this discussion we will consider the
characteristics of various metallic gear materials according to their general
classifications.

Cast irons
Cast iron is one of the most commonly used gear materials. Its low cost,
ease of casting, good machinability, high wear resistance, and good noise
abatement property make it a logical choice. The primary disadvantage of
cast iron as a gear material is its low tensile strength, which makes the gear
tooth weak in bending and necessitates rather large teeth. ASTM numbered
cast irons between 20 and 60 are commonly used for gears. It should be
pointed out that the corresponding AGMA numbered cast irons have the
same tensile strength as the ASTM ones.
Another type of cast iron is nodular iron, which is made of cast iron to
which a material such as magnesium or cerium has been added. The result
of this alloying is a material having a much higher tensile strength while
retaining the good wear and machining characteristics of ordinary cast iron.
Very often the combination of cast iron gear and a steel pinion will give a
well balanced design with regard to cost, strength, and wear.

Steels
Steel gears are usually made of plain carbon steels or alloy steels. They
have the advantage, over cast iron, of higher strength without undue increase
in cost. However, they usually require heat treatment to produce a surface
hard enough to give satisfactory resistance to wear. Unfortunately, the heat
treatment process usually produces distortion of the gear, with the result that
the gear load is not uniformly distributed across the gear tooth face. Since
alloy steels are subject to less distortion due to heat treatment than carbon
steels, they are often chosen in preference to the carbon steels.
Although it is not the purpose here to discuss the various heat treatment
methods and their effect on the properties of materials so treated, the designer
should be aware of possible problems arising from the use of heat treated
materials.
 59]
Section 10-22: Nonmetallic Gears

Gears are often through-hardened by water or oil quenching in order to

increase their resistance to wear. If a low degree of hardness is satisfactory,
through-hardening is probably the most desirable heat treatment process to
be used because of its inexpensiveness.
Case hardening is used for gears that require a hard surface and for which
good accuracy is not necessary. The case hardening process results in gears
that have much harder surfaces than cores. The advantage of case hardening
is that, while the surface becomes hard and wear resistant, the toughness
associated with the core remains. The effect of case hardening a gear tooth
was seen in Figure 10-29. The shaded strip along the surface of the tooth is
the hardened case, while the major portion ofthe tooth (lighter area) remains
unaffected.
Carburizing, cyaniding, nitriding, flame hardening, and induction harden-
ing are some of the processes commonly used to produce the case hardening
effect. If great accuracy is required, the gear must be ground. For problems
in which corrosion resistant surfaces are desirable, nitriding is the process
most often used. However, nitriding is a relatively expensive process and is
used only if other processes do not produce satisfactory results. Flame
hardening and induction hardening are the methods usually used to harden
larger gears.
To sum up this discussion of specially heat treated steels, the designer
should make use of carbon or alloy steels without special heat treatment if
the relatively low surface hardening of these untreated materials is satisfactory.
As the hardness requirement increases, the heat treated steels should be
considered, with the least expensive steel that satisfies the strength and wear
requirements being chosen.

Nonferrous metals
Copper, zinc, aluminum, and titanium are materials used to obtain alloys
that are useful gear materials. The copper alloys, known as bronzes, are
perhaps the most widely used. They are useful in problems where corrosion
resistance is important and also where large sliding velocities exist. Because
of their ability to reduce friction and wear, they are usually used as the
material for making the worm wheel in a worm gear set. Aluminum and zinc
alloys are used to manufacture gears by the diecasting process.

SECTION 10-22
Nonmetallic Gears
Gears have been manufactured of nonmetallic materials for many years.
Rawhide, nylon, various types of plastics, and so on, have been used. The
advantages obtained by using these materials are quiet operation, internal
lubrication, dampening of shock and vibration, and manufacturing economy.
Their primary disadvantages are lower load carrying capacity and low heat
 592
Chapter 10: Spur Gears

conductivity, which results in heat distortion of the teeth and may result in
a serious weakening of the gear teeth.
Recently thermoplastic resins, with glass-fiber reinforcement and a
lubricant as additives, have been used as gear materials. The composite
material has resulted in greater load carrying capacity, a reduced thermal
expansion, greater wear resistance and fatigue endurance. However, a major
problem facing the designer is that gears made of these plastic materials
exhibit a variation in properties that does not depend only on the testing
method used. It is therefore necessary to test each design to determine whether
its performance conforms to the values of material properties used.
In reference [6] a table. containing the properties of reinforced thermo-
plastics is presented. The base plastics, such as styrene acrylonitrite, poly-
carbonate, polysulfane, acetal, polypropylene, nylon, polyurethane, and
polyester, had 30% glass and 15% lubricant, both percentages by weight,
as additives. The properties listed were obtained by testing gears that were
injection molded, 20-pitch, 20-deg pressure angle, full-depth, 23-in. pitch,
diameter, 5-in. face width spur gears.
To sum up this discussion, thermoplastics are desirable engineering
materials because of their versatility. However, because of the difficulty in
predicting their performance, it is usually necessary to test the performance
of the gear on a prototype.

SECTION 10-23
Gear Blank Design
The gear blanks from which gears are manufactured are usually produced
by casting or machining from a solid blank. As was mentioned earlier in this
chapter, there are other processes used, but these two are the most prevalent.
Gears are cut on blanks that are either solid or webbed. Figure 10-36 shows
three types of gear construction, single web, double web, and spoked. When
gear pitch diameters are relatively large, the gear is made in two halves.
Figure 10-37 shows a typical split gear. The decision as to whether to use a
solid or webbed blank depends on the tooth size (pitch) and the pitch circle
diameter. As a guide, maximum solid diameters are
For Pi 3rd '='7 ins for P =420 =164n,
For P = 5,d = 5in.; for P= 6-10, d= 4 in.
For P = 12-20,d = 3 in.

Larger gears are made with arms, hubs, and rims. In other words, very
much the same as flywheels, which are discussed in Chapter 18. The stress
analysis of the rim and arms, is done exactly as shown for the flywheel. In
general, the diameter of the hub is made twice the shaft diameter. The length
of the hub is usually determined by the length of key required. However, the
length should not be made less than the width of the gear. On the other hand,
Figure 10-36 Typical construction showing oval arm (spoke). H type (double
web) and single web gears. [Courtesy Horseburgh and Scott.]

Figure 10-37 Typical split gears showing tongue and groove and pin to insure
accurate and positive reassembly. [Courtesy Horseburgh and Scott.]

593
 594
Chapter 10: Spur Gears

the hub length should not be made excessively large. The usual range for
hub length is 1.5—2 times the shaft diameter. If the hub length exceeds the
upper value, because of the length of key required, two keys should be used.

SECTION 10-24
Gear Trains
Up to this point in our discussion of gears, we have concerned ourselves
with no more than a pair of gears in mesh. Clearly, many applications exist
where many pairs of gears are in mesh. We generally term such a system, a
gear train. The important kinematic relationships that exist for gear trains
may be found in reference [1].
The efficiency of gear trains depends upon the power loss at each tooth
mesh. The amount of loss varies with the tooth characteristics, such as tooth
system, gear material, finish of tooth surface, lubrication, and pitch line
velocity. In general, each mesh should have no more than a 2% loss.
Usually this power loss does not present an overheating problem. However,
as a rough guide, gears having a pitch line velocity exceeding 2000 ft/min
should be checked to see ifenough heat dissipation is possible without external
cooling. Worm gearsets usually present the biggest overheating problems.
In the section on worm gearsets in Chapter 11, a discussion on heat dissipation
is presented. The general procedure shown can be used for spur gear trains as
well as others. The AGMA also suggests equations that can be used to
determine thermal capacity for gear trains.
Excessive noise is another problem associated with gear trains. Although
it is impossible to produce gear trains that are completely free of noise, it is
possible to keep the noise level within acceptable bounds. Noise in gear train
systems is due to several causes, such as vibrations, both lateral and torsional,
bearing noise, inaccurate gear action, and possible noise resulting from
lubricant circulation. Clearly the noise due to inaccurate gear action can be
reduced by improving the precision of the gear teeth. The noise due to
vibrations can be reduced by considering rigidity in the design stage, by
proper balancing, and by avoiding critical speeds.
The designer can eliminate some of the problems associated with torsional
vibrations by considering the following suggestions. Because lateral vibrations
can induce torsional vibrations, all lateral frequencies within 150°% of
maximum operating speed should be checked. Clearly the system should not
be operating at a lateral resonance speed. For gear systems operating at
moderate to high speeds, extremely accurately finished gears are desirable.
All torsional normal mode frequencies within 150% of the maximum
operating speed of the slowest shaft should be calculated. Amplitudes due to
these excitations should be determined in order to assure that they are within
predetermined maximum acceptable amplitude values. Torsional vibration
amplitudes should be measured during the test runs of the gear train to see
if they are in agreement with the calculated values.
 DP)
Problems

Calculations should be made to determine whether the shaft is reasonably

well balanced. Damping in the system may be increased by using journal or
tilting-pad bearings rather than rolling element bearings. The casing used to
enclose the gear train system should be designed for maximum possible
reasonable rigidity.
To sum up the important question of the vibration problems associated
with gear trains, it should be pointed out that, although every rotating system
is subjected to torsional vibrations, when a system has a gear in it the failure
usually occurs in the gear. However, the gear is rather a scapegoat in this
connection because its backlash is usually the only source of rotational
clearance that can producea vibrational failure. It is very often other elements
in the system that are causing the problem, but the only place where the
effect of the vibration can manifest itself is in the gear. Thus, it is clearly
necessary to do a torsional analysis of the entire system and not just of the
gear.

PROBLEMS
The problems in this section dealing with the AGMA method of solution are to be done
using the following assumptions: gears are to be used for general industrial application,
power source is subjected to light shock, driven machine is subject to moderate shock, a
life of 1 x 10° cycles and high reliability is required.
1. A pinion having 30 teeth, a circular pitch of 0.2618 in. and rotating at 2000 rpm
is to drive a gear rotating at 500 rpm. Determine the diametral pitch, number of
teeth of gear, and center distance.
2. A 20-deg full depth spur gear has 30 teeth and a diametral pitch of 4. Determine
the working depth, base circle radius, and outside radius.
3. A pair of spur gears in mesh have diametral pitches of 3, and a velocity ratio of +.
How many teeth must the gears have if the center distance is to be approximately
12 in.?
4. What is the approximate center distance for an external 20-deg full depth gear
having a circular pitch of 0.7854 in. that drives an internal gear having 80 teeth,
if the velocity ratio is to be }?
5. What is the contact ratio for two meshing, full depth spur gears having 145-deg
pressure angles, having addendums of 0.25in., and a velocity ratio of 4. The
pinion has 30 teeth.
6. Two meshing 143-deg pressure angle full depth spur gears have a circular pitch
of 0.7854 in. The pinion has a pitch diameter of 7 in. and the velocity ratio is to
be 0.6. (a) How many teeth does each gear have? (b) What is the whole depth?
(c) The clearance? (d) Root diameter? (e) Outside diameters? (f) Addendum circle
radius? (g) Will interference occur?
7. A gear having 30 teeth and a circular pitch of 1.0472 in. is in mesh with another
gear having a pitch diameter of 15 in. The teeth are 143 deg full depth. Determine
(a) The pitch diameter of the first gear. (b) The number of teeth of the second gear.
(c) The addendum. (d) The maximum addendum for the first gear for which no
interference will occur.
 596
Chapter 10: Spur Gears

8. The pinion of two meshing 144-deg full depth gears has 15 teeth and a circular
pitch of 1.0472 in. The velocity ratio is to be 0.25. Is interference present? If so,
what value should the actual pressure angle be made in order to eliminate the
interference?
©. A 20-tooth steel pinion is to drive a 30-tooth steel spur gear. Values of diametral
pitch and pressure angle are 2 and 20 deg, respectively. Tooth proportions are
standard stub involute. (a) Will the gear-tooth action in approach and recess be
satisfactory from the standpoint of interference? Compute the following : (b) Length
of contact path. (c) Angular speed ratio of follower to driver.
10. For the gear set shown in Problem Figure 10 the pinion, gear 1, delivers 20 hp
at 3000 rpm, has a diametral pitch of 3, 30 teeth, and rotates in the counterclockwise
direction. Gear 2, an idler, has 60 teeth, and the driven gear, 3, has 40 teeth. All
gears are 20-deg full depth. (a) What are the pitch diameters for the gears? (b)
Assuming 100% efficiency for the gears, what is the torque on shaft 3? (c) Draw a
free body diagram of the forces acting on each gear, and calculate the resultant
forces. ;

Problem Figure 10

11. Repeat Problem 10 for clockwise rotation of the driver.

12. For the gear train shown in Problem 12, the shafts | and 3 are to be 15 in. apart.
The driver gear A has 20 teeth, a diametral pitch of 5, and is rotating at 2500 rpm.
The velocity ratio between gears A and D is to be 4. Gear D has 70 teeth and a
diametral pitch of 7. All gears are 20-deg full depth. (a) How many teeth do gears B
and C have? (b) What is the torque acting on each shaft if the input horsepower
is 90? (c) what are the resulting shaft reactions if gear A rotates clockwise?

Problem
&
Figure 12
 S97.
Problems

13. Repeat part c of Problem 12 for counterclockwise rotation of gear A.

14. For the planetary gear train shown in Problem Figure 14, the pinion 1 rotates
clockwise at 1000 rpm. It has 30 teeth, a diametral pitch of 6, a 20-deg pres-
sure angle, and is to transmit 5 hp. Gear 2 has 60 teeth. What torque can the
arm deliver? Draw free body diagrams of all the components and show the
forces acting.

VAM

Problem Figure 14

Problem Figure 15
 598
Chapter 10: Sp ur Gears

15. A planetary transmission consists of a ring gear R (the input), three planets, a
fixed sun gear, and the planet carrier C (the output) (see Problem Figure 15). whew
diametral pitch is 10 and the pressure angle is 20 deg. 15 hp is transmitted with
input speed ng = 1500 rpm clockwise. The pitch diameters are 5 in. for the sun
gear, 24 in. for the planets, and 10 in. for the ring. (a) Determine the numbers of
teeth and speeds n, and n,. (b) Find the torque on input and output shafts and
sun gear. (c) Draw a free body diagram of the top planet. Indicate all forces in
pounds.
\e. In the planetary gear system shown in Problem Figure 16, gears B and C are keyed
toa shaft carried on revolving arm E. Arm E turns about the axis of gears A and D.
The tooth numbers are as follows: gear A, 100 teeth; gear B, 20 teeth; and gear C,
50 teeth. (a) If all the gears have a diametral pitch of 5, and a 20-deg pressure
angle, determine the number of teeth on D. (b) If gear D rotates at 1500 rpm clock-
wise and transmits 8hp, determine the angular speed (rpm) and direction of
rotation of arm E. (c) What torque can the arm deliver? (d) Draw a free body
diagram of all components and show the forces acting.

A (fixed)

Problem Figure 16

17. Determine a suitable pitch and width for a pair of meshing 20-deg gears, which
are to transmit 8 hp. The center distance may be taken as 10 in. and the velocity
ratio is to be 3. The pinion is rotating at 150 rpm and is made of mild steel, SAE
1030, while the gear is made of cast iron, grade 50. Consider strength only.
18. Check Problem 17 by the AGMA method.
19. A tentative design of an external spur-gear set specifies the following: Gear tooth
system, 20-deg full depth; Pinion material. steel, BHN = 200; Gear material.
phosphor bronze; diametral pitch. 6 ; Face width of gears, 0.60 in. ;Pitch diameter
 599
References

of pinion, 3.00 in.; Pitch diameter of gear, 5.00 in. Class 2 gears (gears cut with
great care) are to be used.
The design requirement is that the pinion transmit 13 hp to the gear at a pinion
speed of 1200 rpm. Design bending stress values are as follows: pinion, 0.70 times
the endurance limit in reversed bending; gear, 15,000 psi. (Note: These values
include the effect of stress concentration.)
Determine the beam strength of the (a) pinion tooth and (b) gear tooth. (c) Are
the gears satisfactory from the standpoint of beam strength? (d) On the basis of
wear strength, are the gears satisfactory?
Use the Lewis and Buckingham equations.
20. Repeat Problem 19 using the AGMA method. Use S,, = 10,000 psi; S,, = 38,000
psi.
‘21. Two meshing gears are 20-deg full depth, have a diametral pitch of 5. The gear is
made of phosphor bronze, has 20 teeth, rotates at 500 rpm. The pinion is made of
SAE 1040 steel and rotates at 125 rpm. Both gears have face widths of4 in. Based
upon strength only, what is the maximum horsepower that can be transmitted?
Use Lewis equation method.
22. Repeat Problem 21 using the AGMA method.
23. For the horsepower calculated in Problem 21, will the gearset be safe with regards
to wear?
24. Repeat Problem 23 by the AGMA method.
25. A pair of meshing gears have face widths of 25in., and diametral pitches of 4.
The gears are both made of a steel having a BHN of 400, and are 20-deg full
depth. The pinion is rotating at S00 rpm, has 30 teeth, and the velocity ratio is
to be 4. Assuming average operating conditions, what is the maximum horsepower
that can be transmitted, based on wear only?
26. Repeat Problem 25 by the AGMA method.
27. Design a gearset to transmit 50 hp at a pinion speed of 1200 rpm. Assume average
conditions of operation, and a velocity ratio of0.25.

REFERENCES
[1] C. E. Wilson and W. J. Michels: Mechanisms-Design Oriented Kinematics.
American Technical Society, Chicago, IIl., 1969.
[2] Wilfred Lewis: Investigation of the Strength of Gear Teeth. Engineers’ Club of
Philadelphia, Phildelphia, Pa., Oct. 1892.
[3] T. J. Dolan and E. L. Broghamer: A photoelastic study of stresses in gear tooth
profiles. Bull. No. 335, Eng. Exp. Sta. University of Illinois, Urbana, IIl., 1942.
[4] Earle Buckingham: Analytical Mechanics of Gears. McGraw-Hill Book Co., New
York, 1949.
[5] Harry Palton: Basic Fundamentals of Gear Lubrication. Gleason Works, Rochester,

[6] John Theberse: A guide to the design of plastic gears and bearings. Mach. Des.
(Feb: 5, 1970):
[7] D. W. Dudley: Gear Handbook. McGraw-Hill Book Co., New York, 1962.
[8] AGMA Publications.
[9] D. W. Dudley: Practical Gear Design. McGraw-Hill Book Co., New York, 1954.
 600
Chapter 10: Spur Gears

[10] The Internal Gear. Fellows Gear Co., Springfield, Vt., 1956.
[11] The Involute Curve and Involute Gearing. The Fellows Gear Co., Springfield, Vt.,
1955.
[12] The Art of Generating with a Reciprocating Tool. The Fellows Gear Co., Springfield,
Vt., 1958.
[13] P. L. Balise: Spur Gear Design Manual. University of Washington, Seattle, Wash.,
1957.
Helical, Worm,
Bevel, and
Other Gear Types

SYMBOLS

c = center distance P = diametral pitch in transverse

C,, = combined heat transfer plane
coefficient p = transverse circular pitch
F,, = bending load P, = axial circular pitch
F, = dynamic load P, = normal circular pitch
F, = normal force r, = velocity ratio
F, = radial or separating force V, = pitch line velocity
F, = tangential load Y = Lewis form factor
F,, = wear load W = helix angle
n= rpm A, = lead angle of worm
N, = number of teeth g = transverse pressure angle
N; = virtual number of teeth of bevel ¢, = normal pressure angle
gears
N,, le = virtual or formative teeth of
helical gears

In this chapter we will discuss the important gear types other than spur
gears used in design. The format employed in the previous chapter will again
be followed. Namely, a brief review of the kinematics involved will be followed
by a detailed discussion of the factors that must be considered in designing
systems using these gear types. However, because much of the discussion of
the previous chapter applies equally well to the present chapter, it will not
be necessary to go into as great detail in these areas.

601
 602
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

SECTION 11-1

Helical Gears
Figure 11-1 shows helical gears in mesh. The pair shown is mounted on
parallel shafts, the most common situation for which helical gears are used.
However, helical gears are sometimes used for nonparallel, nonintersecting
shaft applications. When used in this manner, they are known as crossed
helical gears. Figure 11-2 illustrates a pair of crossed helical gears in mesh.
We will discuss crossed helical gears in detail after considering helical gears
on parallel shafts.

Figure 11-1 When a pair of meshing helical gears are mounted on

parallel shafts, as shown, they must be of opposite hand. [Courtesy
Horseburgh and Scott.]
 603
Section 11-2: Helical Gear Tooth Loads

Figure 11-2 When meshing helical gears are mounted on

nonparallel shafts, they are known as crossed helical gears.
Crossed helical gears may be of the same or opposite hand.
[Courtesy Eaton Corporation, Industrial Drive Division.]

Before continuing with our discussion, it would be well to repeat the

criteria mentioned in Chapter 10 for determining whether spur gears or
helical gears are to be used for a given design. Spur gears are used for low speed
applications and those situations where noise control is not a problem. Use
of helical gears is indicated when the application involves high speeds, large
power transmission, or where noise abatement is important. We consider the
speed to be high when the pitch line velocity exceeds 5000 ft/min or the
rotational speed of the pinion exceeds 3600 rpm.
Whereas spur gears have their teeth cut parallel to the gear axis, helical
gear teeth are cut in the form of helices making a constant angle with respect
to the gear axis. Since it is obvious that the helix can slope either in the upward
or downward direction, the terms right hand and left hand helical gear are
used to distinguish between the two types. The rule for determining whether
a helical gear is right or left handed is the same as that used for determining
right and left handed screws. The helical gear shown in Figure 11-3 is right
handed, whereas the pinion in the figure is left handed.

SECTION 11-2
Helical Gear Tooth Loads
o mesh,
they must have the
ds, inaddition tohaving the same
One prime eminent of spur gears is that their initial contact upon
meshing is a line contact. The resulting instantaneous line contact produces
shock effects that tend to limit the magnitude of the load that can be trans-
mitted and result in rough, noisy operation. This problem is alleviated when
 604
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Figure 11-3 The gear shown is right handed, the pinion left handed. Because of
the large size of the gear, it was manufactured in two halves and then assembled as
shown. When gears are made in this manner, they are known as split gears.
[Courtesy Illinois Gear Division, Wallace Murray Corporation. ]

helical gears are used because the initial contact is a point, which becomes a
line of increasing length as contact continues.
A icra associated with the use of helical gears is that the helix
sults in a thrust loadin addition to the usual tangential and separating
loads. The ire components of the normal load acting on a helical gear
shown in Figure 11-4 can be written as follows:
= F cos'd, cos'v
F. =F, tang = F sin 9,
F hrust
nut = £, tan = F, cos @, sin ys
Figure 11-4 (a) Helical gear loads.
(Courtesy New Departure—Hyatt
Bearings Division, General Motors
Corporation.] (b) Components of helical
gear normal load.
Figure 11-5 (a) Train of continuous tooth
herringbone gears. In the continuous variety of
herringbone gears, the teeth are cut up to the center
of the gears. (b) Herringbone gears cut with a center
space, or clearance, for the cutting tool. [Courtesy
Illinois Gear Division, Wallace Murray
Corporation. ]

606
 607
Section 11-3: Helical Gear Terminology

Any Helical on Parallel Shafts

7° H. Driven L. H. Driver

Thrust Thrust

Thrust Thrust Thrust

NS 8. H,Driver R.H. Driven R.H. Driven

Figure 11-6 The diagram shown can be used to determine the direction of the
axial thrust load, which is always present with helical gears. The thrust loads must
be considered when choosing thrust washers or bearings. [Courtesy Insco
Corporation, Groton, Mass.]

The thrust load requires the use of bearings that can resist thrust loads as
well as radial loads. The need for a thrust resistant bearing can be eliminated
by using a herringbone gear. This is simply a helical gear with half of its face
cut with teeth of one hand while the other half has teeth of the opposite hand.
It should be clear that the thrust loads for each set of teeth will cancel each
other.
Figure 11-5 shows two typical herringbone gear types. direction
The in

| é ‘to that of ine aver Figure | iL= may

be theditto epienuine ie direction of the thrust load for helical gears on
parallel shafts.
Example 11-1. A helical gear set consists of a left handed driver, gear 1, having a
25-deg pressure angle, and a 30-deg helix angle, an idler, and a driven gear,
the centers of all gears are in line. Assuming that a 1000-1b load is to be trans-
mitted, determine the resultant loads on the shafts.

Solution:
Finrust = FF, tan y
= 1000 x tan 30 deg = 577 Ib
F. = F,tand = 1000 x tan 25 deg = 466 lb
Clearly, the resultant thrust load on an idler gear is zero. @

SECTION 11-3
Helical Gear Terminology
Figure 11-7 shows a helical gear with some of the more important geo-
metric quantities indicated. The helix angle wy is the angle between a line
drawn through one of the teeth and the center line of the shaft on which the
 NI

wo ~ ~
a
 609
Section 11-3: Helical Gear Terminology

gear is mounted. There are no standard values for helix angles because the
gears are rarely used interchangeably. The usual range of values used for the
le is betwee! 1 30 deg. Gears are sometimes made with helix
angles falling outside this range. However, since the thrust load varies
directly with the magnitude of the tangent of the helix angle, there is an upper
limit to the magnitude of the helix angle in order to prevent excessive thrust
loads. A lower limit is also necessary to ensure smooth transference of load.
Although spur gears involved only the diametral and circular iS helical
gear geometry requires additional pitches. The normal circular pitch p,
the distance between corresponding points of adjacent fete as reds in
aee acaba to the helix (the B-B plane in Figure 11-7). The

Figure 11-7 Helical gear geometry.

 610
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Equations (11-1) through (11-6) represent the geometric relationships

between the various pitches.
Dp, = PCOS (11-1)

Py = pcot (11-2)
p= = (11-3)

where P = diametral pitch in transverse plane, N, = number of teeth of the

gear, d is the diameter ofthe pitch circle, and P,, is the normal diametral pitch.
Poh=nx (11-4)

P,P, = 7 (11-5)
P
a 11-6
a COS'Y ( )
In order to ensure a smooth transference of load, the face width of a helical
-gear is usually made at least 20% greater than the axial pitch. As will be seen
later in this section, the face width is determined by the load acting on the
gear. The 20% greater than axial pitch is merely a suggested minimum value
for face width. In fact, some designers prefer that the face width be at least
twice the axial pitch.
Two pressure angles are associated with helical gears, one is measured in
the transverse plane (A-A plane in Figure 11-7) and the other in the normal
plane (B-B plane in Figure 11-7). Figure 11-8 shows the tooth profiles in the
normal and transverse plane.

Section A-A (transverse plane)

Section B-8 (normal plane)

Figure 11-8 The transverse and normal pressure angles are illustrated
by drawing sections A—A and B-B in Figure 11-7.
 611
Section 11-3: Helical Gear Terminology

Simple trigonometric relationships can be used to arrive at equation (11-7)

which relates the three angles associated with helical gears.

tan ¢, = tan¢ cos w (11-7)

where ¢, is the normal pressure angle, ¢ is the transverse pressure angle, and
w is the helix angle.
Other geometric quantities are similar to those for spur gears.

ie Nip _ Nn (11-8)
Tt m™cos W

d, SF d, p DAN, 4E Ne)
= = —(N No) = ; :
2 2 CN aut.) 2n cosw
. s (11-9)
pa t1 + t2

~ 2P_ cos y

Example 11-2. Two helical gears are to connect shafts 6 in. apart. The pinion has a
diametral pitch of 6, a normal diametral pitch of 7, and a pressure angle of
20 deg. If the velocity ratio is to be 4, determine the number of teeth of each
gear and the normal pressure angle.

Solution: The velocity ratio r, = N,,/N,,. Then

1 N"
2 Ne or N,,
= 2N,,

Tl Tl
Pp = Poo
— = —G = 0.5236
0.5236 in
in.

c=, +N,,)
2n
Cue OX 2x a
N
petNee) RS 2N,
OT =—=
pale aGRISEA = 7
2
SN = 72
So that N,, = 24teeth and N,, = 48 teeth.
For the normal pressure angle
IP &
cos Pee 0.8

Ww = 31 deg
and
tan f, = tan gd cos w = tan 20 cos w
= 0.364 x 0.857
Duis degu @
 612
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

SECTION 11-4

Virtual or Formative Number of Teeth

Referring back to Figure 11-7, the plane normal to the gear teeth, B-B,
intersects the pitch cylinder to form anellipse. The gear tooth profile generated
in this plane, using the radius of curvature of the ellipse, would be a spur gear
having the same properties as the actual helical gear. The radius of curvature
of an ellipse is given by
rset (11-10)
© ~ 2cos? y
where r,
The nu
known as |
number of teeth can be found by a
N,, = P,2r, (11-11)
where P,, is the normal diametral pitch. Therefore
d
N,, =P, x 2X =
: mS ire? w
But since
P N
Pa and P=—
cos w d
we get
Reals di) ON;
'e cosW cos?W cos*> &
or
N,
>SS N= ae) (11-12)
my le,

In other words, the equivalent number ofteeth, can be obtained by dividing

the actual number of teeth of the helical gear by cos? w. The equivalent number
of teeth is used to determine the Lewis factor Y in the bending stress formula
for helical gears to be discussed shortly.

SECTION 11-5

Dynamic Load for Helical Gears

The dynamic load acting on helical gears can be approximated by equation
(10-40).
78 V,
Bie Str
 613
Section 11-6: Bending Strength of Helical Gears

SECTION 11-6

Bending Strength of Helical Gears

The , Strength ears can be obtained by using a modified
cal gee
form of the Lewis equation or Cr using the AGMA formula. py 5”

The Lewis equation (slightly modified) 1, Qe

Set SbY aes Coe

(11-13)
Me cKP*
ih

can be used to determine the bending strength of helical gears, provided that
the form factor Y is obtained from Table 10-2, using the formative or equiva-
lent number of teeth. The normal diametral pitch is used because the tooth
load causing the bending stress is normal to the tooth surface in the normal
plane. F, must then be equal to or greater than the dynamic load F, as
calculated by the equation from Section 11-5.
The same AGMA equations used for spur gears (10-25) and (10-26) apply
to helical gears. The correction factors are obtained from the table and
figures as indicated below.’ Equations (10-25) and (10-26) are repeated below
for convenience.

FBR OPK as
an pen Ome ee Sa LTE 10-25
oO; K,bJ ( 0 )

Su Ke
Se ok 10-26

The overload factor K, can be obtained from Table 10-4. The size factor
K, is equal to unity for helical and herringbone gears. The dynamic factor
K,, can be determined from the appropriate curve in Figure 10-21. The load
distribution factor K,, can be obtained from Table 10-5.
The discussion of the geometry factor presented in Chapter 10 on spur
gears applies equally well to helical gears. However, the figures used for spur
gear geometry factors do not apply. Figures 11-9 through 11-13 can be used
for helical gear geometry factors.
The factors to be used in equation (10-26) as applied to helical gears are
obtainable from the tables and figures given in Chapter 10, namely, Table
10-7 for S,,, Table 10-8 for K,, and Table 10-9 or Table 10-10 for Kz. Kr is
either taken as unity or equation (10-27) may be used if appropriate.

' Extracted from AGMA Information Sheet—Strength of Spur, Helical, Herringbone, and Bevel
Gear Teeth (AGMA 225.01) with the permission of the publisher, the American Gear Manufac-
turers Association, 1330 Massachusetts Ave., N.W. Washington, D.C. 20005.
 614
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

ate Pv, , _Np Generating rack

N* 0.952 Pa 2|
Value for Z is for an element of indicated = Za as
numbers of teeth and a 75 tooth mate. g
Normal tooth thickness of pinion and gear €
tooth each reduced 0.024 in. to provide &
0.048 in. total backlash for 1 P,,4.
0.70
44
R ——
Poa

0.60

500 £
0.50 150 3

60 6
o
Jfactor,
Geometry
30 is
0.40 =
20° 2

0.30 =}= = :
0° 5 10° 15 20 25° 30 35°
Helix angle, w

Figure 11-9 Geometry factor (J) 15 deg normal pressure angle—indicated

addendum. [Courtesy American Gear Manufacturers Association. ]

Py N Generating rack
4N = 0.952
d="Py eateae hs 3
Value for Z is for an element of indicated &S =Ss Ee
number of teeth and a 75 tooth mate. ‘2
Normal tooth thickness of pinion and gear s =
tooth each reduced 0.024 in. to provide 0.048 in. S 20
total backlash for 1 Pag. =
0.70

R = 20
0.60 nd

Factors are for

€ shaved teeth
0.50 & cutwitha
‘S preshave hob
Geometry
J
factor, 5
no}

0.40 E
z

0.30

Helix angle, w

Figure 11-10 Geometry factor (J) 20 deg normal pressure angle, standard
addendum preshave hob. [Courtesy American Gear Manufacturers Association. }
 615
Section 11-6: Bending Strength of Helical Gears

Pn, N Generating rack

Maca pee ON eB
0.952 Py
Value forZ is for an element of indicated
numbers of teeth and a 75 tooth mate.
Normal tooth thickness of pinion and gear
tooth each reduced 0.024 in. to provide 0.048 in. height
Tooth
total backlash for 1 P_,.
0.70
157
R=-
Prd
0.60
=
< Factors are for
° .
2) teeth cut with
£ a finishing hob
= oe Sas the final
= & machining
8 C3 operation.
2 S
0.40 8
=
=|
2

0.30 5 = - -
0 5 10 15 20 2be 30° aie"
Helix angle, w

Figure 11-11 Geometry factor (J) 20 deg normal pressure angle, standard
addendum finishing hob. [Courtesy American Gear Manufacturers Association. ]

Phe aa ety
N ~ 0.95Z Ps Generating rack
Value for Z is for an element of indicated
numbers of teeth and a 75 tooth mate. z
Normal tooth thickness of pinion and gear S
tooth each reduced 0.024 in. to provide &
0.048 in, total backlash for 1 P,,4. &
0.70 bK
3

x 0.60
> s
2 500 9%
8 150 = Factors are for shaved
> 0.50 60 © teeth cut with a
a 30 5 preshave hob.
E
1o)
0.40

0.30 averwewes U3. l a !

0° 5° 10° 15° 20° 25 30 35
Helix angle, y

Figure 11-12 Geometry factor (J) 22 deg normal pressure angle. [Courtesy
American Gear Manufacturers Association.]
 616
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

1.10

1.05
———} | 500

150
<
1.00 75 5
Pa Be Soe 50 85
oO 2
a o

K
factor,
Modifying
0.95 | 30
£2
a 08
20e

20
0.90

PE 5 10
ene ex15 20
ce i 25
| mma.
30 35
Helix angle, w

Figure 11-13 J factor multipliers for 15 deg normal pressure angle. The
modifying factor can be applied to the J factor when other than 75 teeth
are used in the mating element. [Courtesy American Gear Manufacturers
Association. ]

SECTION 11-7

Surface Durability of Helical Gears

The equations that were discussed in Chapter 10 on spur gears for wear
loads can be used to determine the wear load on helical gears.
The Buckingham equation for allowable wear load, equation (10-32), can
be used with only slight modification. Namely,
ae avOk
(11-14)
cos* w

The symbols in the above equation have the same meaning as defined in
equation (10-32), except that K is determined by using the normal pressure
angle, @,.
As was the case with spur gears, the alternate wear load equation for helical
gears is taken from the AGMA bulletin cited previously.
Equations (10-33) and (10-35) are repeated here for convenience.
FGae. CC.
a: Cy | eae (10-33)
C.C
O.< Su - | (10-35)
 617
Section 11-7: Surface Durability of Helical Gears

W, = tangential load, Ib
e = alignment error under load, in./in.

Z = length of action
P, = base pitch, normal to involute, transverse plane

5.0

4.0

ee z
ssc i 1/5/
Force,
C,
coefficient,
error
Pe8/1
| |
1.0 2.0

b = face width
in.

Ln

Figure 11-14 Helical gear load distribution factor (under load), C,, [Courtesy
American Gear Manufacturers Association. ]

The curves and table previously presented in Section 10-17 apply equally
well except for the following :
1. Instead of Figure 10-30, which illustrates the effect of different rates of
misalignment, Figure 11-14 should be used if enough information is available.
 618
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

2. The geometry factor I is not presented in tabular or graph form by the

AGMA, and it is thus necessary to calculate J for helical gears by the use of
equation (11-15) as given in the AGMA bulletin

|Mee we
meee: (11-15 )

where C, = curvature factor at pitch line, and My = load sharing ratio.

C, is given by
Ce M, (11-16)
2 MPA
where: @ = operating transverse pressure angle in degrees and M, = gear
i 1
ratio = —.
he
Nae Pn,
PAZ
T0957 ( )
where p,,, is the normal base pitch in inches, and Z is the Icngth of action in
transverse plane in inches. Equation (10-7) may be used to calculate ‘the
value for Z, where the addendum is equal to 1/P,,, and p,, = p, cos @,.

Example 11-3. How much horsepower can be safely transmitted by a pair of helical
gears, 20-deg full depth, 25-deg helix, having a normal diametral pitch of 5,
both made of SAE 1040 steel, and having a face width of 3 in. The pinion is
rotating at 2000 rpm has 20 teeth, and the velocity ratio is to be | to 5. Deter-
mine the maximum horsepower that can be transmitted (a) using the modified
Lewis and Buckingham equations, and (b) the AGMA method based on
strength only.
Solution:
Part a. We will first use the modified Lewis equation (11-13) to determine the
allowable strength in bending. Because both gears are made of the same
material and the pinion will have a smaller Lewis form factor, it will be the
weaker.
From Table 10-3 S$, = 25,000 psi. To obtain the Lewis form factor Y, we
must calculate the formative number of teeth. From equation (11-12)

ee ee ee, ee Be 20
== 7etecun
‘e cos? W (cos 25)> —-(0.906)?_—«0.744
From Table 10-2, Y = 6348.

concentration
factor may be taken as |.Therefore, from equation (11-13)
SbY — 25,000 x 3 x 0.348
ey wen Lx
= SP) II
The allowable wear load will now be determined from equation (11-14)
_ d,bQK
” cos?
 gle
Section 11-7: Surface Durability of Helical Gears

From equation (11-7)

tan d, = tan dcos y
= tan 20 cos 25
= 0.364 x 0.906 = 0.33
fh, = 18.3 deg
From equation (11-6)
P= Pe cosay = o2~ 01906= 453
N, 20
ie sare oars
Ge Pp eiey E
a gee Ne
a Sia oN()
N=
12 SN, t = S & AO) = ICO wan
Since
0 = ES Ls 2 x 100
IN oN 20 + 100

200
pope
From Table 10-3, BHN = 202. From Table 10-11, K =79. Therefore
_ a)bQK
om COST)
443) 167 579
(0.906)?
= MI Ne
The dynamic load can be determined from equation (10-40)
ra 18 + WV p
(die 78 t

Because the allowable load in wear is less than the allowable in bending, it
is used for F,.
_amdn_
2x 44 x 2000
Peotl2 2 12
= 2304 ft/min
USE i oe ele
SE N/V, 8/2304
alll bs
y
'"p
safe horsepower = 33,000

_ 131233,000
x 2304 = NLS ajo)
 620
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Part b. AGMA strength equation from equation (10-26)

= SaKr
Sad K7Kp

From Table 10-7 for BHN of 202 interpolation yields S,, = 27,000 psi. For
indefinite life K, = 1; Ky = 1 is reasonable and Kz = 1.5 from Table 10-9.
Therefore
27,000 x 1 :
Soa = ar eh ae = 18,000 psi

from equation (10-25)

i oe. where g, is equal to S,4.

K, = 1.5 assuming moderate shock and a light shock power source. K, = |

for helical gears. From Figure 10-21 curve 2, K, = 0.79. From Table 10-5,
Korea
For ¢, = 18.3 deg, we can estimate J by interpolation, using Figure 11-9
for 15deg and Figure 11-10 for 20deg. For ¢, = 1S5deg, J = 0.42; for
ob, = 20 deg, J = 0.45. Therefore,
for ¢, = 18.3 deg, J = 0.44. But because the
number of teeth in the mating element is not 75, correction factor K must be
applied to J. From Figures 11-13, K = 1.01 (approximate). Then
J = 1.01 x 0.44 = 0.444
_ 18,000 x 0.79 x 3 x 0.444
io 1:55 4535er < 12
= 2320 1b
Therefore

FV
bP = 33 000
= t Pp

2320 x 2304
> ass es 132 hp

compared to 91.6 for Parta. @

SECTION 11-8
Crossed Helical Gears
When helical gears are used to transmit power between nonparallel, non-
intersecting shafts, they are known as crossed helical gears. Figure 11-2
depicted a pair of crossed helical gears in mesh. It should be emphasized that
crossed helical gears are not a breed in themselves but rather are ordinary
helical gears used in a nonparallel shaft application. In order for two helical
gears to operate as crossed helical gears, they must have the same normal
pitch P,,, and normal pressure angle, ¢,,. In other words, in the crossed gear
application, the gears do not need to have the same helix angle, nor must they
be of opposite hand. In fact, in most crossed gear applications, the gears have
 62]
Section 11-8: Crossed Helical Gears

the same hand. Crossed helical gears have point contact, rather than the line
contact of regular helical gears. Although the running-in period tends to
extend the point contact into line contact, the contact still remains poor, and
crossed gears are usually used only for problems involving small transmissions
of load.
The important kinematic relationships associated with crossed helical
gears, may be summed up by the discussion of the following equations. The
relationship between the helix angles of the gears and the angle between the
shafts on which the gears are mounted is given by
= Wie ws, (11-18)
for gears having the same hand, and

2= Wi — Wp (L119)
for gears having opposite hands, where £ = shaft angle.
A very common use for crossed gears is the one where the shaft angle is
90 deg. Clearly, for this situation, the helical gears must be of the same hand.
Equation (11-20) can be used to calculate the center distance between
crossed helical gears.
lafeNe N t2
(C= 11-20
Ne cosW, cosw, (

Figure 11-15 The diagram shows a pair of meshing crossed helical gears.
The relationship between the helix angles and the shaft angle is shown.
 623
Section 11-9: Worm Gears

where c = center distance; P,, = normal diametral pitch; N, = number of

‘teeth; w = helix angle.

_N,, _ 4, cos,
=
CINE dacosys
fies!
he)
t2

The sliding velocity, V,, acts along the tooth surface, in other words, in the
tangential direction and can be obtained from equation (11-22).
V, = Vy sinw, + V; siny, (11-22)
Figure 11-15 illustrates the situation given by the above equation.
As was indicated previously, the most common application for crossed
helical gears occurs for shaft angles of 90 deg. For this special case, the
velocity situation is as shown in Figure 11-16, and the sliding velocity is
given by

V.= aaeheAe (i238)

“cos Wp. eos Why *
Figure 11-17 may be used to determine the direction of the thrust load for
crossed helical gears.
Crossed helical gear teeth do not conform to any specific standard as far
as tooth proportions are concerned. With regard to contact ratio (number of
teeth in contact) the designer is advised to use a value of at least 2, since the
teeth have point rather than line contact.
The forces that act on crossed helical gears are the same as those for worm
gearsets to be described in Section 11-9.
As for wear and strength, the bending stress does not limit the design if the
teeth have a reasonable pitch. Thus, there is no need to analyze the bending
strength. An allowable wear load calculation can be found in reference [1].
The efficiency of crossed helical gears is determined by the same procedure
used in Section 11-9, dealing with worm gears.

SECTION 11-9
Worm Gears
Although crossed helical gears can be used for applications involving
nonparallel, nonintersecting shafts, they are rather limited in their load
transmission capacity. In addition, large speed ratios, on the order of say
200 to 1, could not be reasonably achieved in one reduction because of the
size of gear required. For example, if the smaller gear had 20 teeth, the larger
would have to have 4000 teeth. A much more feasible solution to the problem
would be the use of aworm gear set.
A worm gear set such as shown in cada11-18 consists oftheworm, which
is very similar to a screw, and the worm gear, v . helical gea
shafts upon which the worm and the gear are mounted are Peele 90 deg
 624
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Figure 11-18 A single enveloping worm gear set.

(Courtesy Illinois Gear Division, Wallace-Murray
Corporation. ]

apart. One of the advantages associated with the use of worm gears is that
the tooth engagement occurs without the shock prevalent in other gear
types. In fact the meshing of two teeth occurs with a sliding action which,
while resulting in very quiet operation, may produce overheating.
Worm gears may be either single or double enveloping. In a single en-
veloping set, Figure 11-18, the helical gear has its width cut into a concave
surface, thus partially enclosing the worm when in mesh. An example of the
double enveloping type is shown in Figure 11-19. In addition to having the
helical gear width cut concavely, this type also has the worm length cut
concavely. The result is that both the worm and the gear partially enclose
each other. A double enveloping set will have more teeth in contact and will
have area rather than line contact, thus permitting greater load transmission.

gy OF
™,

Figure 11-19 A double enveloping

worm gear set. [Courtesy Michigan Tool
Company. ]
 625
Section 11-10: Worm Gear Terminology

All worm gears must be carefully mounted in order to assure proper

operation. The double enveloping or cone type is much more difficult to
mount than the single enveloping.
The worms are usually made of case hardened alloy steel, and the gear is
usually made of one of the bronzes. The worm is usually cut on a lathe,
whereas the gear is hobbed.

SECTION 11-10
Worm Gear Terminology
The terminology used to describe the worm is very much like that used in
giscussine Doses screws. Th

NO! For hans that are

90 deg apart, the axial pitch of the worm is equal to the circular pitch of the
gear. Most worms have multiple threads so that the lead is obtained by
multiplying the number of threads (or number of teeth) by the pitch.
The more important kinematic equations applying to worm gearing are
listed below.

tan/,, = = —s (11-24)
nd Ww

where

V,, = pitch line velocity of gear

V,,, = pitch line velocity of worm
i,, = lead angle of worm, the angle between a tangent to the pitch helix
and the plane of rotation
l lead = N,. X P,,, = number of teeth or threads of worm times the
axial pitch of worm.
d w = pitch diameter of worm
Relations involving the angles of the worm and gear are
A, + Wy = 90 deg (11-25)
where y,, = helix angle of worm
also
A, + W, = 90 deg (11-26)

and for 90 deg shaft angles

Aw =W, (11-27)
The center distance is given by

g (11-28)
 626
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

The AGMA recommends the following equation be used to check the

magnitude of d,,

Gees nity (11-29)

where p, = circular pitch of gear in inches.

The velocity ratio is given by
Neen
~ = 11-30
any w N tg md g ( )

Example 11-4. A triple-threaded worm, having a lead of 3in., meshes with a gear
having 45 teeth and a normal circular pitch of 0.9455 in. Find the center
distance between the shafts if they are 90-deg. apart.
Solution: Since] = N,, x p,.,

ci mivller
Ne wads,Sores§ lin

From equation (11-6), P, = P/cos w,*

cosy
e =~ses =—#
© Pag Pa
but p,,, = p,. Therefore

Soa eae OPES Fons

Pik Pas l
W, = 19 deg
Therefore from equation (11-27), 2,, = w, = 19 deg, and from equation (11-24)
/
tand, =
td,

ee ee
“"ntand,
2 x 0.344
eA OuiiTe

. ee
MP TRIOS
But
= l
eae
Thus
l 3
d,=—= ~ = 14.33in
Tr, WS Fag

From equation (11-28)

Vide Ge oor aoo
a 2
= 8.57in. @
 627
Section 11-12: Dynamic Load

SECTION 11-11

Worm Gear Strength

In worm gear sets,
worm gear teeth. The simplified Lewis equation is therefore applied to the
worm gear in determining strength.
SYb
F, = —— = Sbyp, (11-31)
n

where the terms in the equation have the usual meaning.

Values for the Lewis form factor can be obtained from the following table.
It is suggested that the values be used only if the sum of the worm and worm
gear teeth exceeds 40.

SECTION 11-12

Dynamic Load
The dynamic load can be estimated by using the equation (11-32)

_ (1200 + ¥,, (11-32)

: 1200 t

where V,,= pitch line speed of the gear in feet per minute, and F, = trans-
mitted load as determined by the horsepower applied to the gear.
It should be pointed out that the horsepower on the gear is the output
horsepower, which will depend on the efficiency of the set. However, since
the efficiency is not known at the initial design stage, the usual procedure is
to use the input horsepower, realizing that the F, value thus obtained will be
higher than is the actual case. But, since we require F, to be greater than or
equal to F,, this procedure is clearly conservative.

Table 11-1 Normal Pressure Angle

a RR SE ES ES

$y» deg. Y ye

14 1/2 0.314 0.100

20 0.392 0.125

25 0.470 0.150

30 0.550 0.175
 628
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

SECTION 11-13

Wear Load Equation

An approximate equation suggested by Buckingham [1] is usually used to
determine the allowable wear load.
F, = d,bK' (11-33)
where d, = pitch diameter of gear;b = face width of gear
;and K' = constant
which depends on the material and geometry of the gear. Table 11-2 may be
used to obtain values of K'.
The usual requirement is that F,, > Fy.

SECTION 11-14
Efficiency of Worm Gears
In order to discuss the efficiency of worm gears properly, it 1s necessary to
consider the tooth loads that are acting. The forces acting on worm gears are
shown in Figure 11-20. The normal force F, acts perpendicularly to the
tooth surface and is broken up into the following three components :
F, = transmitted force which is tangential to worm and axial to the
gear— F (cos'@ sin Ay,
Finrust = axial thrust load on worm and tangential or transmitted force on
worm gear = F,, cos ,, COS Ay
F, = the radial or separating force = F, sin @,.
Figure 11-21 may be used to determine the direction of the thrust load on a
worm gear set.

Table 11-2 Wear Constant; K’, for Worm

Gears with 0, 20 deg‘

Worm Gear K!

Steel, 500 BHN Bronze 80

Steel, 250 BHN Bronze 60

Steel, 500 BHN Chilled bronze 115

Steel, 500 BHN Cast iron 50

SSE EI SSL SN RE RESETS PET SS OT

"The values for K! given above are typical values. The wear load should only be
estimated from equation (11-33), A more complete set of values may be found
in reference [1].
 629
Section 11-14: Efficiency of Worm Gears

Figure 11-20 Bearing loads due to worm gearing. [Courtesy New Departure—
Hyatt Bearings Division, General Motors Corporation.]

The efficiency of a worm gear set is defined as the output horsepower

divided by the input horsepower. The frictional force acting between the
worm and the gear is equal to fF, and is resolved into components fF, cos /,,
along the axis of the worm and fF, sin J, along the axis of the gear.
The resultant force along the axis of the gear (in the tangential direction
for the worm), and therefore the input force of the worm, is equal to F, cos @,
sini, + fF, cos 4,,, while the resultant force in the tangential direction of
the gear (in the thrust direction for the worm), and therefore the output
force on the gear, is equal to F, cos ¢,, cos /,, — fF, sin 4,,.
The input and output power can thus be determined as follows:
power; = (F,cos @¢, sind, + fF, cos4,)V,

power, = (F, cos ¢, cos J,, — fF, sin 4,,)V,,,

 630
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Thrust
ae

R. H. Driver R.H. Driver

Thrust
——
i
R.H. Driver R.H. Driver

R.H. Driver R.H. Driver R.H. Driver R.H. Driver

Thrust
—-———
Thrust
_—_——s A
(
Thrust Thrust
= —s
Right Hand

|
Thrust
ee

Thrust
——
7,

rc .H. Driver
L.H. Driver
Thrust N Thrust
— —_—\—

Thrust
—_——

Left Hand

Figure 11-21 Direction of thrust load for a worm gear set. [Courtesy Insco
Corporation, Groton, Mass.]

where V,., = pitch velocity of worm and V,, = pitch line velocity of gear.
Therefore

_ power, _ F,(cos , cos A,, — f sin allge

ff = ;
: power; F,(cos@¢,sind,, + fcos a,)\y
_ (cos ¢, cos A,, — f sin A,) tan A,
~ ((cosousin AA + f cos 10)
cos g, — ftani,
= tand,
cos @, tani, + f
(11-34)
 631
Section 11-15: Thermal Capacity of Worm Gearsets

Another useful form is

COS @, J lady,
(11-35)
WcosG) fF cots,
Comparison of this equation with the power screw efficiency equation,
shows that they are the same.
Most references give values for the coefficient of friction as a function of
the sliding velocity, V,.

Vo = a
(11-36)
cos A,,
There seems to be no agreement as to a uniform set of values for the
coefficient of friction. The following equations are suggested by Faires [2] as
a reasonable compromise for a carburized and ground worm driving a
phosphor bronze gear.
0.155
f= or where3< V, < 70 ft/min (11237)

or

0.32 .
f=To36 for 70 < V, < 3000 ft/min (11-38)

Another consideration regarding the frictional force in worm gearsets is

the question of whether the gearset is reversible or not. Most sets are speed
reduction units ;1n other words, the worm is the driver. The gearset can be a
speed increasing unit if the gear is the driver. Thus a reversible gearset is one
in which either the worm or the gear can be the driver. If the frictional force
developed is large enough, the gearset will be irreversible or self-locking. In
general, when the lead angle of the worm is 10 deg or less, the set will be
self-locking. This self-locking effect can be an advantage in applications
where a braking action is needed. However, the designer is cautioned to
include a secondary braking device in the design, because otherwise a gear
failure will result in the complete loss of braking power.
Finally, it should be obvious that the irreversible gearset will have a much
lower efficiency than a reversible set because of the large frictional force
developed.

SECTION 11-15

Thermal Capacity of Worm Gearsets

One of the major problems associated with worm gearsets is the question
of how much heat is developed during operation and whether the gear case
is capable of dissipating this heat. In fact, most worm gear units have their
horsepower capacity limited by the heat dissipation ability of the casing.
 632
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

The transfer of heat from the casing is accomplished by both radiation and
convection. In arriving at an equation to determine how much heat can be -
dissipated, such factors as housing area, temperature change between
lubricant and the ambient air, and a combined heat transfer coefficient must
be considered. The usual heat transfer equation can be written as follows:
IH =s Ge Arie (11-39)

where

H = the energy dissipated through the housing (lost energy or frictional

energy) ft-lb/min
C., = combined heat transfer coefficient, ft-lb/[(min)(in.7)(°F)] (Figure
11-22)
A, = area of housing exposed to ambient air, in.”
At = temperature difference between oil and ambient air, °F
The cooling surface area of the housing can be estimated by equation
(11-40) which is recommended by the AGMA.

A, = A5e (11-40)

0.5

0.4

.
ne
A)

Bs ie
ow
se o

en,
© c
Sc
eee
25 02 —
&
iS
°
8
8
ORS)

0.1 =

0
10,000 20,000
Housing area, A, in.2

Figure 11-22 Plot of heat transfer coefficient versus

gear housing area.
 633
Section 11-16: Worm Gear Standards and Proportions

where A, = housing cooling area in square inches and c = center distance

in inches.
The heat that must be dissipated from the casing can be determined by
considering the frictional or lost horsepower. Since

ff pehp,
=
. hp;
and
hp, = hp; ae Prost

e hp; = hp; = hpjost

or
NPist = 2p; — ep, — hp — <)
Then the energy which must be dissipated is given by
H, = hp1 — e) x 33,000 (11-41)
Clearly the heat energy developed H, as determined by equation (11-41)
must be equal to or less than the heat energy dissipation capacity H as
determined by equation (11-39).

SECTION 11-16
Worm Gear Standards and Proportions
There are no real standards or proportions that are accepted universally.
However, some suggested values, which may be of some use to the designer
are now presented.
With regard to the axial pitch of the worm, the AGMA suggests the
following values as standards:
iE seis Reel aed} 1 ft 3
49 1162 82% 25 8» rey ly tees 15, i252

The choice of pressure angle and lead angle must be made, keeping in mind
that too large a lead angle for a specific pressure angle will result in excessive
undercutting of the gear tooth. Buckingham [3] recommends the following
limiting values for lead angles.

Lead Angle, X,,

Normal Pressure Angle, ¢,, GmINnaE Te ca

14 1/2 deg 16 deg

20 deg 25 deg

25 deg 35 deg

30 deg 45 deg
RE ARES RL TN SI LIES ELE DOT LED LESLIE ELE ELIEL LEAP AML
 634
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

A reasonable value for the lead angle has been suggested as 6 deg of lead
angle for each thread of worm. Thus a quadruple threaded worm would have
a lead angle of 24 deg.
The face length, or axial length of the worm may be approximated by
equation (11-42)
N,
Ai oat
L=p, 50 (11-42 )

As far as the gear face width is concerned, care should be exercised by the
designer to prevent the width from becoming too large. Because of the non-
uniform distribution of load across the face of the gear, the larger the width
the greater will be the difference between the average load and the maximum
load. A reasonable proportion would seem to be that the width should be no
more than the radius of the outside circle of the worm.
Finally, it should be stressed that the suggested proportions given above
are merely that, suggested values. They can help the designer start his
solution, but they are relationships that can be deviated from. Rather than
going into an extensive discussion ofa typical worm gearset design procedure,
the following example will be used to demonstrate one possible approach to
the problem.

Example 11-5. Design a worm gearset to deliver 15 hp from a shaft rotating at

1500 rpm to another rotating at 75 rpm.
Solution: Assume that the normal pressure angle is to be 20 deg. The lead
angle A,, should therefore not exceed 25 deg (from the table on page 633).
Allowing 6 deg per thread of worm, the worm could have 4 or less teeth. Let
us try a triple threaded worm.
n, 75 1 N,
CS aRISOU) uD Ole LN
Therefore
N,, = 20 x N,,, = 20 x 3 = 60 teeth
Assume a center distance of 10in. The magnitude for center distance is
based upon actual limitations in design or is based upon designer’s judgement.
c0-875 10°-875
hnBS 1 es 3.41 in. Try d,, = 4 in.

Since d,, © 3p,

p, = Ady = 1.33in. = py,. Try py, = 1.

Poa 3 fate
g
D g

Pee Ne 0 = 119i ty
= =P, Sadie
Op
dn 4oelot
actual c =
2 ry D

LSS rb.
 635
Section 11-16: Worm Gear Standards and Proportions

Check d,, = 11.55°'875/2.2 = 3.87 in. Therefore d,, of 4 is satisfactory.

lead = N,, x py,
= 3 x 1 = 3in.
l 3
tan = = 0.239
ar mwa, qx 4

A,, = 13.43 deg

Aw = W, = 13.43 deg
JP 3.1416
= £ SS —

"@ cosy, cos 13.43 es

rev ye veel al 1 ft
Vo =e YS g
ve min : rev ‘ 2 - 12 in.
TS % He SX IOs! :
aD = 375 ft/min

T= hp x 63,000 — 15 x 63,000
n = 75
= 12,600 in.-lb

Ta 12,600
t = —_
d,/2 = ——
19.1/2 = i310
3 0 It) = 18thrust worm

(1200 + V,,)
ae ea
ee eg
Fi 1200 :
(1200 + 375)

Assume b = 5d, = 4 x 4 = 2 in. from Table 11-1, Y = 0.392. From equa-

tion (11-31)

F,
nse
P F,
Therefore
ba MiB0R 323
Se
Yb 0.392 x 2
= SSI
From Table 10-3 it can be seen that a phosphor bronze gear having an S, of
12,000 psi will be satisfactory, since 12,000 > 7131.
Check for wear. Assume worm is made of hardened steel. From Table 11-2,
K' = 80. From equation (10-33)
Fy = d,bK)
= IO <2 xe SS Mo Ile
Since F,, > Fy, the gearset is satisfactory for wear.
The AGMA recommendation for axial length of worm is
N
454+ —8
os 50
60
=1 45s50 = 3,// in
sha,
 636
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Cooling check. We will first determine the heat developed by using equation
(11.41).
y 1500 2° 5 rad dw J 1 ft
ES Pinte <.
sleW emia

eS
DS 172
oT ait
From equation (11-36)
Vien Ol
* “cosa, © cos 13:43
= 1615 ft/min
From equation (11-38)

0.32 0.32
f = Fase = Gepaose = 0.0224
F,, = F,, cos ¢, cos 4,
oye es ate
"cos @¢, cosa, cos 20 cos 13.43
= a4 53.5010
power; = (F,, cos ¢, sind, + fF, cos a,)\,
= (1433.3 x 0.940 x 0.232 + 0.0241 x 6041 x 0.973)1571
= 540,130 ft-lb/min = 16.37 hp
cos g, — f tani,
~ Cos ob, + f cota,
cos 20 — 0.0224 x tan 13.43
~~ cos 20 + 0.0224 cot 13.43
0.938 — 0.0224 x 0.239
~ 0.938 + 0.0224 x 4.19

From equation (11-41)

H, = hp, — e) = 16.37(1 — 0.904)
= 1.57 hp = 51,860 ft-lb/min

We will next use equation (11-39) to determine the heat that can be dissipated.
We will further specify that the gearset is to be enclosed in a case and oil be
supplied to the case. Then from equation (11-40)
A, = 43.2c!7 = 43.2(11.55)!-7
= 2766 in?
From Figure 11-22
C,, = 0.43 ft-lb/(min)(in.*)(°F)
 037,
Section 11-17: Bevel Gears

A reasonable temperature difference would be 100°F (ambient air ~ 80°F,

oil film ~ 180°F). Then from equation (11-39)
Jal = C,Al.Gh
= 0.43 x 2766 x 100 = 119,000 ft-Ib/min
Therefore 51,860 ft-lb/min are developed while 119,000 ft-lb/min can be
dissipated.
Overheating will not bea problem. @

SECTION 11-17
Bevel Gears
The last important type of gear used to transmit power between non-
parallel, intersecting shafts is the bevel gear. Bevel gears can perhaps be best
described as being conical gears. In other words, whereas the gears discussed
to this point were cut on cylindrical blanks, bevel gears are cut on conical
blanks.
There are four primary types of bevel gears, shown in Figures 11-23
through 11-26.

Figure 11-23 Straight bevel gears. [Courtesy Eaton Corporation,

Industrial Drives Division.]
 638
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Figure 11-25 Spiral bevel gears.

[Courtesy Gleason Works, Rochester,
N.Y.]

Figure 11-26 Hypoid bevel gears.

[Courtesy Gleason Works, Rochester,
N.Y.]
 639
Section 11-18: Straight Bevel Gears

SECTION 11-18
Straight Bevel Gears
The simplest of the four types of bevel gears mentioned above is the so-
called straight bevel gear. The name is derived from the fact that the teeth are
cut straight, have a taper, and if extended inward, would intersect each other
at the axis. Straight bevel gears are used primarily for relatively low speed
applications with pitch line velocities up to 1000 ft/min and where smoothness
and quietness are not an important consideration. However, with the use of
a finishing operation, such as grinding, speeds up to 15,000 ft/min have been
successfully handled by straight bevel gears.
The terminology and important physical characteristics associated with
bevel gears are shown in Figure 11-27.

Apex of pitch cone

Face cone

Pitch cone Addendum angle

Dedendum angle
Root cone

Addendum

Dedendum

Outside pitch diameter

Inside pitch diameter

~ Outside diameter

Figure 11-27 Bevel gear terminology.

 640
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Bevel gears are also classified according to the pitch angle of the gear.
Figure 11-28 shows a pair of regular bevel gears with a shaft angle of 90 deg.
Most bevel gear applications involve shaft angles of 90 deg, but there are
applications for which the angle is something other than 90 deg. When the
shaft angle is 90 deg and the gears are of the same size (the velocity ratio is
equal to 1), the gearset is known as a miter gearset, and each gear has a pitch
angle of 45 deg.
When the pitch angle of a bevel gear is less than 90 deg, as is the case with
the gears shown in Figure 11-28, it is known as an external bevel gear. Gears
having a pitch angle of 90 deg, Figure 11-29, are known as crown gears, and
those having pitch angles greater than 90 deg are called internal gears.
Straight bevel gears may be cut with a wide range of pressure angles, such
as 145 deg, and so on, but the basic standard pressure angle is 20 deg. The
working depth is 2/P in., the clearance is (0.188/P) + 0.002 in., the pinion is
cut with a long addendum whereas the gear has a short addendum, and the
teeth are not stubbed.
The wearing characteristics improve as the number of teeth is increased.
It is suggested, therefore, that the minimum number of teeth for most bevel
gears be 13, whereas for miter gears the limiting value should be taken as
16 teeth.
Straight bevel gears are the most economical of the various types of bevel
gears to be discussed.
Finally, a modified type of straight bevel gear called a coniflex gear has
been developed, Figure 11-30. It has generated and ground teeth and is
formed to have localized lengthwise tooth contact. The result is that the
concentration of load on the ends ofthe teeth, due to assembling or deflection
under load, is prevented.

Figure 11-28 Regular bevel gears with 90 deg shaft-angle.

 641
Section 11-18: Straight Bevel Gears

Figure 11-29 Bevel gears having a pitch angle of 90 deg are known as
crown gears.

The velocity ratio for bevel gears is given by

r= Nariven (1 1-42)

Nariver

As can be seen in Figure 11-28, the shaft angle is equal to the sum of the
pitch angle of the pinion and gear.
Z=T+y (11-43)

Figure 11-30 Coniflex bevel gear. [Courtesy Arrow Gear Company. ]

 642
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

where © = shaft angle; T = pitch angle of gear; and y = pitch angle of

pinion.
The pitch angles can be determined by equations (J 1-44) and (11-45).
sin X
ran" P== (W,JN,) + cos = 11-44
ae
sin &
tan y= (1 1-45)
(N,,/N,,) + cos 2
Clearly if the shaft angle is equal to 90 deg, the equations become

tanT = Neg
ey gee (11-46)
tanany
y = N,.
N, —!2 (11-47)

SECTION 11-19
Formative or Equivalent Number of Teeth
In the discussion of helical gears, it was pointed out that the tooth profile
in the normal plane would be a spur gear having an ellipse as its radius of
curvature. The result was that the form factors for spur gears applied,
providing that the formative number of teeth were used in determining the
tabular values. |
A somewhat similar situation exists with regard to bevel gears. The tooth
profile in the plane perpendicular to an element of the pitch cone will have
a radius equal to the back cone radius, shown in Figure 11-27. It is then a
relatively simple matter to derive equation (11-48) for determining the
formative number of teeth.
N
No == and “Ni == (11-48)
cos
where N; = formative number of teeth; N, = actual number of teeth;
I’ = pitch cone angle of gear, and y = pitch cone angle of pinion.

SECTION 11-20

Strength of Bevel Gears by Modified

Lewis Equation
The geometric shape of bevel gear teeth results in several problems when
attempting to derive an equation for the beam strength. As can be seen in
Figure 11-31, the load acting on a bevel gear tooth varies linearly along the
face of the tooth. It should also be clear from the sketch that the tooth
 643
Section 11-20: Strength of Bevel Gears by Modified Lewis Equation

\.asot

oe
Figure 11-31 The linear variation in tooth load and tooth thickness for
bevel gears.

thickness also varies linearly along the face of the gear. This variation, in turn,
results in linear variations of circular and diametral pitch. Because of these
variations, it is necessary when applying the Lewis equation to bevel gear
teeth to consider an element dx along the face of the tooth, for which these
variables can be considered constant. Figure 11-32 shows the element dx,
as well as the other terms we will use in deriving the strength equation. In the
figure, L is the cone distance, b is the face width, and r is the pitch radius.
The Lewis equation is now applied to the element, dx.
Yd
ue — (11-49)
x

Figure 11-32 The linear variation of the pitch and pitch radius.
 644
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

where dF, = incremental allowable bending force acting on dx; Y = Lewis

form factor; and P, = diametral pitch of tooth at distance x from apex of
pitch cone.
Multiplying both sides by r, and integrating we get
td
[ran = [= (11-50)
But the { r, dF, is equal to the total torque T transmitted by the gear.
Consideration of Figure 11-32 indicates that the circular pitch decreases
linearly as x decreases. Since the diametral pitch varies inversely as the
circular pitch, it follows that the diametral pitch varies inversely with x.
Therefore, because the largest value of circular pitch occurs at x = L, the
largest value of P, will occur at x = L — b.
This fact can be written as follows:

where P = diametral pitch at x = L.

It will also be clear from an inspection of Figure 11-32 that the following
relationship holds because of the similar triangles involved.

Substituting these relationships into equation (11-50) gives us

SY
SY((r/L)xdx
SPE Vee

and since dx varies from x = (L — b) to x = L we get

Se ie
T= a | x? dx
odBola ie
_ SYr (=)
Pisa ee
ye [= Sh = *|
Pre 3
eS [= — L? + 3L*b — 3Lb? + |
PI? 3
SYrb De abe
P= P f-—+—,
2+ aa 2
(11-51)
Before continuing with our derivation, it would be well to consider
limitations upon the values of maximum face width. If the face width is made
too large, the inevitable deflections that occur in a bevel gear installation will
 645
Section 11-22: Allowable Wear Load for Bevel Gears

Cause a greater concentration of load at the small end of the tooth. Because
this is the weakest part of the tooth, it is obvious that failure will occur
much more readily than if the face width had been kept smaller.
Most designers limit the maximum value of the face width to one third
the cone distance. In other words,

DS =
3
It is clear, therefore, that the maximum value of b?/3L? = 1/27. A con-
sideration of the terms in the parenthesis of equation (11-51) indicates that
b?/3L? is small enough compared to the other terms so that it may be
neglected. If we also divide both sides by r, we get the equation for the
allowable bending load,
SYb b
yep |eS =(|
Fo S| (11-52)
11-52

where Y is the Lewis form factor as determined by using the formative

number of teeth.

SECTION 11-21
Dynamic Load for Bevel Gears
The dynamic load equations given for spur gears, equations (10-38)
through (10-40), may be used to obtain reasonable values for the dynamic
loads on bevel gears. The pitch line velocity V, to be used in these equations
is obtained by using the mean pitch diameter.
This approach then requires that F, > F,.

SECTION 11-22
Allowable Wear Load for Bevel Gears
The allowable wear equation is based upon the Hertz contact stress
equation and the Buckingham application of it to gears. Equation (11-53)
may be used to estimate the allowable wear load.
d,KQ’
jy, HRS (11-53)
cosy
where d, = pitch diameter measured at back of tooth.

2Ni.
2 Ne FING
(N;, and N;, refer to formative number of teeth.) The other terms in the
equation are the same as previously discussed. Again it is required that

inceBs
 646
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

It is suggested that the designer use this method involving F,, F,,, and F,
only in the preliminary design stage and then finalize the design by the AGMA
method to be discussed in the next section.

SECTION 11-23
AGMA Method for Designing Bevel Gears
The recommended procedure for analyzing bevel gears is that presented by
the AGMA.” The equations are similar to those presented in the discussion
of spur gears, with some of the values of the correction factors applicable to
bevel gears only.
Equation (10-25) is repeated here for convenience

ar == i
ay,
A A
(10-25)

o, = calculated stress at root of tooth, psi

F, = transmitted load, lb, based upon the large end pitch diameter
= overload correction factor; Table 10-4 should be used
= diametral pitch at large end of tooth
K, = size correction factor, Figure 11-33 gives the values of the size factor
for bevel gears
K,, = load distribution correction. When the deflections or displacements
of gear and pinion are known Figure 11-34 can be used ;when this
information is not available, Table 11-3 should be used. The smaller
values are for those cases where the mountings are extremely rigid,
whereas the larger values are to be used if
the rigidity is questionable
K,, = dynamic factor, Figure 10-21 is to be used
b = face width, in.
J = geometry factor, Figure 11-35 gives the values for the geometry
factor for a 25-deg straight bevel gear. Charts giving values for many
other combinations ofpressure and shaft angle can be obtained from
the Gleason Works, Rochester, N.Y.

It is now possible to calculate the allowable bending stress by the AGMA

formula. The AGMA maximum allowable design stress discussed in Chapter
10 is here repeated.

(10-26)

? Reference and acknowledgement cited in Section 11-6.

 WNWIxXe|)anijejas Uawase|dsIp
Wosyay} uoljisod
JO} YyoIyMseab 39e}U0d
sey uaaq paysiqeysa az1S ‘4039e4
Sy
Jayyte) jea!qan
40 Buoje auod (juawa}a

AA
COO § |

Figure 11-34
SEeSoNNNea
(METS
=)
DS 2
teSSSaneNNWwee), V7
Nn: SELES 2) Z,
Genii
CCCCCCCCONNAEC,
2 SSRs es
i
e
oo
o EDGES 1) Va
; — /
= ae try beas
ae.
Witte
Ga MNINGG
be
CVC. 5

Load distribution factor, K,,

eaca 2

Load distribution factor for bevel gears. [Courtesy AGMA. ]

647
 648
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Table 11-3 Bevel Gear Load Distribution Factors, K_,

Application Both Members One Member Neither Member

Straddle Mounted Straddle Mounted Straddle Mounted

General industrial 1.00-1.10 1.10-1.25 1.25-1.40

Automotive 1.00-1.10 1.10-1.25

Aircraft 1.00-1.25 1.10-1.40 1.25-1.50

SOURCE: AGMA,
NOTE: Because of the smaller bearing loads for straddle mountings, the overall designs are sometimes smaller
and lighter for equal rigidity. Frequently, space limitations dictate the type of mounting for a particular
installation. Usually one member of the pair can be straddle mounted, but it is not always feasible to
use a straddle mounting for both members. An overhung mounting may be just as good as, or better
than, a straddle mounting arrangement. However, rigid supports as well as adequate capacity are re-
quired for both types of mountings,

where

Sq = Maximum allowable design stress, psi

Sat allowable stress for the material, psi; Table 10-7 or Figure 10-24 can
be used
K, = life factor, Table 10-8 may be used; see paragraph 2, p. 561.
K, = temperature factor, same values suggested in Chapter 10 are to be
used
Kr = factor of safety (reliability factor), Tables 10-9 and 10-10 are to be
used.
Number of teeth in mate

100
20 30. 4050 100

90
90

80
80

70
70

60

50

40

30

gear 20
desired
is
factor
of
Number
which
for
in
teeth
geometry

10
0.200 0.240 0.280 0.320 0.360
Geometry factor, J

Figure 11-35 Geometry factor for coniflex straight bevel gears with 25 deg
pressure angle and 90 deg shaft angle. [Courtesy Gleason Works, Rochester,
N.Y.]
 649
Section 11-23: AGMA Method for Designing Bevel Gears

Table 11-4 Elastic Coefficient, C,—Localized Contact (for most

bevel gears)

Gear Material and Modulus of Elasticity

Pinion Material

and Modulus of Tin Bronze

Elasticity, £
30 X 108 19 X 106 17.5X 106

30 X 106 2800 2450 2400

Cast iron 19 X 108 2450 2250 2200

Aluminum bronze 17.5 X 106 2400 2200 2150

Tin bronze 16X 106 2350 2150 2100

SOURCE: AGMA
Poisson’s ratio = 0.30
NOTE: When more exact values ofE are obtained from roller contact test, they can be used.

As was noted in our discussion of spur gears, the calculated stress g,,
equation (10-25), must always be less than or equal to the maximum allow-
able design stress as determined by equation (10-26).
The AGMA fundamental wear equation is repeated here.

= G i LesO CAiCe
peas GaG.
Michi 2

ON SERFS ES (ee)
where

a, = calculated contact stress number

C, = coefficient depending on elastic properties of materials, equation
(10-34) may be used to obtain values for C, or Table 11-4 may be
used
F, = transmitted tangential load, lb, based upon the large end pitch
diameter
C, =overload factor, Table 10-4 may be used
C,, =dynamic factor, Figure 10-27 may be used
d =pinion pitch diameter at large end, in.
b =net face width of the narrowest of the mating gears, in.
C, =size factor, discussion given for spur gears applies, except that Figure
11-33 should be used
C,, = load distribution factor, Figure 11-36 is used when deflections are
known, and Table 11-3 is used when they are not known (K,,)
I = geometry factor, Figure 11-37 gives values for the geometry factor
for a bevel gear having @ = 25 deg, and & = 90 deg. More complete
charts are available from the Gleason Works, Rochester, N.Y.
C, = surface condition factor, same discussion and values given for spur
gears apply.
 650
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

fa PEELE
0.028
i | ew ee ng aL lal
oe
a

el
0.024

\ es
0.020

0.016 fallal
ai a AN
0.012

rua
aeid Ne
Bee eae
ua ‘aia geaqc0
BNE
Zeb
Zanes
fromhas
which
for along
or
cone
contact
position
Maximum
displacement
the
relative
Are et eeeT ff | cointee
gearelement)
vertical
(either
established
been

pee a i sw |
sot a ee eer
Bebe 1.0
ial
1.2 1.4 1.6 0 2 2.8 3.0 3.2
Load distribution factor, C,,

Figure 11-36 Load distribution factor for bevel gears. [Courtesy AGMA. ]

Number of teeth in gear

50 60 70 80 90 100
50

45

40
40

35

30
30

of
Number
pinion
in
teeth 25

20
20

15

10
0.020 0.040 0.060 0.080 0.100 0.120 0.140
Geometry factor, /

Figure 11-37 Geometry factor for coniflex straight bevel gears with 25 deg
pressure angle and 90 deg shaft angle. [Courtesy Gleason Works, Rochester,
N.Y.]
 651
Section 11-24: Tooth Loads on Straight Bevel Gears

The calculated stress number o, can now be evaluated and must satisfy the
equation

C,C
an Ry ee (10-35)
CrCrR

Sac = allowable contact stress number, the values in Table 10-14 may be
used
C, = life factor, Figure 10-33 can be used
Cy = hardness ratio factor, Table 10-15 or Figure 10-34 may be used
C;, = temperature factor, same values suggested for spur gears should be
used
Cr =I factor of safety, Table 10-16 may be used.

Therefore, if equation (10-35) is satisfied, the bevel gears will be safe in wear.
The maximum allowable transmitted horsepower (based on wear) is given
by equation (10-37)
bIC SC, Cae
P= Ny v |ac C, | (10-37)
<2 i 26 WOUGLGRCHOENCS CFOs

SECTION 11-24
Tooth Loads on Straight Bevel Gears
The usual assumption made in determining the loads acting on bevel gears
is that the resultant tooth load acts at the midpoint of the tooth face. The
inherent error in this assumption is so slight that it may safely be made.
Figure 11-38 shows the normal tooth load and the three components into
which it is usually resolved. It is clear from the diagram that a thrust load is
developed in bevel gears.

Example 11-6. A pair of bevel gears is to be used to transmit 12 hp from a pinion

rotating at 360 rpm to a gear mounted on a shaft which intersects the pinion
shaft at an angle of 70 deg. Assume the pinion is to have an outside pitch
diameter of 8 in., a pressure angle of 20 deg, a face width of 1} in., and the gear
shaft is to rotate at 120 rpm, Determine (a) the pitch angles for the gears, (b) the
forces on the gear, and (c) the torque produced about the shaft axis.
 652
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

Solution: Part a.

d, = 3d,
= 3 x 8 = 24in.
Therefore
ry = 8/2 = 4in. r, = 24/2 = 12 in.
2, = Maes
sin X sin 70
ar = = 1138
'p
+ + cos 70
+ cosx
N tg
I = 54.3 deg
pet S70 343= 1ST deg
14. ,
Li Res ede
r sin P= 12 ~ sin 54.3 = 11.4 in.

S. ee 7 = 38i
linear
Spey te = 3.8 1n.

Part b.
hp x 33,000 12 x 33,000 x 12
t
= Sosille
a mdn ~ at x 7.6 x 360
{OR
Pemes ReOb POLS) Orr
LeaCa: @ cos20 0.94
Finrust = F, Sing sin = F tan ¢@sinT
= 553 x tan 20 x sin 54.3
= 553 x 0.364 x 0.812 = 163 1b
F = F,sing@cosT = F, tangcosT
= 553 x tan 20 cos 54.3
— 5980 3645-a 0847 —wlil26)1b
Part c.

d 22.8
torque = F, x > = 553) % Sop ai 6304 in.-lb @

SECTION 11-25

ZEROL Bevel Gears

A ZEROL bevel gear is somewhat like a straight bevel gear, except that
the teeth are curved. They can also be thought of as being spiral bevel gears
whose spiral angle is zero. The big advantage ZEROL gears have, compared
 653
Section 11-25: ZEROL Bevel Gears

nee BS
pride -
F thrust Sry
G ?
_ oral

Figure 11-38 Straight bevel gear tooth loads. [Courtesy New Departure—Hyatt
Bearings Division, General Motors Corporation.]

to straight bevel gears, is that they have localized tooth contact. The advan-
tages associated with coniflex gears thus apply also to ZEROL bevel gears.
In addition since the thrust loads are the same as for straight bevel gears, they
may be used in the same mountings.
The applications to which ZEROL gears are applied are very much the
same as those for straight bevel gears. The minimum suggested number of
teeth is 14, one or more teeth should always be in contact, and the basic
pressure angle is 20 deg, although angles of 225 or 25 deg are sometimes used
to eliminate undercutting.
 654
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

SECTION 11-26

Spiral Bevel Gears

Spiral bevel gears have curved oblique teeth, which allow contact to
develop gradually and smoothly. They are, therefore, extremely useful for
high speed applications and for problems where noise and vibration reduction
are important. An analogy for the relationship that exists between straight
and spiral bevel gears is that existing between spur and helical gears. Localized
tooth contact is also achieved, which means that some mounting and load
deflections can occur without a resultant load concentration at the ends of
the teeth.
The standard pressure angle is 20 deg, although 144-deg and 16-deg
angles are used. The usual spiral angle is 35 deg. Because spiral gears are
much stronger than similar sized straight or ZEROL gears, they can be used
for large speed reduction ratio applications at a reduced overall installation
size.
Figure 11-39 shows the normal tooth load, and the components it is usually
resolved into.

—No. Ill vl
WS(syia

F thrust G Si
At Ce el pea

E,P \\\ tte ea Ws Rowe al

ows
a
ee AZ F thrust G II
Brg. No. |V

Figure 11-39 Spiral bevel gear tooth loads. [Courtesy New Departure—Hyatt
Bearing Division, General Motors Corporation.]
 655
~ Section 11-27: Hypoid Gears

SECTION 11-27

Hypoid Gears
Hypoid gears are very similar to spiral bevel gears. The main difference is
that their pitch surfaces are hyperboloids rather than cones. As a result, their
pitch axes do not intersect, the pinion axis being above or below the gear
axis.
In general hypoid gears are most desirable for those applications involving
large speed reduction ratios, those having nonintersecting shafts, and those
requiring great smoothness and quietness of operation. Hypoid gears are
almost universally used for automotive applications.
The pressure angles usually range between 19 and 223 deg. The minimum
number of teeth suggested is eight for speed ratios greater than 6 to 1, and
six for smaller ratios.
Figure 11-40 shows the tooth load situation for hypoid gears.

F thrust

Figure 11-40 Hypoid gear tooth loads. [Courtesy New Departure —Hyatt
Bearing Division, General Motors Corporation. ]
 656
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

PROBLEMS
1. A left handed helical gear is to transmit 30 hp at 900 rpm. The gear has 42 teeth, a
30-deg helix angle, a 20-deg transverse pressure angle, and is rotating in the clock-
wise direction. Draw a sketch of the gear, showing all the forces acting. Let P be 7.
2. A left handed helical gear having 20 teeth is driven by a motor at 2000 rpm in the
clockwise direction. The gear has a normal pressure angle of 20 deg, a helix angle
of 30 deg, a normal diametral pitch of 10, and a normal force of 80 lb. The gear
meshes with another helical gear having 50 teeth. Draw a sketch of the second gear
showing all the forces acting, and calculate the torque acting.
3. In the Figure Problem 3, gear | has 25 teeth, is left handed, has a normal diametral
pitch of 5, is to transmit 15 hp at 1400 rpm clockwise, has a @ of 20 deg, and a w
of 30 deg. Gear 2 has 40 teeth, and the velocity ratio for gears 1 and 3 Is to be 0.5.
What are the forces acting on each gear? What is the torque on each shaft?

Problem Figure 3

“4. For the reverted gear train shown in Problem Figure 4 the input and output shafts
have the same center line and 225 hp is to be transmitted. ¢, for all gears is to be
20 deg.

Problem Figure 4
 657
Problems

Complete the following table, specifying the helix angles so that there is no net
thrust on shaft 2.

Tangential Radial
rpm No. of Teeth | Helix Angle and Hand} d, in. Bonen: Rerees

“Specify load applied to gear and direction up, down, right, left, in, out.

. A pair of meshing helical gears are mounted on parallel shafts 10 in. apart. Their
normal diametral pitch is 6. If the pinion has 30 teeth and the speed ratio is 0.5,
what is the required helix angle?
. Two meshing helical gears of the same hand are mounted on shafts that are 90 deg
apart. What is the center distance if the pinion has 30 teeth, a helix angle of 30 deg,
and a normal diametral pitch of 4? The velocity ratio is to be 0.4.
. Two meshing helical gears are both made of SAE 1020 case hardened and WQT
steel, and are mounted on parallel shafts 10 in. apart. The gears are to have a
velocity ratio of 0.33, are 20 deg full depth with 35 deg helix angles. The diametral
pitch is 20 and the face width of the gears may be taken as 15 in. What is the maxi-
mum horsepower that can safely be transmitted? Use the modified Lewis equation
for strength and the Buckingham equation for wear. The pinion speed is equal to
2000 rpm.
. Repeat Problem 7 using the AGMA method and the assumptions given in the
preface to the problems of Chapter 10.
. The pinion of a pair of helical gears mounted on parallel shafts is to transmit 5 hp
at 8000 rpm. Both gears have 30 deg helix angles, transverse pressure angles of
20 deg, are full depth, are made of SAE 1030 steel and have a face width of 5 in. If
the pinion is to have 25 teeth and the gear 100 teeth, what are the minimum
diameters, and what is the required BHN?
10. Check the gears of Problem 9 by the AGMA method and use the same assumptions
as in Problem 8.
11. A 24-tooth pinion, 6 diametral pitch, is attached to a drive shaft that is rotating at
2400 rpm. This pinion meshes with a spur gear mounted on a shaft that is 6 in.
from the drive shaft and parallel to it. Also, on this shaft is a triple threaded, right
hand worm that drives a 75-tooth worm wheel. On the same shaft with the worm
wheel is an 8-in. hoisting cylinder. The cylinder is to lift a weight by means of a
cable. How long will it take for the weight to be lifted through a distance of 30 ft?
12. Two shafts at right angles, with center distance of 7 in., have to be connected by
worm gearing. They have a velocity ratio of 0.025, lead angle of worm of 42 deg
and a normal pitch of0.4 in. for the worm wheel. (a) Determine the pitch diameter,
lead, and number of teeth of the worm. (b) Determine the number of teeth and
pitch diameter of the worm wheel.
13. A triple threaded worm having a 23 in. diameter receives 30 hp at 1500 rpm. The
worm wheel has 100 teeth and is 12 in. in diameter. What is the tangential force
on the wheel teeth if the efficiency is 80%?
 658
Chapter 11: Helical, Worm, Bevel, and Other Gear Types

14. A 2-tooth worm that has an input of 630 hp at 1800 rpm drives a 50-tooth wheel.
What is the output torque for 80% efficiency?
15. A hoisting system has a drum 30 in. diameter with a 1-in. diameter wire, mounted
on the same shaft with a worm wheel having 90 teeth. A triple threaded worm is to
drive the wheel and raise a 5000 Ib load at 80 ft/min. If the system has an efficiency
of 70%, how much horsepower must be supplied to the worm gear shaft?
\
16. For the gears of Problem 12, if the worm is transmitting 15 hp at 1500 rpm and the
worm gear has a 25 deg normal pressure angle, determine the forces acting on the
worm and on the worm wheel.

17. If the worm in Problem 14 has a pitch diameter of 3 in. and the worm gear has a
circular pitch of 0.8 in. and 25-deg full depth teeth, determine the forces acting on
the worm and the worm wheel.
18. A worm gearset is to transmit power between two shafts that are to be approxi-
mately 8 in. apart. If the velocity ratio is to be 0.05 determine the number of teeth
.in the gears and the required pitch diameters. Assume AGMA recommendations
have been followed.

19. A triple threaded worm having a 23 in. pitch diameter meshes with a worm wheel
having a diametral pitch of 5 and 60 teeth. Determine (a) lead, (b) lead angle, and
(c) center distance.
20. What should be the numbers of teeth for a worm gearset having a velocity ratio of
0.05, and a center distance of approximately 5 in.? The worm lead angle is 15 deg
and the normal diametral pitch is 6.
21. Design a worm gearset to transmit 20 hp at a worm rpm of 1500 and a velocity
ratio of 0.04.
22. A quadruple threaded cast steel worm (0.2% C WQT) is to be used with a bronze
worm wheel to give a velocity ratio of 0.04: 20 hp is to be transmitted at 1000 rpm
worm speed. Determine the pitch and face width based on strength and wear if the
normal pressure angle is 20 deg. Will special cooling be needed?
23. Design a speed reducer to transmit 2 hp at 1800 rpm of worm. The bronze worm
gear 1s to be 20-deg normal pressure angle and a speed ratio of 0.025 is desired.
The worm is to be cast steel.
24. Design a speed reducer to transmit 20 hp at 1250 rpm of the worm and to have
a velocity ratio of 0.06.
25. A gear train consists of a triple threaded worm that drives a worm wheel having
75 teeth. A bevel gear is mounted on the same shaft as the worm wheel. The
meshing driven bevel gear has 75 teeth and the velocity ratio for the gears is to be
0.4. How many revolutions will the worm make for two revolutions of the larger
bevel gear? If the worm has an input of 3 hp at 1500 rpm and the efficiency of the
gear train is 90%, what will be the output torque of the gear train?
26. A pair of bevel gears are mounted on shafts that are 90 deg apart. The set is to trans-
mit 10 hp at 300rpm. The pinion has a 6-in. outside pitch diameter, 2-in. face
width, a diametral pitch of 6, and is 20 deg full depth. For a velocity ratio of 0.4,
draw a free body diagram ofthe gears showing all forces acting. What is the torque
produced about the gear axis?
27. Repeat Problem 26 for a 60-deg shaft angle.
28. Ifthe gears of Problem 26 are made of cast steel, WQT, will they be satisfactory from
the strength and wear viewpoint? Use the Lewis and Buckingham equations.
 659
References

29. Repeat Problem 28 using the AGMA method. Assume one gear is straddle mounted,
the gears are to be used for a general industrial application, power source has light
shock, driver machine is subject to moderate shock, a life of 1 x 10° cycles, and
high reliability is required.
30. Repeat Problem 28 for the gears of Problem 27.
31. A pair of meshing steel bevel gears are mounted on shafts 90 deg apart. Both are
20 deg full depth, have a diametral pitch of 5, a strength of 20,000 psi, BHN of 300,
and a face width of 25 in. If the pinion has 30 teeth and is turning at 1500 rpm, what
horsepower can the gears transmit if the velocity ratio is to be 0.5? Draw a free
body diagram for the pinion and its shaft, showing all forces and moments acting.
It may be assumed that the apex of the pitch cone is at the center ofa 20-in. long,
simply supported shaft. The left hand bearing only is able to resist a thrust load.
Use the Lewis and Buckingham equations.
32. Repeat Problem 31 using the AGMA method. Assume the conditions of Problem
29)

REFERENCES
] E. Buckingham and Ryffel: Design of Worm and Spiral Gears. The Industrial Press.
[2] V. M. Faires: Design of Machine Elements. The Macmillan Company, New York,
1965.
| Wwey E. Buckingham: Analytical Mechanics of Gears. McGraw-Hill Book Co., New York,
1949.
Spiroid Gearing. Illinois Tool Works, Chicago, IIl., 1966.
gS D. L. Seager: Dynamic behavior of helical gears. ASME 69-VIBR-16.
4)
How to Test Bevel Gears. Gleason Works, Rochester, N.Y., 1955.
Bevel and Hypoid Gear Design. Gleason Works, Rochester, N.Y., 1956.
Belt and Chain Drives
SYMBOLS

c = center distance, in. In = naperian logarithm

e = base of naperian logarithms n = speed, rpm
f = coefficient of friction N, = number of sprocket teeth
F, = maximum belt tension, Ib p = pitch, in.
F, = minimum belt tension, lb T = torque. in.-lb
F, = inertia force on belt, Ib w’ = belt weight, Ib/in.
g = acceleration of gravity, in./sec? B = half of included angle of V-belt
H = transmitted horsepower, hp 0, = angle of contact, rad
L = pitch length of belt, in. @ = angular velocity, rad/sec

Belt and chain drives offer a maximum of versatility as power transmission

elements. They allow the designer considerable flexibility in location of
driver and driven machinery, and as is the case
with gear drives. Another advantage of chain drives, and particularly of belt
drives, is that they reduce vibration and shock transmission. Furthermore,
belt drives are
Large flat leather belts were in general use many years ago, when the usual
practice was to use one large engine to drive several different pieces of
machinery. Some leather belts are in use at this time as well as flat steel,
rubber, plastic, and fabric belts. Light, thin flat belts are practical on high
speed machinery where vibration may be a problem.
A smooth surfaced flat belt must operate with higher tensions to transmit
the same torque as a V belt. Positive-drive type flat belts, commonly known
as timing belts (Figure 12-1), have evenly spaced teeth on the inner surface
and operate on toothed pulleys. The speed ratio isinversely proportional to_
the number of pulley teeth. Required belt tensions are low and, consequently,
bearing loads are reduced.
V belts are probably the most common means of transmitting power
between fractional horsepower electric motors and driven machinery.
V belts are also used in other household, automotive and industrial applica-
tions. Multiple-belt drives with capacities up to a few hundred horsepower
660
 Figure 12-1 Timing® belt drive (Licensed TM).
Drives of this type are designed to transmit fractional
horsepower up to 600 hp or more. Belt speeds range
up to 16,000 ft/min. [Courtesy Dodge Manufacturing
Division, Reliance Electric Co.]

Figure 12-2 Adjustable speed belt drive. If the center distance between shafts is
increased, the sides of the drive pulley separate, decreasing its effective radius.
Drives of this type are designed to transmit from } to 20 hp at 1750 rpm. [Courtesy
Dodge Manufacturing Division, Reliance Electric Co.]

661
 662
Chapter 12: Belt and Chain Drives

Figure 12-3 Variable speed chain drive. [Courtesy Link-Belt Enclosed Drive
Division, FMC Corporation. ]

are available for industrial use. Conventional V belts are made of rubber
covered with rubber impregnated fabric and reinforced with nylon, dacron,
rayon, glass fiber, or steel tensile cords. Automotive and agricultural belt
drives are generally designed for 750-1000 hr average life, whereas industrial
belt drive design is based on longer service. Most often, both driver and
driven pulleys lie in the same vertical plane, and the speed ratio is constant.
Quarter-turn drives are used to transmit power between horizontal and
vertical shafts, using deep groove pulleys and relatively long center distances.
Smooth flat belts and V belts depend on friction for traction on the pulleys,
and some slippage is inherent in their operation. Therefore, speed ratios are
not precise. Variable-pitch pulleys are used to change the input to output
speed ratio ofa V belt drive (see Figure 12-2). Some variable-pitch drives will
change speed ratios when the belt is transmitting power. Most heavy duty
pulleys (sheaves) are made of cast iron or formed steel.
Chain drives generally have far greater life expectancies than belt drives,
up to 20,000 hr with lubrication and regular maintainance. For heavy loads,
 663
Section 12-1: Belt Drives

roller chain and inverted tooth chain are used on toothed sprockets. Roller
chain, available in single and multiple strand form, is used at speeds up to
2500 ft/min. Inverted tooth chain is used at speeds up to 4000 ft/min. Both
are available with capacities over 1000 hp. Figure 12-3 shows a variable
speed chain drive.

SECTION 12-1
Belt Drives

Flat belt drives

Consider a flat belt drive operating at its maximum capacity. Taking an
element ofthe belt as shown in Figure 12-4, the (enon force is limited by the
normal force dN (the pulley reaction) and the coefficient of frictionf;The
local belt tension is F. In the direction of a tangent through the center ofthe
element, equilibrium of forces yields
(F + dF) cos (d0/2) — fdN — F cos (d6/2) = 0 (12-1)
Simplifying, and noting that for small angles the cosine approaches unity, we
obtain
dF =fdN (12-2)

(a) Forces acting on an elementary

section of flat belt on the verge
slipping .

FE-AdF dN

do

Angle of wrap
and belt tensions
Driver
at the smaller
pulley.

Figure 12-4 Flat belt drive.

 664
Chapter 12: Belt and Chain Drives

For radius R measured to the belt center and belt weight w’ Ib/in. of length,
the length of the elementary section of belt is R dO and its mass is

eRe
g
For angular velocity «, the normal acceleration is w?R and the inertia force
F.d@ where we define
, 2 R2

k=—— 2 (12-3)
Equilibrium of forces in the radial direction yields
dN + F.d@ — (F + dF + F)sin(d6/2) = 0 (12-4)
We first observe that for small angles the sine of the angle approaches the
angle itself. Next, the higher order term dF d@ may be neglected. Finally,
using equation (12-2) to eliminate dN from equation (12-4) and separating
the variables, we obtain
dF
F — F,
= f do (12-5)
Referring to Figure 12-4b, the solution of equation (12-5) is obtained by
integrating from minimum tension F, to maximum tension F, through the
angle of contact of the belt, 9 = 0 to 0, rad. Thus, we have

1
n (=)
F, —F
-
—po,c a rf)
(12-6)
ae

or

oe
Ee fo
et 14

for the belt operating at maximum capacity. Note that the smaller angle of

V-belt drives
For V belts, the inertia effect F. is given by equation (12-3) where R is the
pitch radius of the pulley, somewhat less than the outside radius. The weight
density of a typical V belt is approximately 0.05 lb/in.? The normal force is
applied to the sides of the belt since the belt rides against the pulley sides
(Figure 12-5). Let the normal force on each side be dN’/2 producing a total
friction force fdN’ on an element of V belt. The radial force dN in equation
(12-4) is replaced by dN’sin# for pulley included angle 28. Making the
indicated substitutions, we see that the coefficient of friction
f in equations
 665
Section 12-1: Belt Drives

Figure 12-5 V belt.

(12-6) and (12-7) may be replaced by f/sin B. The V-belt tension relationships
are then given by

ry = E; ) f 9,

gy (7= 4 ~ sin B tees

or

eae
a of Oi/sin p (12-9)
:

Transmitted power
For belt drives, the torque on a pulley is given by

T=(F, — F)R (12-10)

where pitch radius R is MSU from the center of a pulley to the neutral
axis of the belt. The transmitted horsepowe

(12-11)
~ 63,025
where n is the speed of a given pulley in revolutions per minute and T is the
torque on the same pulley in inch-pounds. Most belt drives are designed to
produce a speed reduction. ea rotation speeds are inversely proportional
to pulley radii. Belt strength limits maximum ion F, and the minimum
tension F, depends on the tae of contact, the coefficient of friction, and
whether we are using a flat belt or a V belt.

Belt drive design

Since V belts vary considerably in cross section and amount of reinforce-
ment, the design of V-belt drives is usually based on tables given by the
manufacturers. However, the following alternate method may be used. Let
the input and output speeds and horsepower of a belt drive be given. Suppose
 666
Chapter 12: Belt and Chain Drives

we are to design the drive. Using equation (12-11), we may find torque at the
smaller pulley as a function ofits revolutions per minute, n,

63,0230»
1 ad n.-lb (12-12)
ny

Radius R, of the smaller pulley may be tentatively selected. The radius of the
larger pulley, R, is found on the basis of the ratio of pulley speeds. We use

Kee
tees= any (12-13 )

where n, and n, are, respectively, speeds of the smaller and larger pulleys. If
we estimate belt weight, the inertia effect F. is given by equation (12-3). In
order to reduce the problem to finding one unknown tension, we rewrite
equation (12-10) in the form
G
P= be 12-14

When equation (12-14) is used to eliminate F, from equation (12-9), the

results may be rearranged to give maximum tension

ye T,
FF-=
1 EFc od
ae Fee |R, (12-15 )

where
y = efaiising (12-16)
When it is necessary to estimate V-belt tensions, a ratio between 3 to 1 and
5 to 1 is reasonable. For example, we may estimate

from which

T IP
F,= 1255 and F, = 0.255

If a flat belt is used, sin B = 1. For standard V belts, 6 ranges from 17 to

19 deg, producing a greater friction effect. Thus,

. We may obtain the approximate value of the smaller angle of

contact 8, by sketching the drive as in Figure 12-4b. Unless an idler pulley
is used, 0, will generally be equal or less than z rad. When convenient, the
rotation direction is selected so that the lower section of belt has the max-
imum tension, producing a slightly greater angle of contact. This would be
the case for clockwise rotation if the smaller pulley is the driver in Figure
12-4b.
 Table 12-1 Service Factors for V-Belt Drives

Driven Machine Dri ver

AC Motors: High Torque, High Slip,

The types listed below are representative AC Motors: Normal Torque, Squirrel Cage, Repulsion-Induction, Single
samples only. Select the group listed below Synchronous, Split Phase. Phase, Series Wound, Slip Ring.
_ whose load characteristics most closely ap- DC Motors: Series Wound, Compound
proximate those of the machine being con- DC Motors: Shunt Wound, Wound.
sidered. Engines*: Single Cylinder Internal
Engines*: Multiple Cylinder Internal Combustion.
If idlers are used, add the following to the Combustion. Line Shafts
service factor. Clutches
Idler on slack side (inside the belts)... None
Idler on slack side (outside the belts) . . 0.1 Intermittent Intermittent
Idler on tight side (inside the belts)... 0.1 Normal Continuous Normal Continuous
Service, Serie Servi Service, : :
Idler on tight side (outside the belts) . . 0.2 3-5 hr Ns COIS 3-5 hr Service, Service,
a 8-10 hr 16-24 hr Daily on 8-10 hr 16-24 hr
Daily or ;
Daily
Seasonal Seasonal ay ray

Agitators for liquids

Blowers and exhausters
Centrifugal pumps and compressors
Fans up to 10 hp
Light duty conveyors
a : '
Belt conveyors for sand, grain, etc.
Dough mixers
Fans over 10 hp
Generators
Line shafts
Laundry machinery is)
Machine tools
Punches-presses-shears
Printing machinery
Positive displacement rotary pumps
Revolving and vibrating screens
.
Brick machinery
Bucket elevators
Exciters
Piston compressors
Conveyors (drag-pan-screw)
Hammer mills 1.4
te
Paper mill beaters
Piston pumps
Positive displacement blowers
Pulverizers
Saw mill and woodworking machinery
Textile machinery
.

Crushers (gyratory-jaw-roll)
Mills (ball-rod-tube)
G8) 1.6 1.8
Hoists
Rubber calenders-extruders-mills

Chokable equipment?
2.0 2.0 2.0
Fire hazard conditions*

SOURCE: Dodge Manufacturing Division, Reliance Electric Co.

* Apply indicated service factor to continuous service engine rating. Deduct 0,2 (with a minimum service factor of 1.0) when
applying to maximum engine rating.
tWhere fire hazards are prevalent and fire prevention laws apply, it is recommended that drives be designed using a service
factor of 2.0 on the HP rating of the motor.

667
 668
Chapter 12: Belt and Chain Drives

V belts are generally made from reinforced rubber and the ae eebelt
strength 1is Baus by maximum tension F;. ye

Typical service factors are given in

Table 12-1.
, but
a much lower value occurs in the presence of oil or grease. If the belt is not
permitted to slip, the static coefficient of friction applies. If the belt is allowed
to slip, that is, if it is used as a clutch, then the coefficient of sliding friction
applies. The latter value is somewhat lower.
In order to solve equation (12-15), we originally estimated belt cross-
section size to compute the inertial effect F.. If we must select a belt much
larger than our approximation, the above steps may be repeated to find new
values of F. and F, and a new belt selection. Minimum tensionnF, is eiey by
equation (12-14). M aft | by is D' n

When a single V sor is inadequate, a multiple V belt drive may be specified.

Standard pulleys are available with 1, 2, 3, 4, 5, 6, 8, and 10 grooves in a wide
range of pitch diameters. Figure 12-6 shows a multiple V belt drive.

Figure 12-6 Multiple V belt drive. [Courtesy Dodge

Manufacturing Division, Reliance Electric Co.]
 669
Section 12-1: Belt Drives

Belt length, center distance, and contact angle

For a belt drive employing two pulleys of the same radius R, with center
distance c, the theoretical length of the belt is L = 2(c + 2R,) and the contact
angle is 0, = mrad. If pulley pitch radius is used, the result will be the pitch
length of the belt, measured along its neutral axis. Note, however, that belt
length and pulley radius must be adjusted when ordering because belts are
identified by outside circumference and V-belt pulley sizes are based on
outside diameter.
For unequal pulley pitch radii R, and R,, referring to Figure 12-7, we see
that the angle « is given by

and that the angle of contact

0, =x — 2orad (12-18)
on the small pulley.
The length of a straight section of belt
a =[c? — (R, — R,)7]"? in. (12-19)
and the total pitch length of the belt is
L=2a+ R,(x — 2a) + R,(a + 2a) in. (12-20)
Approximating the dimension a by two terms of a binomial expansion and
substituting sin « for a, we obtain

(R3 = R,)
L = 2c + a(R, -+ R,) + re are (12-21)

Figure 12-7 Belt length, center distance, and contact angle.

 670
Chapter 12: Belt and Chain Drives

from which center distance

bP bP ORES Ry) (12-22)

Cy
4

where b = L — n(R, + R,).

Sometimes the center distance c between pulleys is governed by the size of
the machinery involved or restricted in other ways. If not, a reasonable value
O

Se =3R,+R, or c=2R, (12-23)

where R, is the pitch radius of the larger pulley.
This value of c is used in equation (12-21) to find L. If a standard belt is
selected, the actual pitch length is substituted into equation (12-22) to find
center distance. The manufacturer’s specifications should be consulted to
relate pitch length and pitch diameter to nominal sizes of belts and pulleys.
Wherever possible, belt drives should be designed with provision for center
distance adjustment, unless an idler pulley is employed, because belts tend
to stretch in use. When adjustment is provided, the accuracy of length and
center distance calculations is not critical.

Example 12-1: Design of a Belt Drive. Required. A belt drive to transmit

30 hp when input speed is n, = 1750 rpm and the driven pulley is to rotate at
approximately 1040 rpm. The load is steady, service is continuous, and a high
degree of reliability is required.
Design decisions. One or more V belts will be used for this drive. We note that
the pulley pitch radii R, = 3.7 and R, = 6.2 in. will produce an output speed
ny = 1042 rpm, which 1s satisfactory. These pitch radii correspond to standard
size pulleys with outside diameters of 7.5 and 12.5 in. Suppose we tentatively
select a V-belt cross section that is nominally 3 in. wide by 4 in. deep weighing
0.012 lb/in. of length. We will specify the drive components based on our
calculations rather than using the belt manufacturer’s recommendations.
Solution: In the absence of other criteria, equations (12-23) will be used to
obtain a reasonable center distance. For this problem, the first equation
governs and the distance between pulley centers will be approximately
c = 17.3 in. The pitch length of the belt given by equation (12-21) is about
66 in. for this value ofc.Using the nearest standard size listed in manufacturers’
catalogs, we might specify a belt of 67 in. nominal length, which corresponds
to a pitch length of 66.6in. Using equation (12-22), we recalculate center
distance c = 17.6in. Of course, provision should be made for adjustment for
installation of belts and for belt stretching.
Using equations (12-17) and (12-18) we obtain « = 8.16 deg and the angle
of contact 0, = 163.7 deg = 2.86 rad for the small pulley.
Using equation (12-12) we find driven torque T, = 1080 in.-Ib. Angular
velocity of the drive pulley is 27n,/60 = 183 rad/sec. Then, based on our
tentative belt section, F. = 14.3 1b from equation (12-3). It will be assumed
that the coefficient of friction f will not be less than 0.2 and that the pulley
angle / will be 18 deg. From equations (12-15) and (12-16), we obtain y = 6.36
and maximum tension F, = 361 lb. Using equation (12-14), we find minimum
tension F, = 69 lb.
 671
Section 12-2: Chain Drives

The nature of the loading is complicated because the belt is subject to

varying tension as well as bending as it goes around the pulleys. Furthermore,
a typical V belt is not uniform in section but made of rubber fabless by
synthetic fiber cords. We do know, however, based on tests, that a 3 by 44 in.
V belt can operate satisfactorily over long periods with maximum tensile loads
of 200 lb. If we apply a safety factor of 1.5, because a high degree of reliability
is required, we would have a working maximum tensile load of 133 lb per
belt. Applying a service factor of 1.3 to F, (for continuous service) we obtain
a maximum tensile force of 471 lb applied to the belts. Four belts carried on
multiple pulleys would therefore be satisfactory. If a single large belt is chosen
instead, larger pulleys anda greater center distance would be required, changing
the solution entirely.
Based on our selection of four V belts, the input and output shafts would be
_ subject to a bending load equal to the vector sum F,+ F,= 430 |b (almost
S the same as the scalar sum) applied, roughly, at the center of each multiple
. pulley. The input shaft torque T, = 1080 in.-lb and the output shaft HORS
T, = 63,025H/n,= 1813 in.-Ib. ‘e

SECTION 12-2
Chain Drives
Chains are used for power transmission and as conveyors. They can be
and where precise speed ratios must be maintained.
Although location and alignment tolerances need not be as precise as with
gear drives, the best service can be expected when both input and output
sprockets lie in the same vertical plane.
The ratio of output speed n, to input speed n, is given by

aR Ne
beNy (12-24)
12-24
t2

where N, and N,, represent the number of teeth on the input and output
sprockets respectively.
For a tentative center distance c between shafts, chain length L may be
approximated by equation (12-21) of the preceding section where R, and
R, refer to pitch radii of the input and output sprockets respectively. We
then select an exact chain length for a whole number ofchain links (preferably
an even number) and recompute center distance from equation (12-22) of
the preceding section. The length of an individual link from pin center to
pin center is the pitch p. The pitch radius of a sprocket with N, teeth may
be defined by

_ Nip (12-25)
2n

Angle of contact for the chain is given by equation (12-18) of the preceding
section. For the small pulley, it is desirable to have an angle of contact
not less than 120 deg. For speed ratios n,/n, > 3 this restriction is met if
 672
Chapter 12: Belt and Chain Drives

c = 2(R, — R,). For smaller n,/n, ratios, a practical center distance will
fall between a value that just permits the sprockets to clear and
c = 2(R, + R,). When longer chains are used, it is advisable to investigate
the need for idlers on the slack section of chain.

Inverted tooth (silent) chain

Figure 12-8 shows inverted tooth or silent chain. Chain pitch is defined
as in Figure 12-9. Smaller pitch chain (p = Zin. and 4in.) is commonly
available in widths from 4 to 4 in., whereas 13-in. pitch chain is available in
widths from 3 to 16 in. The chain usually has guide links on the sides or in
the center to keep it on the sprocket. Inverted-tooth chain is made of steel,
and sprockets are made of steel in the smaller sizes and cast iron in the
larger sizes. Inverted tooth chain is typically quieter than roller chain and
may be operated at higher speeds because there is less impact force when
chain links engage the sprocket. Lubrication is generally provided and, at
full load, drive efficiency may be as high as 99%.

Figure 12-8 Inverted tooth (silent)

chain. (a) A silent chain drive. Drives of
this type are possible with speed ratios as
high as 10:1. The guide link engages
grooves cut in the sprocket teeth. Most
silent chain drives are of this type.
[Courtesy Link-Belt Chain Division,
FMC Corporation. ] (b) Side-guide silent
chain. The alternate guide links at the
outer edges of the chain straddle the
sprocket face. [Courtesy Ramsey
Products Corporation, Charlotte, N.C.]
(a)
 673
Section 12-2: Chain Drives

Figure 12-9 Chain pitch (inverted tooth or silent chain).

Roller chain
Roller chain is available in single and multiple strands, as shown in
Figure 12-10. The chain is made up of side plates, pins, bushings, and rollers.
Pitch p is measured from pin center to pin center. Single strand roller chain
for industrial drives is generally available in pitches from
j to 3 in. and multiple
strand chain in pitches from ? to 3 in. Conveyor chain is available in larger
sizes. Roller chain is usually made of hardened steel and sprockets of steel
or cast iron, but stainless steel and bronze chains are available where
corrosion resistance is required. Lubrication is generally provided. However,
if lubrication is undesirable, as in food processing, oil impregnated sintered
metal bushings may be used in place of rollers.

Chordal action
Roller chain and inverted tooth chain are made up of links of finite
length and, as these links engage the sprocket teeth, the chain velocity
varies. Figure 12-11a shows the pitch line of a chain in its highest position,
a distance r, from the sprocket center. Part (b) of the figure shows the pitch
line of the chain in its lowest position, a distance r, from the sprocket center.
If a constant driving speed w rad/sec is maintained, the pitch line velocities
of the chain in its highest and lowest positions are
Ob,a = OF a (12-26)

and
 674
Chapter 12: Belt and Chain Drives

(a) (b)
Figure 12-10 Roller chain. (a) A single-strand roller chain drive. [Courtesy
Link-Belt Chain Division, FMC Corporation.] (b) A double-strand roller chain
drive. [Courtesy Acme Chain Division, North American Rockwell.]

respectively, and the velocity changes cyclically N, times per sprocket

rotation
for N, sprocket teeth. Referring to Figure 12-11a, we see that

(12-27)

(a) (b)
Figure 12-11 Chordal action. (a) Chain in highest position. (b) Chain in lowest
position.
 675
Problems

Expanding the cosine term, the change in height becomes approximately

ofa
(12-29)
GOR
Inverted tooth chain sprockets generally have no less than 17 teeth. The
radius change is then, approximately 6/r, = 1.7% for the worst case. Roller
chain sprockets are commonly available with as few as eight teeth, and the
corresponding radius change is 6/r, = 7.7% for the worst case, corre-
sponding to a velocity change of almost +4°% from the mean. The corre-
sponding accelerations result in severe fatigue loads on chain components.
The effects of chordal action can be alleviated by sprocket and link design.
One design uses an involute form of sprocket tooth for inverted tooth
chain. Another inverted tooth chain design uses two-part link pins that roll
upon one another and effectively change chain pitch by a small amount as
the chain engages the sprocket.

Failure of chain drives

Most industrial roller chains and inverted tooth chains are designed for
service of many thousands of hours if adequate lubrication is maintained.
Engagement and disengagement of the chain with sprocket teeth, however,
results in fatigue loading, particularly when one of the sprockets is small,
and chordal action contributes to the problem. The link plates are subject
to varying tension and, thus, must be designed to prevent tensile fatigue
failure. At low speeds, failure is: more probable in the link plates than
elsewhere, particularly in the case ofroller chain. Roller chain roller bushings
are subject to impact against sprocket teeth and are more susceptible to
failure at high and moderate speeds. At very high speeds and loads, lubrica-
tion breakdown may cause instantaneous seizing or galling at the chain
joints. Current standards, however, allow operation at very high speeds
and loads (the range where galling is possible) if the design calls for an
adequate lubrication system.

PROBLEMS
1. Find maximum torque at the small pulley of a V-belt drive if maximum belt
tension cannot exceed 300 lb. The included angle of the pulley groove is 35 deg, the
coefficient of friction is 0.2, the small pulley has an angle of wrap of 170 deg, a
pitch diameter of 8 in., and a speed of 4000 rpm. The belt weighs 0.01 lb/in.
2. Find the maximum torque at the small pulley in Problem 1 if maximum belt ten-
sion cannot exceed 1500 Ib and the belt weighs 0.05 |b/in.
3. A single V-belt drive with included angle 36 deg is to have a capacity of 16.55 hp
based on a coefficient of friction of 0.2 and estimated belt weight of 0.012 Ib/in.
Speed is to be reduced from 1750 rpm to 1170rpm using a drive pulley pitch
radius of 3.5 in. Shafts are 16 in. apart. (a) Find belt tensions at full load. (b) Find
shaft torques and bending loads. (c) Find initial tension when the drive is not
operating.
 676
Chapter 12: Belt and Chain Drives

4. The capacity of a V-belt drive is to be 10 hp based on a coefficient of friction of 0.2,

a drive pulley pitch radius of 4in., a driven pulley pitch radius of 8 in. and an
input speed of 1725rpm. The shafts are 18 in. apart. Estimated belt weight is
0.012 lb/in., included angle of pulleys is 36 deg. (a) Find required belt tensions.
(b) Find shaft torques and bending loads. (c) Find belt tensions when the above
drive is transmitting only 2 hp. Assume no adjustments have been made.
. A single V belt with a pitch length of 79.6 in. is to transmit 10 hp. Input speed
is 1720 rpm and output speed is 860 rpm. The drive pulley has a pitch radius of
4 in. and the included angle of the pulleys is 36 deg. Assume the belt tension on
the tight side is 5 times the tension on the slack side. (a) Find the center distance
of the shafts. (b) Find belt tensions at full load. (c) Find torque and bending load
on the shafts. (d) Find the centrifugal effect based on a 5 V section belt weighing
0.012 Ib/in. (e) Find the minimum coefficient of friction required.
. Analyze the belt drive of Problem 5 if 5 hp is to be transmitted and output speed
is to be 1720 rpm. Assume belt tension on the tight side is 2.5 times the tension
on the slack side. :
. A thin plastic flat belt is to be designed to transmit 12 hp. Input pulley radius is
5 in., input speed is 3000 rpm and output speed is 1670 rpm. Use a center distance
of 24 in., and a 0.02-in. thick by 2-in. wide belt with a density of 0.1 Ib/in.* Assume
the coefficient of friction is not less than 0.3. (a) Find torque at the small pulley.
(b) Find large pulley radius and contact angle for both pulleys. (c) Find inertia
effect and belt tensions. Find maximum belt stress without including bending
stress.

. Find stress in the belt in Problem 7 if the drive is designed for a capacity of 5 hp.
Find bending and torsion loads on the pulley shafts.
. The output shaft of a speed reducer rotates at 100 rpm and transmits a steady
10 hp to a 5-in. diameter pulley overhung by 1.75 in. Assume minimum belt tension
equals one third maximum tension. Using a corrected endurance limit of 45,000 psi
and a yield point of 105,000 psi, find the required shaft diameter based on the
maximum shear theory. Use a safety factor of 3.6.
10. The countershaft in Problem Figure 10 has two 6-in. diameter flat belt pulleys.
It rotates at 1000 rpm and transmits a steady 50 hp. Assume minimum belt tension

Problem Figure 10
 677
References

equals one half maximum tension. Both belts are approximately vertical. Use a
corrected endurance limit of 40,000 psi and a yield point of 100,000 psi with a
factor of safety of 3 for the shaft. Use a stress concentration factor K, = 1.4 at
the step. Find the required shaft diameters.
11. The countershaft in Problem Figure 10 has two 7-in. diameter V-belt pulleys. It
rotates at 1160 rpm and transmits a steady 7.0 hp. Assume minimum belt tension
equals one third maximum tension. Both V-belts are approximately vertical. Use
a corrected endurance limit of 40,000 psi and a yield point of 100,000 psi with a
factor of safety of 3 for the shaft. Use a stress concentration factor K, = 1.4 at
the step. (a) Show moment and torque loading on the shaft due to the V-belt
drive. (b) Determine diameters D, and D, based on the Maximum Shear Theory—
Soderberg Criterion. (c) Find the required minimum coefficient of friction if the
included angle of the pulleys is 38°, the angle of belt contact 180° and the belt
weight 0.017 Ib/in.
a2: If pulleys are to turn in opposite directions, a belt may be crossed in the form of
a “figure eight.” Find approximate belt length in terms of pulley radii and center
distance.
43. Find the horsepower capacity of a V-belt drive with a 5-in. pitch diameter driver
pulley having a 155-deg angle of contact, 35-deg included angle, a coefficient of
friction of 0.12, 1800 rpm driver speed, 900 rpm driven speed, belt weight of 0.05
Ib/in. and a maximum allowable tension of 500 lb.
14. Find the capacity of the drive in Problem 13 if belt tension is limited to 350 Ib.
15. A 5-in. pitch inverted tooth chain operates on a 20-tooth drive sprocket rotating
at 5000 rpm. The driven sprocket rotates at 1250 rpm. Find the recommended
minimum center distance.
16. Find the minimum center distance for a 3-in. pitch roller chain operating on a
16-tooth drive sprocket rotating at 4800 rpm and a driven sprocket rotating at
2400 rpm.
17. A 12-tooth ?-in. pitch roller chain sprocket rotates at 1800 rpm. Find the approxi-
mate amplitude and frequency of chain motion due to chordal action.

REFERENCES
[1] Engineering Steel Chains. American Steel Chain Association, St. Petersburg, Fla.,
1971.
[2] Roller Chain Engineering. Rex Chainbelt, Inc., Milwaukee, Wis., 1962.
[3] C. E. Wilson and W. Michels: Mechanism—Design Oriented Kinematics. American
Technical Society, Chicago, Ill., 1969.
[4] Design Manual for Roller and Silent Chain Drives. American Sprocket Chain
Manufacturers Association, Park Ridge, Ill., 1955.
[5] Belt Conveyors for Bulk Materials. Conveyor Equipment Manufacturers Association,
Cahners Publishing Co., Boston, Mass., 1966.
 Brakes and Clutches
SYMBOLS

C = specific heat PE = potential energy

E, = frictional energy r = radius
f = coefficient of friction T = torque
en OLce v = volume
F, = actuating force V = velocity
F, = normal force w = width
g = acceleration of gravity W = wear
J = mass moment of inertia « = angle of contact
KE = kinetic energy @ = subtended angle
n = speed, rpm o = actual stress
p = pressure T = shearing stress

In this chapter we will discuss two very similar machine elements: the
brake, a frictional device that absorbs the kinetic energy of moving bodies
and thus controls their motion, and the clutch, a friction device whose
primary function is to transmit power on an intermittent basis.
We will first turn our attention to clutches, which are generally classified
as mechanical (Figure 13-1), electrical (Figure 13-2), and hydraulic (Figure
13-3). There are numerous subtypes of these three classifications that will
also be discussed.

SECTION 13-1
Positive Contact Clutches
The type of clutch that we designate as mechanical falls into two categories :
friction and positive contact. We will first consider the positive contact type,
an example of which is shown in Figure 13-4. As is evident from the example
shown, this type of clutch transmits power from the driving shaft to the
driven shaft by means of jaws or teeth.
678
 Figure 13-1 A mechanical clutch that has
rotary cam actuation. [Courtesy Formsprag
Company. }

Figure 13-2 The clutch shown is of the

electromagnetic stationary coil tooth type.
The direct tooth-to-tooth engagement
prevents slippage. Typical applications
include packaging and printing machinery,
machine tool spindle drives, speed change
mechanisms, elevators, heavy duty
conveyors, and rolling mill drives.
[Courtesy Formsprag Company. ]

679
 Figure 13-3 The clutch shown is a
heavy duty oil-through-the-shaft-type,
multiple disc hydraulic clutch. It operates
to pressures of 500 psi. [Courtesy
Formsprag Company.]

Figure 13-4 Serrated tooth positive

contact clutch. [Courtesy Dodge
Manufacturing. }

680
 681
Section 13-1: Positive Contact Clutches

The advantages ofthese positive con lutches are that they will not slip,
they develop very little heat because they do not depend upon friction, and
they are generally lighter and less costly than a friction clutch of similar
torque capacity. The disadvantages are that positive contact clutches cannot
be engaged at high speeds, shock accompanies engagement at any speed,
and they require some relative motion in order to engage when both driving
and driven shafts are at rest. Positive clutches, although not as widely used
as friction clutches, do find important use in such applications as automotive
transmissions, business machines, presses, and household appliances.
In order to demonstrate the procedure to be used in analyzing positive
clutches, we will use the square jaw clutch as an example. Figure 13-5 shows
half of a square jawed positive clutch. In operation, the half shown slides
along the shaft upon which it is mounted until it engages with the similarly
shaped other half. Clearly, some device, such as a spring or a hydraulic or
pneumatic device, is needed to keep the two halves of the jaw in axial contact.
As far as stress analysis is concerned, the jaws are subjected to bearing and
shearing stresses.
The force acting on the jaw that produces these stresses depends upon
the horsepower and speed that the clutch is to transmit. Equation (13-1)
can be used to determine the torque acting on the clutch.
h 63,0
Te hp x 63,000 (13-1)
n

where T = torque in inch-pounds; hp = horsepower to be transmitted;

and n = revolutions per minute.

Figure 13-5 Square-jaw positive-contact clutch.

 682
Chapter 13: Brakes and Clutches

At this point the average force, assumed to be acting at the center of each
jaw, can be determined from equation (13-2).
je
(13-2)
k(r, + r))/2
where F = average force in pounds; k = number of jaws on one member;
r, = outside jaw radius in inches; and r; = inside jaw radius in inches.
The shearing stress, in pounds per square inch, acting on one jaw is then
given by
— F
(13-3)
— Onrs tr) Ox t
2 360
where t = thickness of jaw, in inches and 6 = angle subtended by one jaw
in degrees.
Clearly, 27(r, + r,)/2 represents the average circumference of the jaws.
The bearing stress can be found from equation (13-4).

E
c.— r (13-4)
where b = length ofjaw.
As was pointed out earlier in this discussion, positive clutch engagement
is accompanied by shock, which can be rather severe. As a result, the stresses
calculated by the previous equations must be multiplied by appropriate
k factors. Some designers make the assumption that only one of the
jaws takes the entire load, thus making the design even more conservative.

SECTION 13-2
Disc Clutches
The first of the friction type clutches to be discussed is the disc clutch,
also known as an axial or plate clutch. This type of clutch is able to transmit
torque from the input to the output shaft because of the frictional force
developed between the two plates or discs. Figure 13-6 is a simplified
sketch showing the basic components of this type of clutch.
The input disc is free to move axially along the shaft but is pinned, splined,
or keyed to the shaft so that it must rotate with the shaft. Clearly, the torque
that can be transmitted will depend upon the frictional force developed,
which in turn depends, among other factors, upon the axial force developed
between the discs. The axial force can be applied in several ways, such as by
mechanical means (levers, springs, linkages), hydraulic or pneumatic
pressure, or electromagnetic means.
The primary adv: antage s of friction clutches are that, because they can
slip relative to each other, there is very little shock during engagement and
 683
Section 13-2: Disc Clutches

Input shaft
Output shaft

Friction material

Figure 13-6 Disc type clutch.

they can be used for high speed engagement applications. The disadvantages
are that they do slip (not suitable for applications that require positive
transmission), they do wear out (requiring replacement of friction material),
and heat is developed (might require external cooling).
Before proceeding with the analysis of disc clutches it is necessary to
consider the physical situation that occurs when the discs are in contact
and transmitting power. If the discs or plates used in the clutch are relatively
flexible,it is possible to obtain relatively uniform pressure on the friction
surfaces. On the other hand, if the plates are relatively rigid, the wearing of
the friction surface is approximately uniform after an initial wearing-in
has taken place. In analyzing the clutch, either the assumption of uniform
wear or uniform pressure must be made. , SO
the designer must decide which assumption more closely approximates the
particular clutch he is analyzing: An alternative favored by some designers
is to use only the uniform w ssumption, because it turns out to be the
It is more conservative because clutch capacities
obtained from the equations based upon this assumption are lower than
those obtained from the uniform pressure approach. We will present both
and allow the designer to use his own judgment as to which is more appro-
priate for a specific problem.

Uniform wear
One disc of the clutch to be analyzed is shown in Figure 13-7. The wear
may reasonably be assumed to be directly proportional to pressure intensity
and velocity at a particular point on the clutch. This velocity is directly
 634
Chapter 13: Brakes and Clutches

Friction material

\NN
S
\\
>

Figure 13-7 The actuating force needed to engage a disc clutch.

proportional to the radius. Letting W be the wear, we have

W ‘cc pV cc ipr

where p = pressure intensity in pounds per square inch at some radius r in

inches. Therefore
W = kpr

where k is a constant. Thus

—=pr=K
k
where K is a constant, since both W and k are constants.
Because pr is a constant, it is obvious that the maximum pressure must
occur at the minimum radius, r;. Thus

K= DI = Prax! i Ol Ue Draast tf

The numerical value for the maximum permissible pressure depends upon
the type of friction material used. Table 13-1 lists some values for maximum
pressure as well as values for coefficient of friction for some typical clutch
and brake materials.
The actuating force, F,, is the force pressing the two discs together and
is normal to the friction surface. It can be obtained by multiplying the
pressure between the friction surfaces by the area of the surfaces. However,
since the pressure is not constant with respect to r, we must write the force
for the elemental ring shown in Figure 13-7 and integrate between appropriate
limits.
 685
Section 13-2: Disc Clutches

Table 13-1 Brake and Clutch Material Properties

Maximum
Maximum Drum Coefficients Allowable
Material 8
Temperature, “F of friction, Pressure,
psi

Metal on metal 500-600 200-250

Wood on metal 200 50-90

Leather on
metal 150-200 15-40

Molded blocks 500-600 100-150

Asbestos on
metal in oil 0.35-0.45 50-150

Sintered metal
on cast iron
in oil

Thus, the equation for actuating force required is

|
Ope | p2nr dr

Fo r.

= i Bee dr
ip
i

= DTD pases = ri) (13-5)

In order to arrive at an equation for the torque capacity of the clutch
(the frictional torque developed), the frictional torque for the elemental
annular area must be integrated over proper limits. Thus,

T= {pour dr

or

T= TT Pax" ~~ r?) (13-6)

wheref = coefficfent of friction.
The designer often finds it convenient to have an equation available
relating torque capacity and actuating force. Since F, = 227 ;Dmax("y — ri),
from equation (13-5), and equation (13-6) can be written

T= TUT Pmax\"o = rr, ty ri)

equation (13-7) is valid.

ie ofPea) (13-7)
2
 686
Chapter 13: Brakes and Clutches

Uniform pressure
Under the assumption of uniform pressure, every part on the clutch face
can be subjected to the maximum permissible pressure. Clearly, under this
assumption the wear will not be constant. Proceeding in the same manner
as was used previously, the actuating force equation Is

pe i PesanDocedr
> Ranta ) te) (13-8)

The torque capacity of the clutch is then

{P= | 1fPmax20t dr
ri

= 37Pmax("o — Ti) (13-9)

And the torque capacity expressed in terms ofthe actuating force becomes

Peementee?
ip ends
ese p?
a3; 10)
Friction clutches are also made in the form of rims or drums. As can be
seen in Figure 13-8, the friction material is placed around the outer surface
of the drum. Since the analysis of this type of clutch is similar to that for
drum brakes, to be discussed in Section 13.6 of this chapter, we will not
discuss that type of clutch at this time. An advantage that the disc clutch
enjoys over the drum clutch 1s that multiple discs can be used with a resultant
increase in torque capacity and with the same actuating force as is required
for a single disc. Other advantages are the absence of centrifugal effects and
efficient heat dissipation surfaces.

Figure 13-8 A friction clutch of the rim or

drum type. [Courtesy Formsprag Company. ]
 687
Section 13-2: Disc Clutches

Figure 13-9 A typical multiple-disc dry clutch. This type of clutch is air or
mechanically actuated. [Courtesy Twin Discs Incorporated. ]

The multiple disc plate clutch of the type shown in Figure 13-9 can have as
many as 50 or 60 discs, with the discs being alternately driving and driven
discs. Although the torque capacity equations we developed were for a
single pair of friction surfaces, they can be used for multiple discs by merely
multiplying the values obtained by the number of active surfaces.

Example 13-1. A single plate disc clutch, with both sides of the plate effective, is to
be used in an automobile. The friction material that the clutch is made of has
a coefficient of friction of 0.3 and a maximum allowable pressure of 15 psi.
If the outside radius of the clutch is 10 in. and the inside radius is 8 in., determine
the torque that can be transmitted and the actuating force required of the
springs.
Solution: We will solve this problem using both the uniform wear and the
uniform pressure assumptions in order to verify the earlier statement that the
uniform wear assumption results in a more conservative estimate.
Uniform wear. From equation (13-6)
T= TT Pras(le a; r?)

= x 013 x) 8 x) 15(10% = 87)

4070 in.-lb for one side effective
 688
Chapter 13: Brakes and Clutches

Therefore for two sides effective

T= 2 x 4070in.-lb = 8140 in-lb
Now from equation (13-5)
F, = 227;Pmax("o — M1)
2n x 8 x 15(10— 8)
I 1507 Ib

or by equation (13-7) for one side

For two sides then

and
Fe i - 8140
= S07 Ms)
“f(r, +1;) 0.3110 + 8)
Uniform pressure. From equation (13-9)

T = 31fPmax("o — 1?)
= 4n x 0.3 x 15(10° — 83)
= 4437 in.-lb for one side effective
For two sides effective
T =2 x 4437 = 8874 in-lb
By equation (13-8)

Ey = [Dmg 7)
=n 15, x (107.—
87)
1696 lb
Thus, it is obvious that the uniform wear assumption gave a smaller torque
capacity and actuating force; therefore, it is the more conservative of the two
assumptions. @

SECTION 13-3
Cone Clutch
Another type of axial clutch is the cone clutch, shown in Figure 13-10.
Cone clutches have the advantage of being able to transmit a larger torque
than disc clutches with the same outside diameter and actuating force.
The reason for this capability is the increased frictional area and the wedging
action that takes place. Cone clutches find their widest use in relatively low
peripheral speed applications.
Figure 13-10 Cone clutch.

Friction
material

Figure 13-11 Free body diagram of one half of a cone clutcn

showing the normal and actuating force.

689
 690
Chapter 13: Brakes and Clutches

In Figure 13-11 a free body diagram of one half of a cone clutch is drawn.
The activating force F,, which is usually produced by means of a spring,
must be large enough to produce the normal force required as well as to
overcome the frictional force f F, present during engagement.
Summing the horizontal forces will result in the actuating force equation
(13-11)
F,=F,sine+fF,cosa or F,=F{sina-+fcosa) (13-11)
If we assume that the resultant frictional forcefF,,acts at the mean diameter
of the clutch cone, we can obtain the following equation for the frictional
torque, or torque capacity of the clutch.

Th = F
d av
ii n. DD

but

ae = it ae r;

thus

Flr, +
poe ri) (13-12)

The torque can also be expressed in terms of the actuating force by

combining equations (13-11) and (13-12).

= he ( a r;)

~ sina+fcosa 2

-? f(r, “te r)F,

~ Asina + f cos a) ae

An analysis similar to the uniform wear assumption made for disc clutches
in Section 13.2 results in the Blowin oer ants (13-14) through (13-16).

2nr; :
jae aa
sin o%
(13-14)

F, ra
= LRT Damani = ri) (13-15)

THT P max iy UG ai r;
T=“ (r? — r?) = Fad(ro +Pd) (13-16)
Sin o% 2 sino

The angle « has a minimum value of 8 deg and, although there is no upper
limit, a typical value would be about 12 deg.
 691
Section 13-4: Other Types of Clutches

SECTION 13-4

Other Types of Clutches

The functions performed by the mechanical clutches, discussed in the
previous sections, can be performed equally well by electric clutches, which
utilize magnetic fields to produce torque transmission. The basic operational
procedure involved in electro-magnetic clutches can be illustrated by
Figure 13-12.
Electric clutches are usually classified into the following categories:
friction disc, serrated tooth, hysteresis, magnetic particle, and eddy current.
Another general type of clutch is the hydraulic or pneumatic clutch, which
has the advantage of minimizing shock and vibration. Figure 13-13 is an
example of a pneumatic air clutch.
There are many applications such as washing machines, textile machinery,
pumps, appliances, helicopter rotors, and so on, for which it is desirable
to have automatic, smooth engagement between driven and driving member
whenever some minimum speed is exceeded. The ideal clutch for this type
of problem is the centrifugal clutch, such as the one shown in Figure 13-14.

External frame

Armature

Friction material

Flux path

Splined hub

Center
line

Clutch and Clutch Coupling Flux Path

Figure 13-12 Clutch and clutch coupling. The

clutch-coupling and the clutch consist of three basic elements :
the field, the rotor and the armature. The field and its coil are
held stationary. The rotor is normally driven by a prime mover
such as an electric motor and the armature is attached to the
load. When the field coil is energized, a flux path is set up and
the armature is magnetically attracted to the rotor. Through
friction, the armature and the load to which it is connected will
lock in at the same speed as the rotor, as long as the field coil is
energized. When the field coil is de-energized, the armature
disengages, no torque is transmitted, and the armature and load
come to rest. [Courtesy General Time Corp., Thomaston, Conn. ]}
 Figure 13-13 Pneumatic air type
clutch. [Courtesy Horton Manufacturing
Company, Inc., Minneapolis, Minn.}

Figure 13-14 The centrifugal type clutch illustrated engages when some
minimum speed has been exceeded. [Courtesy Mercury Clutch.]

692
 693
Section 13-4: Other Types of Clutches

Figure 13-15 The clutch shown is

designed for low speed overrunning and
indexing applications. [Courtesy
Formsprag Company. ]

The last type of clutch we will discuss is the overrunning or freewheeling

clutch. They are generally classified into three categories, roller, sprag, and
wrapped spring. This clutch permits torque transmission in one direction
only and thus is useful in such applications as feed mechanisms, bicycle
coaster brakes, automotive clutch freewheeling, and so on.
As a typical example of an overrunning clutch, the Formsprag clutch is
shown in Figure 13-15. The description on p. 694 is by courtesy of the
Formsprag Company, Warren, Mich.
To conclude this discussion of clutches, it should be emphasized that the
designer will very often find it more economical to choose a standard
clutch from a manufacturer’s catalog rather than going through an analysis
and design himself. In other words, if the designer needs a clutch capable of
transmitting 5000 in.-lb of torque, there are any number of standard clutch
types from which he can choose one whose operating characteristics most
nearly solve the particular problem. However, there are many instances
where the designer will have to check the claims made by a manufacturer
or where he will have unique problems that do not lend themselves to
solution by standard items. It is hoped that when these situations arise,
the designer will be able to use the analysis procedures presented to solve
them.
Stripped of such items as gears,
splines, bearings, oil seals and other
attachments, a Formsprag over-
running clutch consists basically of
a cylindrical inner race and a cylin
drical outer race surrounding it, with
an annular space left between the
two races. A full complement of
accurately formed sprags fills this Simple over-running
annular space.
This class of application is typified
by standby and compound drives.
For example, a steam turbine and
a standby electric motor may be
connected to a single driven shaft
through over-running clutches. The
Each sprag is essentially a strut shaft can then be driven by either
placed between the races in such a the turbine or the motor or both,
way that it transmits power from with no further modification of the
one race to the other by a wedging installation. The turbine drive clutch
action when either race is rotated automatically engages when the
in the driving direction. Rotation turbine starts to drive, but auto-
in the other direction frees the sprags matically over-runs when the load
and the clutch is disengaged, or is transferred to electric motor
overruns. Either race may be the
driven member.

Indexing
In this type of application, recipro-
If one race of the over-running cating motion applied to the driving
clutch is securely fixed to a grounded race is transformed into intermittent
member, so that it cannot rotate, motion in only one direction at the
and the other race is free to turn, driven race. For example, if a pinion
the free race will turn freely in one is connected to the driving race, a
direction of rotation, but will be Locked
rack meshing with the pinion can
locked to the ground in the opposite give reciprocating motion to the
direction. If the grounded race is driving race. The clutch will then
connected instead to a moving advance or “index” the work (driven
mechanism rather than to ground, race) on each forward stroke of the
then the over-running clutch will rack, but will not return or back-up
transmit torque to the moving mech- on the return stroke of the rack.
anism in one direction, but will
free-wheel, and transmit no torque
when turned in the opposite
direction.

Holdbacks/Backstops
In backstopping or holdback appli-
In a Formsprag clutch, each ‘’strut’’
cations, one race is always fixed to a
in the previous examples is actually
stationary member. The function of
a fully formed sprag. The cross
the clutch is to permit rotation of the
sectional shapes of these sprags
mechanism, connected to the other
have been developed by Formsprag
race, in one direction only, and to
engineers to mee$ most conceivable
prevent any rotation in the reverse
types of clutch applications. Since
different types of over-running appli- direction at any time. Although the
cations will involve different loading clutch normally overruns most of the
characteristics, clutch technology is
time, it is referred to as a holdback*
or backstop in conveyors, gear
varied to provide the greatest pos-
sible life, torque capacity and reducers and similar equipment be-
cause its function is to prevent
functional characteristics for the
run-back,
three basic types of over-running
clutch applications.

694
 695
Section 13-6: Energy and Power Considerations

SECTION 13-5
Brakes
A brake may be simply defined as a machine element that is used to control
the motion of a moving body by absorbing kinetic energy. This is the case
in the usual applications where a rotating body is being slowed down or
stopped, although there are instances, such as bodies being lowered by
hoists, cranes, and so on, where the energy to be absorbed is potential
energy. A last general class of problem involves the use of brakes to keep
bodies at rest.
Although one important requirement in selecting a brake is its torque
capacity, which was also the case with clutches, another vital consideration
is the ability of the brake to absorb and dissipate heat. This additional
consideration is necessary because clutches usually connect two moving
bodies, whereas brakes connect a moving and a stationary body. In general,
however, the types of brakes to be discussed could also be used as clutches
with some slight modifications.

SECTION 13-6
Energy and Power Considerations
As has already been mentioned, the function of a brake is absorbing
energy and dissipating the resultant heat. The question of the time it takes
for a brake to dissipate heat compared to the time during which it is absorbing
energy is quite important. Another way of saying this is that, although a
brake may act satisfactorily when used at largely spaced intervals, it could
overheat when applied steadily. Most of us at one time or another have
experienced the unpleasant burning smell of overheated brakes when applying
the brake too often in trying to control the speed of a car going down a
long hill. Carried to an extreme, the temperature of the brake lining could
become high enough to cause the brakes to fail completely. The point being
made here is that the rate at which energy must be absorbed and heat
dissipated by a brake is extremely important.

he r ing, ch aerate means that ie peace of

the brake may ie serena curtailed. Clearly, the designer must avoid fading
by designing the brake for a more efficient dissipation of heat.
In considering the energy to be absorbed, it is obvious that the energy
equation will depend upon the type of motion the body is undergoing.
To illustrate, let us consider the following examples.
1. A crane is to lower a body of weight W, from a height h, to a height h,.
The change in potential energy, which is also the energy that must be
absorbed by the brake is given by
APE = W(h, — hy) (13-17)
 696
Chapter 13: Brakes and Clutches

2. A body, of weight W, is translating at a speed V, and is to be slowed

down to a speed V,. The change in kinetic energy is

1W
AKE = 5 lake V+) (13-18)

where g = acceleration of gravity.

3. A body of weight W, rotating at a speed of «; rad/sec, is to be braked
down to a speed of w, rad/sec. The change in kinetic energy is then

AKE = 3J(w? — @}) (13-19)

where J is the mass moment of inertia of the body about its axis of rotation.
The three cases cited above are the most common types encountered.
However, it should be obvious that more complicated problems can be
broken up into combinations of these three types.
At this point, we may assume that the frictional energy developed must
be absorbed by the material adjacent to the brake lining (usually a metal
drum). It is then possible to calculate
the temperature increase for the brake
by means of equation (13-20)
E, = pvCat (13-20)

E, = frictional energy the brake must absorb, ft-lb

p = weight density of the drum material, lb/in.?
v = volume of drum material absorbing the frictional energy, in.*
C = specific heat of the brake drum material,
— ft-lb
(Ib)(°F)
At = increase in temperature of the brake drum, °F.
Faires [1] suggests the following as suitable values for specific heat:
cast iron = 101, steel = 93, aluminum = 19%.
To recapitulate, the energy a brake must absorb is calculated by the
appropriate equations (13-17) through (13-19), and equation (13-20) is then
used to determine the temperature of the drum material. If enough time is
allowed for the brake drum to radiate and conduct the heat that was
developed before the next braking operation, then there will be no buildup
of heat. However, if cooling does not take place rapidly enough, there will
be a higher initial temperature at the start of every braking operation. The
result will still be an equilibrium temperature, but a much higher one. An
exact determination of this equilibrium temperature is not possible because
of the uncertainty about the brake’s environment. Approximate methods
are available, such as the one shown in reference [2].
 697
Section 13-7: Band Brakes

Table 13-1 listed some of the maximum drum temperatures for some
commonly used brake and clutch materials. These temperatures are the
maximum values for steady operation
As will be shown in Section 13.11, the torque capacity of brakes depends,
among other factors, upon the maximum permissible pressure that can be
developed between the braking surfaces. Table 13-1 listed some suggested
values for maximum pressure. The designer must exercise his judgment
with regard to what pressure to use by evaluating the effects of such opera-
tional characteristics as length and frequency of braking.
Most brake lining manufacturers include the effect of the rate of energy
dissipation by giving limiting values of pV for given materials. The units
for pV, energy dissipation rate, are expressed as foot pounds per square inch
per minute (ft-Ib)/(in.7)/(min) or horsepower per square inch (hp/in.”). To give
the designer a feel for the numbers involved, the following pV values are
typical [1]. .
1. Less than 28,000 in applications involving continuous operation, and
inadequate heat dissipation.
2. Less than 55,000 for intermittent operation and poor heat dissipation
but with long periods of rest.
3. Less than 83,000 for continuous application with good heat dissipation.
The values given above are merely guidelines, and the manufacturers’
catalogs should be consulted for values to be used for specific materials
and applications.

SECTION 13-7
Band Brakes
The band brake, Figure 13-16, is perhaps the simplest of the many braking
devices. The braking action is obtained by tightening the band wrapped
around the drum that is to be slowed down or halted. The difference in the
tensions at each end of the band determines the torque capacity.
The relationship between the tensions, F, and F,, is derived in exactly
the same manner as that used for flexible belts, with the exception that the
centrifugal force acting on belts does not act. Rather than go through the
analysis again, we will simply repeat the relationship here.
[fet
om e (13-21)

where
F, = larger tensile force, lb

F, = smaller tensile force, lb

f = coefficient of friction
@ = angle of contact between band and drum, rad
 698
Chapter 13: Brakes and Clutches

Figure 13-16 The simple band brake shown is operated by tightening

the band wrapped around the drum that must be slowed down.

The relationship between the actuating force F, and the slack side tension
F, is obtained by summing moments about point O.

ee ee (13-22)
where a and c are the distances shown in Figure 13-16.
The torque capacity is obtained by summing moments about the center
of rotation of the drum.
T (k= By (13-23)
where r = radius of drum.

Referring again to the analysis of belts, a similar equation for maximum

tension exists

Fy = PmaxWl (13-24)

where w is width of band.

 699
Section 13-8: Differential Band Brakes

SECTION 13-8
Differential Band Brakes
The differential band brake, shown in Figure 13-17, is similar to the
simple band brake just discussed, except that the tight side tension helps the
actuating force. Se eS CG as will be
shown shortly, may even be self-locking. The equations discussed in the
previous section apply, except that the actuating force equation, obtained
by summing moments about the pivot point O, becomes

jh epee ae (13-25)
It is obvious from an inspection of the equation that the actuating force
See is less hon Ve had osen attached at point 0. In fact
oT than « Fs brake 1] ckin: Bea if

Figure 13-17 Because the differential band brake makes use of the
tight side tension to help the actuating force, it is called self-energizing.
 700
Chapter 13: Brakes and Clutches

SECTION 13-9
Disc Brakes
Disc brakes, Figure 13-18, are very much like the disc clutches discussed
in article 13-2. In fact, since the analysis follows exactly that presented
previously, there is no need to spend time discussing them in detail.
Disc brakes are used in heavy duty industrial applications because they
can be designed to dissipate heat quickly and thus have relatively little
problem due to fading. They have also been used to a limited extent in
automotive applications, one reason being that they provide equal braking
torque for either direction of rotation.

Figure 13-18 Disc brakes of the type shown

are used in heavy duty industrial applications
because of their ability to dissipate heat quickly.
[Courtesy Goodyear Tire and Rubber Company,
Industrial Brake Department. ]

SECTION 13-10
Short Shoe Block Brake
A block brake, Figure 13-19a, is considered to be a short shoe brake if
the pressure distribution is constant along the shoe; in other words,
if the |
 701
Section 13-10: Short Shoe Block Brakes

With uniform pressure distribution, the resultant normal force will act at
the center of the shoe. Figure 13-19b is a free body diagram of the forces
acting on the shoe.
The normal force can be determined from equation (13-26)

Fa pee ero, (13-26)

The torque capacity of the brake is then given by
=i r (13-27)

(a) Ete?)
\
\ /

Figure 13-19 (a) The block brake shown is considered to be a short shoe brake
if the angle of contact is small enough to permit the assumption of uniform pressure
distribution along the shoe. (b) The free body figure shows the forces acting on the
shoe and the actuating arm. When the arm pivot is located at O, the brake is
self-energizing; if it is located at O’, it is not self-energizing.
 702
Chapter 13: Brakes and Clutches

Summing moments about the pivot point O, will give the equation for
determining the actuating force.
=M, = F.a—Fd+fFc

Fee a
fe (13-28)
where F, is the actuating force in pounds, and d is the distance shown in
Figure 13-19.
The horizontal and vertical components of the pin reaction at O can be
found by summing the horizontal and vertical forces.
RG E (13-29)
ee ee (13-30)
The brake as drawn is self-energizing because the frictional force helps
the actuating force (reduces the actuating force required for a given braking
torque). It should also be obvious that if d = fc, no actuating force is required,
and if d < fc, the actuating force F, 1s negative, which means the brake ts
self-locking and a force is required to disengage the brake once it has been
engaged. The self-energizing effect is useful, but the self-locking effect is
generally undesirable. The usual procedure is to design the brake so that d
is at least 25-50% greater than fc. This will ensure proper utilization of the
self-energizing effect while preventing self-locking.
Una
had been located
RN NEEENAIGS. If the pivot point 0
below the line of action of the frictional force, say at O’
in Figure 13-19b, the moment due to frictional force would oppose that of
the actuating force, and the brake would not be self-energizing. Finally, both
of the situations discussed, pivot at O or at O’, will be reversed if the direction
of rotation is reversed.

SECTION 13-11
Long Shoe External Block Brake

t
. Since most shoe brakes
have contact angles of 90 deg or more, it is clear that a more exact analysis
is required. The obvious problem concerns the determination of the pressure
distribution. Because the shoe is not rigid, it will deflect, and this effect in
addition to the load applied will probably cause the pressure distribution
to be different than that assumed. However, we will make the usual
_
tion, namely, that the pressure varies directly as the distance from the pivot
point, O. This assumption is equivalent to assuming that the wear is propor-
tional to pV, (pressure times velocity).
 703
Section 13-11: Long Shoe External Block Brake

Figure 13-20, depicts a long shoe external contacting brake. The pressure
at some arbitrary angle 6 is proportional to c sin 6, but since c is a constant,
the pressure varies directly as sin 6. In equation form
pox csin@« sin
or p = ksin6@ where k is a proportionality constant, and p is the pressure
at some arbitrary angle 6. If p,,,, 1s the maximum allowable pressure, as
determined by the properties of the brake lining material, then

= p sl Pmax

Sin ssini@y
and

p= = sin 0 (13-31)
Pmax s

sin 8 max

This equation gives the pressure distribution at a particular angle 0.

Obviously for small values of 0, very little pressure is developed, and therefore
the brake is usually designed for 8, > 10 deg. The maximum pressure occurs
at 6 = 90 deg, and if @, is made larger, the magnitude of the pressure
developed decreases. As a result, there is very little braking capacity to be
gained by letting 0, exceed 120 deg.
Finally, theterm sinOa. isevaluated asfollows:
if 6, < 90 deg, sin 6,,., = sin 0,
if 8, > 90 deg, sin 6,,.x = sin 90 deg

Since the pressure is not constant around the shoe, it will be necessary
to analyze an element rd as shown in Figure 13-20. Our immediate objective
‘will be to obtain an equation for calculating the actuating force F,. Obviously,
summing moments about pivot point Owill be the procedure used. Although
it is a relatively simple matter to obtain equations for determining the
magnitudes of the total normal and frictional forces acting on the shoe,
determining their point of application and lines of action is not simple.
Because we are primarily interested in the moments they produce about O,
we will go right to the determination of the total moments.
The moment due to the normal force is
92
M,, =| pwr dé c sin 0
Oy
02 P

={ sin or do csin @
64 sin max

= ee cio) _ 0,) — 3(sin 20, = sin 26,)] (13-32)

Sings,
 704
Chapter 13: Brakes and Clutches

The moment due to the frictional force is

62
Mr, = fpwr dO(r — ccos 8)
9,
02
=|; ie sin Owr dO(r — c cos 8)
PeeSHICU ae

= 5 fasts] ros 0, — cos @,) — (sin? 6, — sin? 0] (13-33)

In max
Rx

w = width of drum

Figure 13-20 The long-shoe external brake shown cannot be analyzed by

assuming a uniform pressure distribution across the shoe.
 705
Section 13-11: Long Shoe External Block Brake

If we now sum moments about the pivot point O, we get

M, +M
Fos — (13-34)
a

- of the pivot point. When using the equation to determine the actuating
force, the algebraic sign obtained for the moment should be used. It is also
obvious that if both M,, and M;, have the same sign, self-locking cannot
occur.
Reversing the direction of rotation of the drum will reverse the sign of
the moment due to the frictional force, and the actuating force equation
becomes

_ (13-35)

Clearly, for this situation, self-locking will occur if Mp, > Mp,
The torque capacity equation for the brake can be obtained by integrating
the product of the frictional force and the drum radius.
62
T; = | fpwr d0r
6

-{ f——-~*
Pmax _sin Owr? dO
SiO es.

iy
oe wr2(cos 0, — cos 03) (13-36)
Finally, the designer will find it useful to have equations for determining
the pin reaction at the pivot point O. Summing forces in the y direction we
get
02 02
I —F, sing + { pwr d0cos0 + | fpwr d@sin 0
) y=
6, 0;

02
= —F sing + Par | sin 0 cos 6 d@
sin 6 max 6;

ea hi i sin?
6dO
aon Goes 0 :

Pen pee Pa Sin’ Gye sill 10,

=— wr —
ae sin 0 max 2 Z

+ 1[50, — 6,)- (sin20, — sin 20,)|! (13-37)

 706
Chapter 13: Brakes and Clutches

Similarly
02 7)

R x Se F,cos ~ | pwr
dé sin @ + fpwr dé cos 6
O, O1

1 Dae :
= F.cos¢ + Ty, (0, — 0,) + —(sin 20, — sin 20,)
sin 0 max 4

+
sin 04) Sin “4 (13-38)
f Z ac

Reversing the direction of rotation will change the sign of the terms
containing the coefficient of friction.

SECTION 13-12
Long Shoe Internal Brake
A type of brake that is widely used in automotive applications is the
internal contacting shoe brake shown in Figure 13-21. The method of
analysis and resulting equations are identical with those discussed in the
previous section [equations (13-32), (13-33), and (13-36)].
The actuating force for a counterclockwise rotating drum, Figure 13-21 is
given by
—M
Fr (13-39)

w = width of drum

OF,

Figure 13-21 The long-shoe internal brake shown is widely used

in automotive applications.
 707
Section 13-12: Long Shoe Internal Brake

Figure 13-22 When the shoes are pivoted as shown and the
rotation is clockwise, both are self-energizing. However, for
counter clockwise rotation, neither is.

Figure 13-23 When the shoes are pivoted as shown, only the
upper shoe is self-energizing. However, reversing the direction of
rotation will make the lower shoe self-energizing.
 708
Chapter 13: Brakes and Clutches

If the direction of rotation is reversed, the equation becomes

M,,, + M,,
Fi= (13-40)
a

As was mentioned previously, drum brake shoes usually do not exceed

angles of contact of 120 deg. The result is that most internal brakes are of
the double shoe type. The shoes are usually pivoted either as shown in
Figure 13-22 or Figure 13-23. When the pivots are located as shown in
Figure 13-22, both shoes are self-energizing; but when the direction of
rotation is reversed, neither is. For the brake shown in Figure 13-23, the
upper shoe is self-energizing and the lower shoe is not. However, if the
direction of rotation is reversed, the lower shoe is self-energizing and the
upper shoe is not. For a given actuating force, the braking capacity with
both shoes self-energizing is clearly greater than if only one were. Automotive
brakes are usually made with both shoes self-energizing, with the result
that braking ability in reverse is much less than for forward motion.

SECTION 13-13
Materials for Brakes
The materials used to manufacture brakes fall into two general classifica-
tions, those used for the drums and those used for the linings. Drums are
usually made of cast iron with some alloying materials added. More expensive
materials such as stainless steel, aluminum, monel, and so on, are used when
good heat conduction is important.
Brake linings on the other hand are usually made of some form of eroestcs
and binders. Most asbestos based linings are molded, although there are
some special applications for woven linings. Linings of this type are normally
not used if the operating temperature is expected to exceed 400°F. For
brakes, in which the temperature is expected to be in the 400—750°F range,
sintered metal linings are usually used. If the temperature range falls into
the 750—1000° F category ceramic particles are added to the sintered metal
linings. In fact, if the ceramic content is high, operating temperatures as
high as 1800°F can be tolerated.
The sintered linings have the advantage of high thermal conductivity, a
long life span, and very stable friction characteristics. The disadvantages
are that they are relatively expensive, they have a lower coefficient of friction,
and they must be manufactured to shape because of their rigidity.
Most linings are attached to the drums by either riveting or bonding.
Although riveting has the advantage of low cost and relative ease and
simplicity of installation, it does have the disadvantage of having the usable
lining thickness dependent upon the depth to which the rivet head can be
countersunk.
Bonded linings afford more friction area and greater effective thickness
but are more expensive and require more elaborate equipment and careful
control when they need to be replaced.
 709
Section 13-15: Brake Actuation

SECTION 13-14
Electrical Brakes
Figure 13-24 illustrates a typical electromagnetic brake. Brakes of this
type have two basic components, the armature and a friction forced magnet
that attracts the armature when power is applied. By making or breaking
an electric contact, the two moving parts are either engaged or disengaged.
Although the electromagnetic brake is the most commonly used electrical
brake, three other types—hysteresis, eddy-current, and magnetic particle
brakes—have some use in certain applications.

SECTION 13-15
Brake Actuation
One of the important problems with which a brake designer must concern
himself is the method by which the brake is to be activated. Mechanical
brakes are activated mechanically, pneumatically, hydraulically, or electri-
cally.
Mechanical activation is used for low torque, low speed applications such
as scooters, conveyors, golf carts, garden tractors, minibikes, and so on.
Clearly, a limiting value on the braking torque that can be developed by

Figure 13-24 The electric brake shown is

very much like an electromagnet in that it has
two basic parts, the magnet and the armature.
When power is applied, the magnet attracts the
armature. The magnet is faced with friction
material, thus eliminating slippage between the
two parts when they are engaged. The two
moving parts can thus be engaged or disengaged
merely by making or breaking an electrical
contact. [Courtesy Warner Electric Corporation]
 Accumulator

Pressure limit and

modulating valve
| (700 psi max.) Pian pee
Pilot line (+ +--+ =
Pump Peres | A
motor driven At : Bleeder
(inlet) Unloading valve A ls

Separate Closed Center ee =

Hydraulic System 2 1.D. REE

hose or tube y
Steering valve or other 5i.d. (min) hose
7, open center devices Accumulator Ai.d. (min) hose or tube—=| or tube
O In-line “oe
adjuster
Bleeder
Pump
motor driven
(inlet)
Brake Brake
Return line
Brake Arrangement
Reservoir
Common To Both Systems
Pressure os
relief valve (_) For Hi-Torque Brakes Only
Combination Open
Center-Closed Center Hydraulic System

Figure 13-25 Typical straight hydraulic system, open center—closed center.

(Courtesy B. F. Goodrich.]

Rotochamber
Brake Brake

Brake valve Rotochamber,

Boy
DIESsets. Bleeder Roe
indicator
Air gauge Master
Master
cylinder
aes Se ae
vlervoi | eee cylinder
Wee

Quick release
valv , 5
€ Oil reservoir
; Governor Air
Air
tank
compressor

Rotochamber
Master | Rotochamber
\ cylinder
Brake Brake

fae Tractor Trailing vehicle

Figure 13-26 Typical air-over-hydraulic system. [Courtesy B. F. Goodrich.]

710
 711
Section 13-16: Design Considerations

this type of activation is the magnitude and length of time the operator can
exert the required pressure. Another disadvantage is that mechanical
linkages usually require lubrication.
Most brakes, however, are activated either pneumatically, hydraulically,
or by a combination of the two. A typical hydraulic brake system is shown
in Figure 13-25. Combination brakes, also called air-over-hydraulic brakes,
of the type shown in Figure 13-26, are used for heavy rolling equipment.
Electrically activated brakes are of two main types. The first, which was
discussed in the previous section, depended upon a friction-faced magnet
attracting a rotating armature. In the other type the brake is spring-set
and is released by magnetic means. The advantage of the second type is
that it is independent of the power. In other words, in the event of a power
failure, the brake will be in operation until the power is restored to release it.
The first type will not hold if there is a loss of power.

SECTION 13-16
Design Considerations
Brakes and clutches are available in commercial units. As is true of the
other machine elements discussed previously, the designer must decide
whether the particular brake or clutch is to be designed and manufactured
or a commercially available unit is to be used. In general, unless large
numbers are required, it is more economical to use commercial units.
Even if the problem requires a nonstandard unit, most companies will be
willing to have their engineering staffs solve the problem.
In the event, however, that large quantities or unusual conditions of
operation necessitate the manufacture of a system, the following general
steps can be used as a guide. The analysis will usually be based upon the
torque capacity required and the energy that must be absorbed and dissipated.
Factors such as required lining area, method of activation, available space,
the characteristics of the friction material, effect of the activating force on
the structure supporting the element, characteristics of the machine of which
the element is to be a part, and the environment in which the machine is
operating must all be considered. The usual result is a compromise of all
these factors so as to ensure an economically competitive product.

PROBLEMS
1. For the square-jaw clutch shown in the Problem Figure 1, r, = 2in., r; = 12in.,
b = 4in.; 75 hp is to be transmitted at 650 rpm. Each of the jaws subtends an arc
of90 deg. In other words, only 180 deg of the clutch surface has jaws on it. Calculate
the shear and bearing stresses if t= ?in.
 a
Chapter 13: Brakes and Clutches

Problem Figure 1

2. Part of the transmission system of an automobile consists of a single disc clutch,

both sides of which are effective ;25 hp is to be developed at 200 rpm. Experience
has indicated that an outside diameter that is 3 times the inside diameter gives
good results. If the friction material of the clutch has a coefficient of friction of 0.3
and can develop a maximum pressure of 15 psi, calculate the inside and outside
diameters as well as the actuating force required. Assume uniform wear.
3. Repeat Problem 2, but assume uniform pressure.
4. Asingle disc clutch having both faces effective is to be used in an application where
60 hp are to be developed at 600 rpm. If the outside diameter is 9 in. and the inner
radius 4in., what will be the actuating force required and the average pressure
for a clutch material having a coefficient of friction of 0.3? Assume uniform wear.
5. Repeat Problem 4 for uniform pressure.
6. A multiple disc clutch has four active faces, it is rated at 40 hp at 250 rpm. The
coefficient of friction is 0.3, outside diameter = 16in., inside diameter = 8in.
Determine the actuating force required and the average pressure between the
discs.
7. A multiple disc clutch has an outside diameter of 6 in. and an inside diameter of
3 in. The clutch material permits a maximum pressure of 40 psi and has a coefficient
of friction of 0.2. Determine the actuating force required, the number of discs
needed, the average pressure, and the actual maximum pressure. Uniform wear
may be assumed and a torque of 2800 in.-lb is to be transmitted.
8. A multiple disc clutch having four pairs of mating friction surfaces operates at a
speed of 945 rpm. The maximum permissible magnitude of the axial actuating
force is 800 lb. Coefficient of friction equals 0.15. The inside and outside radii of
contact of the friction surfaces are 1 in. and 2 in., respectively. If conditions are
such that uniform wear prevails, determine: (a) The torque capacity per pair of
friction surfaces. (b) The total horsepower capacity. (c) The maximum pressure.
(d) The minimum pressure.
9. Derive the torque capacity equation and the actuating force equation for a cone
clutch similar to that shown in Figures 13-10 and 13-11 for the assumption of
uniform pressure.
}— ; : es
Nek
 713
Problems

10. Acone clutch, which is made with an average diameter of 12 in., has friction material
whose coefficient is 0.3, and whose maximum allowable average pressure is 50 psi,
is used to transmit 20 hp at SOO rpm. If « = 8 deg and uniform wear is assumed,
calculate the actuating force and face width.
11. Repeat Problem 10 assuming uniform pressure.
12. Repeat Problem 10 for an « of 12 deg.
13. A cone clutch has an average diameter of 10 in., an « of 10 deg, and a coefficient
of friction of 0.3. If an actuating force of 700 1b is applied, what is the torque that
the clutch can transmit. Assume uniform wear.
14. Repeat Problem 13 for a uniform pressure of SO psi.
15. A drum having a diameter of 5 ft is used to support a wire that is attached to a
weight of 8000 lb. The drum is rotating at 60 rpm, has a radius of gyration of 2 ft,
and weighs 10,000 lb. A brake is to bring the drum to rest in 15 ft. What is the
average braking torque?
16. Repeat Problem 15 but for a drum rotation of 90 rpm.
17. Ifthe brake rim in Problem 15 is made ofsteel, has a 10-in. diameter, 5-in. thickness,
and a 4-in. face width, what will be the temperature rise assuming all the energy is
absorbed by the rim?
18. Repeat Problem 17 for the data of Problem 16.
19. For the drum brake shown in the Problem Figure 19, determine the actuating
force F, required to decelerate the drum at a rate of 120 rad/sec. The coefficient
offriction of the drum may be taken as 0.25 and the moment ofinertia of the rotating
system as 2 ft-lb-sec?.

Rotation

Problem Figure 19

20. Repeat Problem 19, but assume counterclockwise rotation of the drum.
21. For the band brake shown in Problem Figure 19, the angle of wrap is 270 deg,
coefficient of friction is 0.3, and the drum radius is 8 in. What are the tight and loose
side tensions required to have a friction horsepower of 30 hp at 900 rpm.
 714
Chapter 13: Brakes and Clutches

22. A simple band brake similar to that in Problem Figure 19 has an angle of wrap
of 270 deg, a coefficient of friction of 0.3, a 12-in. drum diameter, and a maximum
allowable pressure of 100 psi. If the band width is 3 in., determine the tight and
loose side tensions and the torque capacity. What is the limiting actuating force?
23. The band brake shown in the Problem Figure 23 is to rotate in both directions.
If a frictional horsepower of 100 is to be developed at 500 rpm, determine the
actuating force required for a coefficient of friction of 0.1.

Problem Figure 23

24. For the brake in Problem Figure 24 a frictional horsepower of 10 is to be developed

at 600 rpm. Will the brake be self-locking? The coefficient of friction may be taken
as 0.4. What actuating force is required if a = 10 in.?
. A brake similar to the one in Problem Figure 24 has a diameter of 20 in., an angle
of wrap of 200 deg, the distance from the pivot point to where T, attaches is 1 in.,
and the distance from the pivot point to where T, attaches is 5 in. If the actuating
force is 40 1b, a = 25in., and f = 0.3, will the brake be self locking? What is the
maximum braking torque?
26. Repeat Problem 25 for counterclockwise rotation of the drum.
27. If the brake in Problem Figure 19 is to absorb 30 hp at 700 rpm and the maximum
pressure that can be developed is 70 psi, the drum width is 2 in., and the coefficient
of friction is 0.2, what angle of wrap will be required? What actuating force will be
required?
28. For the short shoe block brake shown in Problem Figure 28 r = 15 in., a = 40 in.,
c = 2in.,d = 15in. For a coefficient of friction of 0.3, a frictional horsepower of
20 at 650 rpm is to be absorbed. What actuating force is required? Can the brake
be self-locking? Calculate the pin reaction at O.
 files)
Problems

Rotation

Problem Figure 24

Problem Figure 28
 716
Chapter 13: Brakes and Clutches

29. Repeat Problem 28 for clockwise rotation. What must c be to have self-locking?
Calculate the pin reaction at O.
30. An external long shoe brake similar to the one shown in Figure 13-20 has the
following dimensions: a = 25 in.,c = 12in.,0, = 10 deg, 0, = 100 deg, w = 3in.,
= 20 deg, r = 8 in., rpm = 600. If the maximum permissible pressure is 100 psi
and the coefficient of friction 0.3, determine the actuating force required, the
horsepower the brake will absorb, and the pin reaction at O.
31. Repeat Problem 30 for counterclockwise rotation.
32. The brake shown in Problem Figure 32 has a coefficient of friction of 0.25, an
allowable maximum pressure of 100 psi, and a face width of 2 in. Determine the
actuating force required, the horsepower that can be absorbed at 200 rpm, and
the reactions at the pin.

Probiem Figure 32

33. Repeat Problem 32 for counterclockwise rotation.

34. The block type hand brake shown in Problem Figure 34 has a face width of 13 in.
The frictional material permits a maximum pressure of 80 psi and a coefficient of
friction of 0.24. (a) Determine the force F,. (b) What is the maximum torque
capacity? (c) If the speed is 100 rpm and the brake is applied for 5 sec at full capacity
to bring the shaft to a stop, how much heat is generated?
35. Repeat Problem 34 for counterclockwise rotation.
36. An internal long shoe brake similar to the one shown in Figure 13-21 has the
following dimensions: radius of drum = 6in., a = 9in., c = Sin. 0, = 10deg,
6, = 120 deg, coefficient of friction = 0.3, lining width = 3 in., maximum allow-
able pressure = 90 psi. What actuating force is required, and how much horse-
power can be absorbed at 400 rpm.
37. Repeat Problem 36 for clockwise rotation.
 FAT.
Problems

Rotation

Problem Figure 34

38. The automotive type brake shown in Problem Figure 38 has a drum of radius
7 in., a coefficient of friction of 0.4, and an actuating force of 600 lb. What is the
torque capacity of the brake?

Rotation

Problem Figure 38
 718
Chapter 13: Brakes and Clutches

39. Repeat Problem 38 for clockwise rotation of the drum.

40. In Problem Figure 38 assume the right hand shoe is pivoted at the top and the
actuating force is applied at the bottom. The left hand shoe remains the same.
What is the torque capacity?
41. Repeat Problem 40 for clockwise rotation.

REFERENCES
[1] V. M. Faires: Design of Machine Elements. The Macmillan Company, New York,
1965S.
[2] Gagne: Torque capacity and design of cone and disc clutches. Mach. Des.,
24(12): 182-187 (Dec. 1953).
[3] Mechanical Drives Issue, Machine Design. The-Penton Publishing Co., Cleveland,
Ohio, 1969.
[4] Peter Black: Mechanics of Machines. Pergamon Press, Elmsford, N.Y., 1967.
[5] H. S. Rothbart: Mechanical Design and Systems Handbook. McGraw-Hill Book
Co., New York, 1964.
[6] A. W. Frehse: Fundamentals of brake design. SAE J., 27.
[7] A. C. Rasmussen: Internal friction, blocks and shoes. Prod. Eng., 18(3): 133
(1947).
[8] L. D. Hagenbook: Design of brakes and clutches of the wrapping band type.
Prod. Eng., 16: 321 (1945).
[9] Z. J. Jania: Friction clutch transmission. Mach. Des., 30 (1958).
[10] C. L. Eksergian: High speed braking. ASME Trans., 73: 935 (1951).
[11] H. A. Borchardt: Designing external shoe brakes. Mach. Des., 32(13): 163 (1960).
[12] R. L. Kotnik: Electromagnetic disc clutches. Mach. Des., 32(16): 113 (1960).
[13] A. J. Bette: Friction materials. Mach. Des., 32(20): 141 (1960).
 Springs
SYMBOLS

C = spring index S,. = endurance limit in shear, psi

D = diameter of wire, in. S.yp = Shear yield strength, psi
g = acceleration of gravity 6, = solid deflection, in.
(386 in./sec”) 6, = working deflection, in.
h, = free height, in. t = shear stress, psi
h, = solid height, in. Tm = mean shear stress, psi
K = spring rate, lb/in. T, = range shear stress, psi
N = factor of safety v = Poisson’s ratio
N, = number of active coils @, = natural frequency, rad/sec
R = mean coil radius or moment arm of load.

Many machine members are designed on the basis of strength. In most

cases, deflection is of secondary importance. Usually deflection is checked
to see whether it is reasonable. In the design of springs, however, deflection
is aS important as strength. A significant deflection is necessary to most
spring applications. These include the storage of energy, isolation of vibra-
tion, application of a steady force or torque, force measurement, and so on.

SECTION 14-1
Torsion Bars
A hollow or solid bar (Figure 14-1) may serve as a spring where significant
angular deflection results from a torsional load. For torque PR, the relative
rotation of the ends is given by
DPR
= GJ rad (14-1)

where G is the shear modulus of the bar, and the relative displacement

b= = $R=—— IRS
~R = —__— (14-2)
14-2

Use)
 720
Chapter 14: Springs

diameter D

Relative
rotation
of ends
-_

Support

(b) Tir) — Tp)

T(7) + 710)

Figure 14-1 (a) Torsion bar subject to load P at moment arm R. (b) Cross
section at B showing approximate shear stress distribution on a horizontal
diameter.

Bending is neglected in both cases. For a solid circular cross section, the
polar moment of inertia

jess
732 (14-3):
and for a hollow circular section of inside diameter D,,
n(D* — D?)
J = —— ry) 14-4
(14-4)
Shear stress due to torsion alone is proportional to distance from the
center of the bar. The maximum value that occurs on the surface is

sey = TD = SPR
O44| mD
ne
for the solid bar.
For the part of the bar that does not lie between supports, there is a
direct shear load P and an average direct shear stress P/A for cross section
area A. For vertical load P, the torsional shear stress and direct shear stress
combine to give the maximum shear stress at the end of a horizontal
diameter nearest to the point of load application (Figure 14-1b). At that
point, the direct shear is somewhat greater than its average value. Based
on the theory of elasticity, it is
(1 + 2v)PD?
Oe Teen (s-°)
 Fel
Section 14-2: Helical Springs under Static and Dynamic Loads

where I = 2D*/64 for a solid section. Adding equations (14-5) and (14-6),
and using v = 0.3, we obtain the maximum shear stress

16PR 0.3075D
—— ae |b SB (14-7)
mD R

where the effects of the end curvature and the effects of bending are neglected.

SECTION 14-2
Helical Springs under Static and Dynamic Loads
Helical springs of various proportions are shown in Figure 14-2. The

mb! s Goneidcn for

example, the compression spring sketched in Figure 14 . If the resultant
load and reaction P lie on the axis of the spring coil, then an element of the
spring is subject to a torque PR where R is the mean coil radius (the average
of inside radius and outside radius). There is no bending moment resulting
from the load P, and deflection is due almost entirely to torsion in the wire.
Using equation (14-2), we have

(14-8)

Figure 14-2 Helical compression

springs. Most of the larger springs have ©
squared and ground ends. [Courtesy
Associated Spring Corporation, Bristol,
Conn.]
 Le
Chapter
14: Springs

Figure 14-3 Helical compression spring with

squared and ground ends.

|
Effective wire length L is a function of the mean coil radius R and the number
of active coils N,:
L = 2nRN, (14-9)
Substituting equation (14-9) and the value of polar moment of inertia for
a solid wire of diameter D into equation (14-8), we obtain the deflection of
a helical compression spring
._ 64N,PR° ai
Gpe
and a spring rate
Pa GDS
(14-11)
6 64N,R*
The number of active coils, N,, may be one or two coils less than the
total number of spring coils N,. The spring of Figure 14-3 has squared
and ground ends, and it is apparent that the end coils do not participate
in the spring deflection. For springs with squared and ground ends, there
are about two inactive coils and we have
N,=N,-2
For squared ends, Figure 14-4, the approximate number of active coils is
given by

Ny = Ne= ts
 723.
Section 14-2: Helical Springs under Static and Dynamic Loads

Figure 14-4 Helical compression spring with

squared ends.

With squared ends, also called closed ends, the end turns he in a plane
approximately perpendicular to the spring axis.

Working deflection
Let us define the working load P,, on a compression spring as the greatest
expected load and the working deflection 0,, as the deflection corresponding
to the working load P,,. Furthermore, referring to Figures 14-5 and 14-6,
we will define the solid deflection as the difference between the free (no
load) height and the solid height

6, =h, —h, (14-12)

The clash allowance is defined by
0, — Ow
‘ye (14-13)

which is a margin of extra deflection divided by the working deflection.

A clash allowance of 0.20 (20 %) is satisfactory for most applications. Using
this value, an overload of 20% will deflect the spring to its maximum deflec-
tion 0, and greater overload will have no effect on deflection or stress. Thus,
with a sufficient safety factor, a compression spring 1s protected against
failure after it reaches its solid deflection. Of course, the spring may fail
to perform its primary function. For example, when compressed to its solid
height, a spring mount can no longer protect a piece of equipment from
shock or vibration damage.

erase h, Figure 14-5 The solid height of a spring.

 724
Chapter 14: Springs

—~ Working 5
. w
a ee Deflection Solid A
Deflection *
————

pec
isares ae
Se SSS
a —— — ae EE Gy
new es xi
Free Height

Figure 14-6 Working and solid deflection of a helical spring.

Shear stress
While shear stress due to torsion in the straight bar (Figure 14-1) is given
by
_ 16PR
Roy po (14-14)

the effect of curvature increases the above value by the factor

1
CA
for a helical spring where the ratio of mean coil diameter to wire diameter
2R
Cc = — (14-15)
D
is called the spring index. Using the above correction factor (called the
Wahl curvature correction) we revise equation (14-7) to read

_ 16PR(AC — 1, 0.615
TD Dead eG
or (14-16)
8CP/4C—1 0.615
— 3 =i
mio 4" 6G
Some designers treat the Wahl curvature correction as a stress concentration
factor and neglect it for static loading. This practice is equivalent to simply
using equation (14-7). However, considering the nature of typical spring
materials and the stress distributions encountered we will use equation (14-16)
for static loading and for the mean and range stress components for fatigue
loading.
 72)
Section 14-2: Helical Springs under Static and Dynamic Loads

Design for static load

Let rking st 1 S.,,/N and working load |
P,, given.
be If we are
required to ence a spring with aan these restrictions, we must make some
assumptions. Suppose we decide that the spring index C will be no less
than 5. Then, from equation (14-16), shear stress
OSPR
a (14-17)
Ds
Selecting a reasonable mean coil radius R and setting t = S,,,/N, the working
strength in shear, we obtain the required wire diameter D for the spring.
If the result is reasonable and if the spring index C is no less than the minimum
selected above, we go to the next step in the design.
Let us assume that the spring rate K is specified (as it might be if the spring
is to be used for vibration isolation). Selecting a material, we may then
rewrite equation (14-11) to obtain the number of active coils

N= se (14-18)

If squared and ground ends are selected, the total number of coils is
N,=N, +2
The solid height is given by

ND (14-19)
If we select a 20% clash allowance, we will reach the solid height with a
20% overload. Thus, solid deflection
P

6, = 1.20—
5 K

and the free (unloaded) height is given by

h, =h, + 6,
The spring should be wound so that the center-to-center spacing between
active coils (the pitch) is

Pa D+ N,
°;

when there is no load.

Example 14-1: Design for Static Loading. Required: Design a set of

helical compression springs to support a load P = 700 lb/spring. The natural
frequency of the system is not to exceed 100 cpm. Design decisions: Let us
select steel with a shear yield point S,,, = 70,000 psi and a shear modulus
 726
Chapter 14: Springs

G = 11.5 x 10° psi. If we use a clash allowance of 20%, the spring will be
protected in case of overloads in excess of 20%, and a low safety factor, say
N = 1.5, will be reasonable. A spring index C = 6 will be used and the coil
ends will be squared and ground.
Solution: Rearranging the second of equations (14-16) and settingt = S,,,/N,
we obtain

D2 =
8PCN/[4C — 1 0.615
Ssyp
4C —4 C
Then for the given values, the result is a wire diameter D = 0.536 in., or we
might use D = ; in. At this point, we must check the availability of this
diameter steel with the assumed yield point. Mean coil radius is

The spring rate K is related to the natural frequency f,(cpm) and supported
mass by

o epi eh OE,
OO Baie
Using the maximum allowed OU tf,= 100 cpm, a supported wee
w = 700 lb, and g = 386 in./sec*, we obtain
2

k= ue = 199 Ib/in. (14-21)

60
Then substituting this value of K in equation (14-18) we obtain N, = 18.8
active coils, from which N, = N, + 2 = 20.8 total coils. The solid height h, is
N,D = 11.7 in. and, with 20% clash allowance, the solid deflection is 120Wes
of the working deflection or
= le
= 125 — 49350

Thus, the free height is given by hy = h, + 6, = 15.93in. @

The design is not complete, however, until we consider surging and

buckling which are treated in the Sections 14.3 and 14.4, respectively.

Design for fatigue loading

If we know the endurance limit in reversed shear stress and the shear
Sode riterion may be misedaieroy on the basis of
shear SEE oythe endurance limitof the spring wire is given for one-way
modified Soderberg procedure is used. The one-way shear test is
based on a shear stress varying ssanateile from zero to a maximum value
S‘. (Figure 14-7a). It is assumed that S{, is the value of shear stress for which
a part is on the verge of failure after an “‘infinite’’ number of cycles. In many
cases, S, may be based on 1 x 10° or 1 x 10’ cycles of shear loading.
Referring to Figure 14-7a, we see that the range and mean components of
the shear loading are both 3S{,. Thus, we will designate points (45. 45%.)
and (S,,,,, 0) as failure points on a modified Soderberg plot, Figure 14-7b.
The line between the failure points will be used to predict failure due to
 MDL
Section 14-2: Helical Springs under Static and Dynamic Loads

Figure 14-7 Fatigue. (a) Fatigue test :one-way shear. (b) Modified Soderberg
criterion for spring.

combinations of mean and range stress. Similarly, we will identify the limit
of safe eave of Lee and range stress by the safe stress line drawn
between points (35;./N, 4S,./N) and (S,,,/N, 0) for safety factor N. Stress
peints to ile on or ole the safe stress line will constitute acceptable
spring design. The equation of the safe stress line is found by substituting
the two stress points in the general equation of a line. The safe stress line
is defined by
Ks
t, = 5 EA yl Nae fy) (14-22)

S., a oe

An alternate form of equation (14-22) yields the safety factor for given
loading
Saas
N= oP 14-23
oR oS caer s.) Seale Ose ( )
If we know the ratio of range to mean stress, it may be convenient to rewrite
equation (14-22) in the form

N
se (14-24)
(25 oe ee Sy Sse)
Ss

se
 728
Chapter 14: Springs

Equation (14-24) is useful if we wish to design a spring when the ratio of

range to mean load is known. Once mean stress t,, is found, we may use it
and the mean load P,, to find wire diameter D. Using equation (14-16),
for example, we may write
16P,,R/4C —1 0.615
3 = =. 14-25
i Gee Fe —4 : re | ( )

or
8P,C/4C —1 0.615
eg 14-26
x le. (32—4 ss C | (
The rest of the design procedure is similar to the design procedure for
static stress. Thene
"he cle vance se based ximum load

Example 14-2: Design for Fluctuating Loads. Required: Design a set of

helical compression springs to support a weight of 700 lb/spring where the load
per spring varies from 600 to 800 Ib as in Figure 14-8a. The natural frequency
of the system is not to exceed 100 cpm. Design decisions: As in the preceding
problem for static loads, let S,,, = 70,000 psi; G = 11.5 x 10° psi; clash
allowance = 20%; N = 1.5; and C=6. Use squared and ground ends.
Furthermore, let us specify steel wire with an endurance limit in one-way
shear of S, = 30,000 psi.

0) 10 20 Tn 40 $1) 50 Ke
(b) N
Figure 14-8 An example of design for fluctuating loads. (a) Fluctuating
load. (b) Graphical solution for mean stress (used to find required diameter).
 729
Section 14-3: Surging of Helical Springs

Solution: The mean load is given by

a eaves Sta ein

= 700 lb

and the range load by

Eps
ee = max min
= 100i
os: 1

Using a modified Soderberg plot, Figure 14-8b, we draw a line through the
origin having a slope

to represent all possible designs for the given loading. The safe stress line
is drawn between the point (3S;,/N, 3S;./N) and point (S,,,/N, 0) on the same
coordinates. The intersection of the two lines, point A on Figure 14-8b,
represents the limiting safe stress. Using this construction, we read the mean
stress value t,, = 30,400 psi. Identical results are obtained by using equation
(14-24). Using the above value of range stress and setting P,, = 700 1b in
equation (14-26), we obtain wire diameter D = 0.665 in. (or say +2 in.). The
spring rate is the same as in Example 14-1 for static loading, and the number
of active coils and the total number of coils are found by the procedure used
in that example. The solid deflection is given by
Jz
os = 1.20
0 K

In addition, the spring must be checked for surging and buckling. @

SECTION 14-3
Surging of Helical Springs
In Example 14-2, we considered a weightless spring of spring rate K
supporting a rigid mass m. The natural frequency of the system is given
by

or

1) EK
= ae cps (14-27)
2n\Vm

The frequency so obtained is the lowest natural frequency and probably

the most important natural frequecy. The spring, however, cannot actually
be weightless, and thus it has vibration characteristics of its own. Vibration
 730
Chapter 14: Springs

effects within the spring are called surging, and the associated frequencies
are found by examining a small element of the spring in harmonic motion.
Let a small element of the spring in Figure 14-9 consist of dN coils. The
length of wire in the element is 27R dN. Thus, for wire diameter D and mass
density p, the element has a mass of
n’D*pRdN
dm ="
For vibration amplitude u at circular frequency w and at any axial location
(coil number N = 0 through N = N.), the inertia force amplitude is given
by
*D?w?upR dN
af, = dma = > (14-28)
The spring rate
rece:
~ 64N,R>
equation (14-11) represents the force per unit displacement of a spring
having N, active coils. The spring force amplitude corresponding to a
change in displacement per coil du/dN is
EG Darda
* 64R? dN
and the change in spring force amplitude over the element dN is

_ GD* du Ai 14-29
* ~~ 64R3 dN? Ore)

Zz

—S-

(a)

F. F, +dF,
<—_—_ es

(b)
Figure 14-9 Surging of helical springs. (a) Surging. (b) An element of
the spring (dN coils).
 731
Section 14-3: Surging of Helical Springs

Equating the sum of the forces on the element to zero, we obtain

d*u A
AN? +cu=0 (14-30)

where

3217
pRt?
D*G
Equation (14-30) is satisfied by

u = Asin(cN) + Bcos(cN) (14-31)

where A and B are arbitrary constants. For a spring fixed at one end and
free at the other, the boundary condition u = 0 is applied to coil zero and
the condition du/dN = 0 is applied at N = N,. We then have the nontrivial
solution B = 0 and cos (cN,) = 0 from which

CN nano
WAnere w= IL, 3 Dyo0c's
The result is a set of surging frequencies

nD G
On) = SRN, iprad/sec

or

nD G
ay ee ee 14-32
Jo) = TégR2N,\ 2p “PS nee
where n = 1, 3,5,.... The fixed-fixed case leads to a similar solution where
A eee

requencies fc ir spring. In most cases, the ower surging

Tequenoyeorrespondingto n = 1,will be well above the excitation frequency
caused by equipment on a spring mount. In high speed engines, however,
surging of valve springs has been known to occur. When surging 1s predicted,
the spring should be redesigned because surging may lead to fatigue failure
or malfunction of the spring. In the case of valve springs, for example, the
valves might tend to oscillate rather than to operate properly.

Example 14-3: Surging... Given: The spring designed for static loading in Example
14-1 will be examined for surging. Assume that the ends of the spring are
essentially fixed. D = 7: in., R = ?2in., G = 11.5 x 10° psi, N, = 18.8 active
coils. The density of steel is 0.28 Ib/in*; mass density p = 0.28/g = 0.000725
(Ib) (sec?)/in.*
 ey)
Chapter 14: Springs

Solution: Using the values given and equation (14-32), we obtain

fin) = 18.6n cps = 1120n cpm
Wane i) = D4) Ooo
The surging frequencies for this spring are relatively low and may correspond
to operating speeds for equipment mounted on the spring. If this is the case,
it may be necessary to redesign the spring. @

SECTION 14-4
Bending and Buckling of Helical Springs
Helical springs are seldom intentionally subject to bending, but the
problem of buckling may be examined by considering resistance to bending.
Consider a spring of coil length L subject to bending moment M (Figure
14-10a). The effect is an angular deflection 0. Part b of the figure shows the
resultant moment M represented as a vector on a cut section of the spring.
The resultant moment is separated into a torsional component M cos@
tangent to the spring wire axis and a bending component M sin ¢ perpen-
dicular to the spring wire axis. We will assume that the helix angle is small so
that the moment vectors lie, essentially, in a plane perpendicular to the center
axis of the spring coil. For a small element of length R dd, the bending
component produces an angular deflection
MRsin
¢ dd
EI (14-33 )

(a)

log
er Figure 14-10 Bending and buckling of
helical springs.

(resultant)

(b)
 733
Section 14-4: Bending and Buckling of Helical Springs

and the torsional component produces a rotation

MR cos ¢ dd
Gy (14-34 )

for elastic modulus E and shear modulus G. Moments of inertia J and J refer
to the wire cross section; they are
nD* nD*
I =— and J=—— 14-35
64 oy ( )
The center axis of the spring coil deflects in a plane perpendicular to the
resultant moment vector M. Other deflection tends to cancel out. Taking the
components ofthe deflection contributions in that plane and integrating over
the entire spring, we obtain

| 0
one ee ¢
ee
cos*¢
area |40 (14-36)
Substituting the values of J and J, using the relationship
E
G = ——
2(1 + v)
and integrating, we obtain
_ 128MN,R(1 + v/2)
0 14-37
ED* ue te
(which is exact if the number of quarter turns is an integer).
By analogy to a simple beam in pure bending, we may write

C== EI,
Me i
(14-38)

where the equivalent moment of inertia is found by eliminating @ from

equations (14-37) and (14-38). The equivalent moment ofinertia of the spring
coil is
EDs
[fo = 14-
4 128N,R(1
+ v/2) ve)
This result may be used directly in the Euler column equations to obtain
critical loads, that is, the compressive loads at which a spring will fail by
buckling. Figure 14-11 shows the critical loads in pounds per spring for
various end conditions. These values are taken directly from the Section 5-10
on elastic stability in Chapter 5. A spring with well-seated squared and
ground ends might be considered to have clamped ends. Ordinarily, several
springs are used together to support a rigid member. If both ends of all the
springs are well seated and squared and ground the critical load per spring is
RED,
Ea z (14-40)
 734
Chapter 14: Springs

P P \ \N

\ IN

L22 LZ

2 2 4 Z /
Pp Es
EG P eRMEICE
BL Poy = -L:
cr 4L2 cr ie

(a) (b) (c)

m2E/
Load per spring P,, = [2
a

(d)
Figure 14-11 Critical loads for various end conditions. (a) Fixed-
free ends. (b) Pinned—pinned ends. (c) Clamped—clamped ends with
lateral restraint. (d) Clamped—clamped ends without lateral restraint.

as given by Figure 14-11d. Since the end condition may be in doubt, we

should apply a safety factor N., to the critical load, allowing compressive
loads of P,,4x = P.,/N.,- If the spring load is likely to exceed P.,/N,,, we must
provide restraints to prevent buckling.

Example 14-4: Critical Load on a Spring. Given: The spring design for static
loading in Example 14-1 will be examined for buckling. Assume a rigid
body weighing 2800 Ib is supported by four springs, each supporting P = 700 lb
and that the ends may be considered clamped as in Figure 14-11d.
Solution: Using the results of example 14-1, we have
Spring rate, K = 199 Ib/in.
Free height, h, = 15.93 in.
Length when loaded, L = h, — (P/K) = 12.4 in.
Diameter of wire, D = in.
Active coils, N, = 18.8
Coil radius, R = 2Z in.
Elastic constants, E = 30 x 10° psi: vy= 0.3
Substituting in equation (14-39), we obtain an equivalent moment of inertia
eq = 0.000266
Using equation (14-40), P,, = 512 lb/spring, and using a safety factor for
buckling N., = 1.5, we obtain a safe maximum load
Prax = 341 Ib/spring
 ae)
Section 14-5: Extension Springs

Obviously the springs cannot support 700 lb each without additional restraint.
Several paths are open to us:
1. We may insert a rigid rod in the center of the spring to prevent buckling.
2. We may enclose the spring in a tube to prevent buckling.
3. We may restrain the rigid body against lateral motion.
Items 1 and 2 above require careful design because we must insure freedom
of vertical motion to allow the spring to operate as a vibration isolator.
Item 3 is equivalent to Figure 14-11c, with

Based on this solution, P,,,, = 1364 lb/spring which is clearly safe. @

SECTION 14-5
Extension Springs
Helical extension springs of various proportions are shown in Figure 14-12.
The design procedures for helical extension springs are quite similar to the
design procedures for compression springs given in Section 14-2. The coils,
however, are usually wound tightly so that there is an initial tension P;.
There is no deflection until the initial tension is exceeded as indicated in
Figure 14-13. Thus, deflection

a 64N,R3(P — P,)
ane (14-41)

for mean coil radius R, wire diameter D, and shear modulus G. Equation
(14-41) is valid for P > P;. The term N, represents the number ofspring coils,
not counting the end coils which are bent to form a hook. The spring rate,
which applied for P > P; is the change in load per unit displacement change
or

Ke (14-42)

Thus, spring load is given by

P= P,; + Ké (14-43)
The bend in the end coil causes a rather complicated stress distribution
that may be approximated by using a stress concentration factor with the
stress equation. Maximum shear stress is then given by
16K ,PR fc = 1 a
Pegs oC me
Toe

where the spring index is defined as before by C = 2R/D.

 736
Chapter 14: Springs

Figure 14-12 Helical extension springs. Stock springs of this type are available
in music wire and stainless steel with wire diameters from 0.007 in. to 0.115 in. and
coil outside diameters up to 1 in. [Courtesy Associated Spring Corporation, Bristol,
Conn.]

Sharp bends should be avoided since the stress concentration factor is, of
course, greater for sharp bends. A reasonable value which may be used for
design is K, = 1.33. Although compression spring deflection is limited by
the solid deflection, there is no safety feature of this type inherent in extension
springs. Thus, hi y factors ) extens: ri
: G shanica 1ed imit deflection. |

K 2 : ; :
Figure 14-13 Helical extension spring. The
| load-deflection relationship.

P;

0
 75 7
Section 14-6: Leaf Springs

SECTION 14-6
Leaf Springs
The simply supported member and the cantilever beam may be used as
springs. For springs with uniform cross sections, we may use the results of
Chapter 5. For the simply supported member with central load P, Figure
14-14a, spring rate

P 48EI 4bh°E
K=— 5 B LB (14-45)

Maximum bending stress

CMe OPI
O x(max) = bh2 a bh? (14-46)

occurs beneath the load. For the cantilever spring, Figure 14-14b, the spring
rate
K= 3EI bheE
re, Aa (14-47)

Maximum bending stress

CRE
Oxtmax) = pz (14-48)

occurs at the clamped end.

Springs are expected to deflect significantly under load. Usually, a soft
spring (one with a low spring rate K) is most desirable, providing the strength
is adequate. The uniform members considered in the above paragraph are
not efficiently designed springs. In each case, bending stress is a maximum at
one point on the surface and decreases to zero elsewhere on the surface. For
example, maximum bending stress occurs at the load for the simply supported
member.

L
Z P

eae |
2

(a) Simply Supported

Z L i

\\
:

(b) Cantilever

Figure 14-14 Flat springs of constant section.

 738
Chapter 14: Springs

Suppose we design a simply supported spring with varying cross section

so that bending stress is the same everywhere on the surface (Figure 14-15).
Bending moment varies linearly from a maximum at the load to zero at each
support
Ex
M =— (14-49)
0<x<L/2 au
Let us vary the section modulus I/c = bh?/6 the same way. Thickness may be
varied parabolically or width may be varied linearly to obtain the desired
‘beam of uniform strength.” We will choose a varying width as the most

. 4
O<x = :

= (a) Spring with

Varying cross
Section.

Zs Constant
P thickness
h

te 2

the.
4 (b) Moment

] (c) Curvature

(d) Slope

— a SE (e) Deflection
w

Figure 14-15 Spring of uniform stress.

 739
Section 14-6:
Leaf Springs

practical solution. The support ends will be squared, however, to provide

for end fastenings and to prevent failure due to transverse shear. Width is
now a function of x (Figure 14-15a):
2b
ele (14-50)
O0<x<L/2 L

(the spring is symmetric about x = L/2). Moment of inertia

E, bh? * b ,h?x
O<x<L/2 aioe 6L

and maximum stress is uniform everywhere on the surface

6M 1.5PL
9 x(max) a bh2 = bh? (14-51)

We now use the curvature-moment equation

(Pip = ile Sab (14-52)

i lo is ;
which applies over the entire spring. Integrating, we obtain slope
dw BPX
EE sy pee eee
dx COT hE
By symmetry, 0 = 0 at L/2 from which the integration constant
sve 3PE Ee
ee aE Se (et eee ares 14-53
Os aig is age b, WE (5 7 ees)
Integrating once more, we obtain the deflection
LS PLX
w = b ASE (L— x) (14-54 )

Maximum deflection occurs beneath the load. It is

1G; OCIS Pie
o=
»|5]= b WE 14-55
oar)
The spring rate is
2.67b ,heE
= —— (14-56)
1G

In order to maintain the same strength, we let b, (equations (14-55) and

(14-56)) = b (equation (14-45)). Then, the spring of varying width has a 50%
greater deflection or a 33 % lower spring constant than the spring of constant
width.
 740
Chapter 14: Springs

Applications
Springs in the form ofa cantilever, Figure 14-14b, are often used as electrical
contacts. Automotive type leaf springs are equivalent tu the ee sheen
inee 14-15. Let he cen
( 7 in Figure UE ES: th PERRO let
ne sum of the widths at a given location be equal to or slightly greater than
the width of the spring in Figure 14-15. If the leaves are lubricated and free to
slip slightly relative to each other, then spring stiffness and strength are as
calculated for the spring of varying width. The Soderberg criterion for tensile
stress may be used in the design of leaf springs subject to varying loads.

SECTION 14-7
Torsion Springs
Helical compression and extension springs are so designated because the
spring coil as a whole is compressed or extended in length along its central
axis. The spring wire in compression and extension springs is loaded
principally in torsion. Helical torsion springs, Figure 14-17, are subject to
torque about the central axis of the spring coil. This induces bending stresses
in the spring wire.

Helical torsion springs

The helical torsion spring shown in Figure 14-18a is wound left hand, and
and the preferred loading direction is as shown, tending to further wind the
coils. The spring is generally used over a supporting mandrel. The mandrel
diameter is smaller than the inside diameter ofthe coil to allow for contraction
when torque is applied to the spring. Residual stresses resulting from the
original forming operation are tensile on the inner surface of the coil. Thus,
ifit is necessary to design the spring so that the applied torque tends to unwind
the coil, the design should be based on a lower bending stress.

Figure 14-16 Leaf spring. This spring is equivalent to the

spring of uniform stress (see Figure 14-15). Before loading, the
spring has an initial curvature.
Figure 14-17 Helical torsion springs. Torsion springs are wound both left hand
and right hand. The straight ends are normally as shown in Figure 14-18a when the
spring is loaded. [Courtesy Associated Spring Corporation, Bristol, Conn.]

aa ts) (a)

Thickness h — b =
Figure 14-18 Torsion springs. (a)
Helical torsion spring. (b) Spiral torsion
spring.

(b)

741
 742
Chapter 14: Springs

Helical torsion spring deflection is principally due to bending moment

which corresponds to torque about the coil central axis. Thus, we have
angular deflection
ML
i =r rad (14-57)

Referring to Figure 14-18, the bending moment about any section is M = aP

and the length L = 2xRN, for N, coils of mean radius R. For circular wire
of diameter D, I = 2D*/64 and the angular deflection is
128aN,PR
ee (14-58)
ED

A torsional spring rate may be defined as _

M ED* ;
Roars = ra = 128N,R Ib-in./rad (14-59)

If we ignore residual stress, the bending stress is

K ,Mc
ey (14-60)

where c is the distance from the neutral axis in bending to the extreme fiber.
For circular wire the bending stress is
32K ,aP
OS =
1D

The stress concentration factor K,, which depends on wire diameter and
curvature, may be obtained by analyzing the spring as a curved beam,
following the methods given in Chapter 18. Maximum stress occurs at the
inner surface. For spring index C = 4, the stress concentration factor
K, = 1.23 and for C = 6, Ky, = 1.14. For curved or hooked ends, K, may
be greater.

Spiral torsion springs

Analysis of a spiral-wound torsion spring (Figure 14-18b) is similar
because moment on any section is given by the torque aP. For total length
L, and moment of inertia J = bh*/12, we have angular deflection

bale
= bho rad (14-61)

and a torsional spring rate

Dien
Kiorsional = Fane lb-in./rad (14-62)
L2L
 743
Section 14-8: Rubber Mounts

Bending stress

Bi KM
Sahaca (14-63)
If
where I/c = bh?/6 and K, is based on analysis of curved beams. If the
minimum value of R/h = 2,stress concentration factor K , = 1.2;for R/h = 3,
K,
= 1.12.

SECTION 14-8
Rubber Mounts
Natural and synthetic rubber shock and vibration mounts are frequently
selected because of their inherent damping qualities and low elastic moduli.
These characteristics tend to dissipate energy and prevent sound transmission.

Compression mounts
The rubber compression mount, Figure 14-19a, having dimensions
a x b x h will be subject to a nominal compressive stress
P

psi (14-64)
ab

P Rectangular rubber block

dimensions: aX bXh

Rectangular rubber block v

~ dimensions: aX bXh |

Compression Mount
fa
Rectangular Shear Mount

(a) (b)

CSS

Cylindrical Shear Mount Constant Stress

Shear Mount

Figure 14-19 Rubber shock and vibration mounts.

 744
Chapter 14: Springs

and will deflect an amount

6 = ——in. (14-65)

for a load P lb/mount. Poisson’s ratio v = 0.5 (approximately) for rubber

and, considering the low elastic modulus E, the mount will tend to expand
laterally. The rubber, however, may be bonded to metal supports that limit
lateral expansion and effectively increase the modulus as a function of
displacement. Equations (14-64) and (14-65) are only approximate, and E
depends on the hardness of durometer of the rubber selected for the mount.

Shear mounts
The rectangular shear mount, Figure 14-19b, will be subject to a nominal
shear stress
Ie
tT = — psi (14-66)
ab '

and will deflect an amount

j=Gc
ie in. 2
(14-67)

in the direction of the load P lb/mount. Shear modulus G, like E, is dependent

on deflection and on rubber hardness.
If the mount is cylindrical in form (Figure 14-19c), shear stress

SPS (14-68)
with a maximum value

psi (14-69)
‘max ~ Qnah
Shear strain

T
y= G (14-70)

and the total deflection of the load is

s=[ ae P ae ert
Dil! | SRG
re aan est)
For more efficient material use, we might let h vary with r to obtain
constant shear stress throughout (Figure 14-19d). In this case, shear stress
P, P
T= are = 2nbh, P*! (14-72)
 745
Section 14-9: Air Springs

shear strain

and deflection

(14-73)

SECTION 14-9

Air Springs
Air springs are particularly effective for isolation of low frequency vibra-
tions. The natural frequency for a single degree of freedom spring-mass
system is given by
1g hk:
a 5 Kew (14-74)
2x Vm

for mass m and spring rate K. The spring rate was independent of the
supported mass for the springs treated above and, thus, natural frequency
was mass dependent. Most air springs have the peculiar characteristic that
natural frequency is practically independent of mass.
Consider the air spring system illustrated schematically in Figure 14-20.
Let the total volume of air or other medium in the reservoir and bellows be
V) and the absolute pressure po at static equilibrium conditions. Let us
utilize the gas law
pV" = constant (14-75)
where n = 1 for a constant temperature process and n = 1.4 for air in an
adiabatic process. The value n = 1 would hold for very low frequencies. Near
the natural frequency (say 1 to 5 cps) much of the heat of compression would
be retained in the system, in which case a value of n between unity and the
adiabatic value would be appropriate.
Using the gas law, equation (14-75), we may write

PoV6 = P(Vo — Ad)” (14-76)

Compressor

Bellows

Throttle
valve

Figure 14-20 Air spring.

 746
Chapter 14: Springs

for pressure p, deflection 6, and volume change Aé where A is the effective

bellows area. The spring rate is given by the force change per unit displace-
ment change or
Ad ;
R= <,_ Ib/in. (14-77)
Rewriting equation (14-76) to obtain p, differentiating, and assuming a large
reservoir (that is, volume change Ad « Vo) we obtain

A’ npo
K= Ib/in. (14-78)
0

For supported weight P = Ap), the natural frequency, from equations

(14-74) and 14-78) is given by

fo 1 [Agn(pi
Bn(Po asPa) =a (14-79)
2n VoPo
where po is the gage pressure at static equilibrium conditions and Pa is
atmospheric pressure. For relatively high pressures (p> > p,) natural
frequency is nearly independent of pp and the supported weight P.

PROBLEMS
1. Design a steel torsion bar for a static load of 400 lb with a moment arm R = 6 in.
Use steel with a shear yield point of 50,500 psi and a safety factor of 2. (a) Find
wire diameter. (b) Find length for a static deflection of 1.5 in.
2. Design a torsion bar for the same application as Problem 1 except that the static
load is 800 lb and R = 4 in.
3. Design a set of helical steel compression springs to support a static load of 350
lb/spring. In order to provide isolation, the natural frequency of the system must
not exceed 100 cpm based on each spring supporting a 350-lb weight. Use a shear
yield point of 70,000 psi, a safety factor of 3, and a spring index of 6. Design for
squared and ground ends and a 20% clash allowance.
4. Design a set of springs as in Problem 3 except that the load varies from 300 to
400 Ib/spring. Use an endurance limit of 30,000 psi in one-way shear.
5. Design a helical compression spring for 1200-lb maximum load and a spring rate
of 350 lb/in. Use an outside diameter of approximately 5 in., 25° clash allowance
and squared and ground ends. Use alloy steel with a yield point of 120,000 psi,
G = 11.5 x 10°, and a safety factor of 1.5.
6. Design a spring for the same requirements as in Problem 5 except that the spring
rate is to be 150 lb/in.
7. Design a helical compression spring for a load that varies continuously from 0 to
900 Ib. Use squared and ground ends, a spring index of 5 and 20% clash allowance.
Design for steel with a shear yield point of 70,000 psi and an endurance limit of
40,000 psi in one-way shear. The safety factor is to be 1.4 and the spring rate
480 Ib/in.
 747
Problems

. Design a spring as in Problem 7 except that the load varies from 450 to 900 lb.

- A small steel helical compression spring must exert a force of 3 1b when its height
is 2.5in. and a maximum force of 10lb when compressed to a height of 2.2 in.
Base design on fatigue loading using a shear yield point of 80,000 psi and one-way
shear endurance limit of 45,000 psi. Use a spring index of 6 (if possible) and a
safety factor of 1.75. Use squared ends.
10. Design a spring as in Problem 9 except the force is 1 Ib at 4 in. height and amaximum
of 5 1b at 3.5 in. height.
11. Design a helical compression spring for a working load of 60 lb and a spring rate
of 120 lb/in. Decisions: Use steel with squared and ground ends. Let the spring
index be 6 and the clash allowance 20%. Allow a shear stress of 30,000 psi. (a) Find
required wire diameter. (b) Find mean coil diameter. (c) Find the number of active
coils. (d) Find the solid height and the free height.
2: Design a spring as in Problem 11 except that the spring rate is to be 60 Ib/in.
13. A helical coil spring is adequate for static load F,. It has N,, active coils. A second
spring is required to support a static load of 2F,. How many coils must it have if
mean coil diameter, spring rate, and working stress are to be the same as for the
first spring? Assume a spring index C = 6 for the first spring.
14. Design a helical steel compression spring for a maximum static load of 25 lb, a
spring index of 8, a spring rate of 40 Ib/in. and a safety factor of 3. Assume atensile
yield point of 120,000 psi; use squared and ground ends and a 20 % clash allowance.
TS. A helical tension spring has a mean coil diameter of 1 in. and a wire diameter of
0.1 in. It has 10 active coils. If asecond spring were made with a mean coil diameter
of 8 in., what wire diameter would be required to yield the same spring rate as the
first spring?
16. Design a helical steel compression spring mount to support 1000 lb/spring. Use
wire with a shear yield point of 60,000 psi and a safety factor of 3. Use a spring
index of 6. (a) Find wire diameter. (b) Find the number of active coils to produce
a natural frequency of 2.cps with the 1000 lb/spring load. (c) Find solid height
and free height based on 15 inactive coils and a 20% clash allowance.
17. A helical compression spring has a spring rate of 28.5 lb/in., 14 active coils, and a
wire diameter of 0.092 in. Find the pitch so that the allowable load (39.6 |b) will just
be reached when the spring is compressed solid. Use zero clash allowance.
18. Find the wire diameter of a helical spring required to carry a 500 lb maximum
load if the spring index is 6 and maximum shear stress cannot exceed 50,000 psi.
19. The shear yield point of a certain spring material is 60,000 psi and the endurance
limit in one-way shear is 40,000 psi. Use a safety factor of 2. A spring mounting is
to be designed for a varying load of 200-400 lb/spring. Find the required wire
diameter for a mean coil diameter of 6 in.
20. A helical extension spring has 88 turns of 0.020-in. diameter wire and an outside
diameter of 0.120 in. The initial tension is 0.301b. Write an expression for load
versus deflection.
21. A helical extension spring has an outside diameter of 6.10 mm and 49 turns of
0.79 mm wire. Write an expression for extension in millimeters in terms of load in
kilograms if the spring is made of steel and the initial tension is 0.227 Kg.
22. Find the first three surging frequencies of a helical extension spring having 63
turns of 0.020-in. steel wire and an outside diameter of 0.120 in.
 748
Chapter 14: Springs

23. A 0.057-in. outside diameter steel compression spring is made of 0.008-in. wire
and has a free length of 0.56 in. and a solid height of 0.217 in. (a) Find the equivalent
moment of inertia. (b) Find the buckling load if the spring has squared ends (not
ground). Assume pinned-pinned end conditions.
24. A steel leaf spring similar to Figure 14-16 has }-in. thick by 3-in. wide leaves. The
spring is 26 in. long between supports and there are four leaves at the center. (a)
Find spring rate. (b) Find stress and deflection due to a 600-Ib load.
25. A helical torsion spring is made of 0.30-mm stainless steel wire and has a 4.42-mm
outside diameter. (a) Find the number of turns required for a 360-deg deflection
due to a torque of 0.54 kg-mm. (b) Find stress.
26. Find the shear stress in'a rubber mount similar to Figure 14-19d where a = 1,
b = 2, and h, = 1.5 in. for a load of 30 Ib/mount.
27. Anair spring with an effective area of 12 in.” is to be designed for a natural frequency
of 2 cps at a gage pressure of 60 psi. Find the. required air volume.
28. (a) Derive an expression for surging in a helical spring with both ends fixed.
(b) Find the first three surging frequencies of a helical extension spring having
63 turns of 0.020-in. steel wire and an outside diameter of 0.120 in. if both ends are
fixed.

REFERENCES
(1] G. E. Mather: Mechanical Spring Design Guide. North American Rockwell,
Clawson, Mich., 1972.
[2] M. F. Spotts: Mechanical Design Analysis. Prentice-Hall, Inc., Englewood Cliffs,
N.J., 1964.
(3] N. P. Chironis (ed.): Spring Design and Application. McGraw-Hill Book Co., New
York, 1961.
Power Screws
SYMBOLS

= cross-sectional area of the column = lead of thread, in.

(or screw), in.” = number of threads
b= thickness of the tooth at its root = column load, lb
diameter, in. = critical buckling load, in.
C= distance from the centroid of = thread pitch, in.
the column to the outside fiber, = mean screw thread radius, in.
in. = yield point, psi
dm= mean screw thread diameter, in. lowering torque, Ib-in.
d major screw diameter, in. = overhauling torque, lb-in.
dx root diameter of the screw, in. = raising torque, lb-in.
E= modulus of elasticity in tension, = load being lowered or raised, in.
psi = thread helix angle, deg
eccentric distance of the column = thread angle, deg
load, in. = angle between F, and OB, deg
= efficiency of the screw (see Figure 15-10)
friction force, lb = bearing pressure on threads, psi
normal thread load, lb = bending stress of screw threads,
= coefficient of sliding friction psi
= compressive stress of the screw,
between screw collar and
psi
support
coefficient of sliding friction = stress induced at the concave
between thread and nut side of the screw, psi
= tensile stress, psi
= depth of thread, in.
= torsional stress, psi
= factor depending on column
tT = transverse shearing stress, psi
supports
= maximum combined shearing
= least radius of gyration, in.
stress, psi
= length between column
supports, in.

Power screws are used for transmitting motion in a smooth and uniform
manner. They also may be thought of as linear actuators that transform
rotary motion into linear motion. The kinematics of power screws is the same
749
 750
Chapter 15: Power Screws

as that for nuts and screws, the only difference being the geometry of the
threads (see Chapter 16). Whereas power screws find application as motion
devices, ordinary screws find application in holding parts together (that 1s,
as fastening devices).
Typical applications for power screws are (1) automobile jacks, (2) lead
screw for lathes, (3) screw type presses, (4) C clamps, (5) valve stems, (6) hospital
beds, (7) nuclear reactor control rod drives, and so on.
Power screws generally have an efficiency of the order of 30-75%,
depending upon the thread helix angle and the coefficient of sliding friction
between the nut and the screw. However, if higher efficiencies (for example,
90% and higher) are required, the designer may use a “ball screw.” Ball
screws were developed by the Saginaw Steering Division, General Motors
Corp., to be used, originally, in the steering mechanism of automobiles. But,
due to an ever widening field of application, they are now available as stock
items from a variety of manufacturers.

SECTION 15-1
Power Screw Thread Forms
The thread forms used for power screws are (1) Acme screw threads, (2) stub
Acme screw threads, (3) 60-deg stub Acme threads, (4) modified square
threads, and (5) buttress threads. The specifications for these threads are
covered by the ANSI Standards for translating screws [1].
Acme threads are the earliest type of power screw threads. They were
developed for use with machine tools. Acme threads are cut as either general
purpose or centralizing threads. The general purpose threads are classified
into three class fits, namely, 2G for general purposes, 3G and 4G for minimal
backlash. Centralizing threads have associated tolerances that limit the
clearance between the screw’s major diameter and the nut. As a result, the

1
2D allowance
Zgon major diam.

<1
Nwy allowance
\

General Purpose Centralizing

Pp = pitch, in.

Figure 15-1 Acme thread forms.

 Tol
Section 15-1: Power Screw Thread Forms

concentricity between the screw and the nut is controlled. Centralizing Acme
threads are classed as 2C, 3C, 4C, 5C, and 6C, depending on the fillet at the
minor thread diameter, with class 2C prividing the greatest amount of end
play (that is, backlash) between the screw and the nut. The basic dimensional
proportions and nomenclature for the Acme thread are shown in Figure 15-1.
Stub Acme screws are intended for applications where a coarse pitch and
a shallow depth are needed due to heat treating. This thread type has only
one class, class 2G for general purposes. Figure 15-2 illustrates the thread
form for stub Acme threads. Basic dimensions for the Acme and the stub
Acme thread series are given in Table 15-1.
The 60-deg stub Acme thread form is illustrated in Figure 15-3, and thread
specifications are listed in Table 15-2.

Figure 15-2 Stub Acme thread forms.

1\
rex
rae\ \\ Internal

8 § veaons \

a \\
SOO
K
SS
QQ
'Y,

=
SS —WWXIiizzz

Yi
YZYy

:
allowance
2

allowance
minor
diam.
n

External

26\= 29°
6 = 14°30’
p = pitch
n = number of threads per inch
N = number of turns per inch
h = 0.3p, basic height of thread”
F on = 0.4224p = basic width of flat of crest of internal thread
Fy, = 0.4224p = basic width of flat of crest of external thread
Fn = 0.4224p — 0.259 X (major-diameter allowance on internal thread)
F, = 0.4224p — 0.259 X (minor-diameter allowance on extenral thread—pitch-diameter
allowance on external thread)

“Modified stub Acme forms exist having h =0.375p and h =0.250p.

Table 15-1 Basic Dimensions of Acme and Stub Acme Thread
Series (all dimensions in inches)
Stub Acme Threads

General-purpose (All
Classes) and Centralizing Classes
Centralizing Classes 5C and 6C
Threads Basic 2C, 3C, and 4C Basic
per Height of Height of
in. Thread, Helix Helix Thread,
Vp h Basic Angle at Angle at h'

Major Basic Basic

Pitch Pitch
Diameter,
Diameter, Diameter
a

0.03125 0.01875
0.03571 0.02143
0.04167 0.02500
0.04167 0.02500

0.05000 0.03000
0.06250 0.03750
0.08333 0.05000
0.08333 0.05000
0.10000 0.06000

0.10000 0.06000
0.10000 0.06000
0.12500 0.07500
0.12500 0.07500
0.12500 0.07500

0.12500 0.07500
0.16667 0.10000
0.16667 0.10000
0.16667 0.10000
0.25000 0.15000

0.25000 0.15000
0.25000 0.15000
0.25000 0.15000
0.25000 0.15000

SOURCE: Colin Carmichael (ed.): Kent's Mechanical Engineer's Handbook,

12th ed. John Wiley & Sons, Inc., New York, 1960.
For general-purpose and centralizing classes 2C, 3C, and 4C, basic pitch diameter E = D —h;
basic minor diameter K =D — 2h,
For centralizing classes SC and 6C, basic pitch diameter E = B —h; basic minor diameter K =B
—2h.
For stub Acme, basic pitch diameter E = D —h’; basic minor diameter K =D —2'.

752
 Hep)
Section 15-1: Power Screw Thread Forms

Table 15-2 Basic Dimensions of 60 deg Stub Acme Thread Series

Width of flat, in.

Depth of Total Thread
Thread Depth of Thickness
Threads _— Pitch p, (Basic) Thread (Basic) ee Root of
per in. in. h=0.433p, (h+0.02p), t=0.Sp, (Basic) Screw
in. in.* in. F=0.250p F, =0.227p

16 0.06250 0.0271 0.0283 0.0313 0.0156 0.0142

14 0.07143 0.0309 0.0324 0.0357 0.0179 0.0162
12 0.08333 0.0361 0.0378 0.0417 0.0208 0.0189
10 0.10000 0.0433 0.0453 0.0500 0.0250 0.0227
9 0.11111 0.048 1 0.0503 0.0556 0.0278 0.0252

8 0.12500 0.0541 0.0566 0.0625 0.0313 0.0284

‘] 0.14286 0.0619 0.0647 0.0714 0.0357 0.0324
6 0.16667 0.0722 0.0755 0.0833 0.0417 0.0378
> 0.20000 0.0866 0.0906 0.1000 0.0500 0.0454
4 0.25000 0.1083 0.1133 0.1250 0.0625 0.0567

SOURCE: Acme threads, ANSI Standard B1.5-1972;

stub Acme threads, ANSI Standard B1.8-1972; buttress
threads, ANSI Standard B1.9-1972.

“A clearance of at least 0.02p is added toh to produce extra depth, thus avoiding inter-
ference with threads of mating part at minor or major diameters,

A square threaded screw (also known as the Sellers’ thread) is the most
efficient type of sliding friction screw as verified from equation (15-14), but it
has a comparatively poor mechanical advantage. In addition, the square
thread (Figure 15-4) is difficult and expensive to machine. Furthermore, it is
not as compatible as the Acme thread for use with a split nut. As a result, a
modified square thread has been established which is shown in Figure 15-5.

4pitch dia.
allowance

Figure 15-3 60 deg Acme stub thread form.

Tie 7
and modified square threads.
The Buttress thread (Figure 15-6) is a type of thread that is intended to
resist loads in only one direction. Also, it is stronger than the other forms
because the tooth thickness at th t f g A
may be seen from Figure 15-6, the side of the thread that the load h
a Slope of 7 deg. This small angle prod minimal radial fe mp nt,
making this thread form approach the efficiency of the square thread. The
buttress thread is easy to cut, making the cost reasonable

Lf
—N
Y
2 pitch

5p
5p+ )
p
0.4563p — (0.17 X clearance)
basic widt lat at crest of thread) = 0.4563p

Figure 15-5 Modified square thread form. [From U.S. Department of

Commerce: National Bureau of Standards Handbook, H28, Part III. Washington,
DICe 957A)
 LOO.
Section 15-1: Power Screw Thread Forms

Table 15-3 Some Nominal Dimensions for

Square and Modified Square Threads

Modified
Square Threads | Square Threads

Nominal Threads Minor Dia., Thickness of the

dia., in. per inch in. Thread at the Root,*
in.

5 0.163 | 0.0544

3 [a] 0.266 0.0680

1
i 1 0.366 | 0.0837
3 0.575 | 0.1087

en
ee
*Neglecting the effect of major diameter and minor diameter allowances, the thickness at the
root of the thread is 0.5436p

Although this thread form has particularly been applied to screwing

tubular members together, it has also found application in the breech
mechanisms of large guns, cannons, and airplane propeller hubs.
Since buttress threads are employed mainly for special applications, there
is no standard size series in common use. However, nominal major diameters
should be selected from the following geometric series:

Inches

1/2 1 1/8 DMP Seli2 1

9/16 11/4 2 3/4 6 14
5/8 1 3/8 3 UT 16
11/16 2) 3 iP) 8 18
3/4 1 3/4 4 2) 20
7/8 2 41/2 10 22
 756
Chapter 15: Power Screws

Internal Thread (Nut)

V 2

5
3 <Z‘a ‘
NLS
yt NX TAKS
2 he Ss 72 \\ a

‘ WIESE — SEBS AS
alse
. ° G
CE

WE 7

AA
AN 0.020p radius,
2 ZB “ZN approx. (optional)

\
wyMY \\ AW
f External Thread (Screw)
\ >

Basic pitch diam

Min pitch diam of internal thread Minor diam of external thread
Pitch diam of external thread Minor diam of internal thread
Nominal (basic) major diam D,, Major diam of external thread

Nominal major diameter D

Height of sharp V-thread h = 0.89064p
Basic height of thread h = 0.6p
Root radius r= 0.07141p
Root truncation s = 0.08261p
Depth of engagement h, = h —G/2
Crest truncation f = 0.14532p
Crest width F = 0.16316p
Major diameter of internal thread (nut) D, = D + 0.12542p
Minor diameter of external thread (screw) K, = D — 1.32542p —G
Height of thread of internal thread (nut) A, = 0.66271p
Height of thread of external thread (screw) A, = 0.66271p

Figure 15-6 Buttress thread form. [From ANSI Standard BI.9-1972.]

Similarly it is suggested that the pitches for buttress threads be selected from
the following geometric series:
Threads Per Inch

20 10 5 OMt/2. 11/4
16 8 4 2 1
12 6 5 Lye
Tee
REREAD
SEF AA

To assist the designer in selecting a compatible diameter-pitch combination,

he may make his choice from the following:
 77,
Section 15-2: Some Definitions

Diameter Range, in. Associated Pitches, threads/in.

From 1/2 to 11/16 incl. 20, 16, 12

Over 11/16 to 1 incl. 16, 12, 10

Over
| to 1 1/2 incl. 16, 12, 10, 8, 6

Over 1 1/2 to 2 1/2 incl. 16, 12, 10, 8, 6, 5,4

Over 2 1/2 to 4 incl. 16, 12, 10, 8, 6,5, 4

Over 4 to 6 incl. 125 10} S5Gh 55453

Over
6 to 10 incl. NG)5 G5 Sy hy Sh, 2 2

Over 10 to 16 incl. 1OF 8565554352 W221 /2ery/4

Over 16to 24incl. S516,

Se tae cly2ae Ll W2, li4ae

SECTION 15-2
Some Definitions
Before deriving the applicable design relations for power screws, we shall
redefine some important terms from kinematics relating to screw threads.
Pitch (p). Defined as the axial distance along an element of the pitch
cylinder, measured from one thread to an adjacent thread (see Figure 15-2).
Lead (/). The axial distance a nut will advance for one revolution of the
screw. Lead is measured along an element of the pitch cylinder. For a screw
with a single thread, the lead is equal to the pitch. For a multiple threaded
screw, the lead is equal to the product of the number of threads (n) and the
pitch (p). We therefore have
1 =np (15-1)
From equation (15-1), we can readily see that a multiple threaded screw will
advance a nut more rapidly than a single threaded screw of the same pitch.
Helix angle’ («). That angle of a thread formed by a plane drawn tangent
to the pitch helix and a plane drawn normal to the axis ofthe screw.
To understand how the lead, helix angle, and the pitch are related, the
reader should refer to Figure 15-7. If one visualizes a single turn of a square
thread as being equivalent to having wrapped an inclined plane around the

l=np
Figure 15-7 A developed screw
thread.

Mean thread circumference = 77d,

! This is also, at times, referred to as the lead angle.

 758
Chapter 15: Power Screws

root diameter of a cylinder, it will appear as shown in Figure 15-7 when

unwrapped from the cylinder. In Figure 15-7, the unwrapped inclined plane
is that which has a thickness equal to the difference between the screw thread
mean radius and the root radius. Thus, the unwrapped inclined plane has a
circumference equal to zd,,, where d,, is the mean screw diameter.
From Figure 15-7, we have
np
2 ——— (15-2)
td, d,,
Figure 15-8 identifies the aforementioned terms on a single, double, and triple
threaded screw.

Lead Lead

(a) Single Thread (b) Double Thread (c) Triple Thread

Figure 15-8 Multiple threaded power screws.

SECTION 15-3
Torque Equation for Power Screws
To assist us in deriving the torque equation, we shall use the model of the
screw jack shown in Figure 15-9. The platform that supports the load W is
an integral part of the screw and is prevented from rotating by being contained
between fixed walls as shown. The rollers between the platform and the walls
imply that the platform (and therefore the screw) cannot rotate but can move
freely up or down without any frictional resistance from the wall. The
platform and screw are raised or lowered by rotating the nut which is
supported by a thrust collar as shown in Figure 15-9.
Assuming that an outside force F applied at a mean radius r,, causes Wto
be raised, let us consider the forces acting on a portion of the thread as shown
in Figure 15-10. Our model shows all of the reactive forces acting at point O
on the surface of the thread that have a helix angle (that is, lead angle) x and
a thread angle @. The resultant force F, is normal to the thread surface, and
has vector components of interest: OD, OB, and OA. Vector OD is the
frictional force F, lying in the plane ofthe thread and is drawn in the direction
shown because the weight is being raised. If the weight were being lowered,
its direction would oppose the motion of the screw and it would have been
drawn opposite to that shown in Figure 15-10. Vector OA is equal in
magnitude, but opposite in direction to the load W. Vector OB is the resultant
 @ e
i
UO Platform
|}
|
@ oS

i NWA
f J. SI
0, Thread angle

= 2 ==: : oles for wrench

ym)

Thrust surface N SSS Se N ro; = collar inner radius

or collar \ eo = collar outer radius
*i Time = Mean collar radius

Thrust collar [ o = LSS

+ Teo
—————— ‘me :

Pein ares UIs}

Figure 15-9 Model of a power screw used as a screw jack.

Figure 15-10 Reactive force diagram on a thread due to raising a load W.

759
 760
Chapter 15: Power Screws

of OA and OD, and forms an angle 6, with vector F, (that is, angle 6, lies in
the plane OCB). If the coefficient of sliding friction between the nut and
screw is f,and we sum the forces in the vertical direction, we have
F,cos 6,cosa = W-+ F, sina
Replacing the friction force F, by its equivalent value f,F,, and then solving
the above relation for F,,, we obtain
W
=== 15-3

cos 8, cos a — f,sino

n

Next, we sum the moment of the forces about the centerline of the screw.
Thus, the torque Tp required to raise the load W (that is, move W up the
inclined plane) is
Ta = Frm = Vm Fp COS & + F, cos 0, sin a)
Replacing F, by f,F, results in

Tr = Tl fsFnCOS & + F, cos 0, sin a) (15-4)

However, the normal reactive force acting on the thrust surface due to W
results in an additional frictional force f.W—where f, is the sliding coefficient
of friction between the thrust surface (or thrust collar) and the surface that
supports the screw. It is assumed that this frictional force acts at the mean
collar radius r,,, (see Figure 15-9). The total torque required to raise W is
then equal to equation (15-4) plus r,,.f.W or
Tr = Wal Sef, COS + F, cost, sin a) + rae f.W

Replacing F, by equation (15-3) and noting that r,, = d,,/2, we get

T= d,W | f,cosa + cos 6, sina Amc t-W

R™ 2. | cos@,cosa
— f,sina 2
Dividing the numerator and denominator of the first term by cos results in

ee E + cos 6, tan | f Amc.t.W

In = z
nf 2 cos 0, — f,tana 2 )

where T, = the torque in inch-pounds needed to raise (that is, lift) the load
W, and d,, is in inches.
In applying equation (15-5), it is necessary to know 6, It can be obtained
from the trigonometric relationship between @,, 0, and a as follows: From
Figure 15-10 |
BC
tan 0 =
OB

But, BC = AE = OA tan 6 = (OB cos «) tan @ and therefore by substitution

 761
Section 15-4: Overhauling

we obtain

tan, = cosa tan 0 (15-6)

In most applications, the helix angle « is relatively small, and so cosa ~ 1.

Therefore, we can let 6, = 0 and equation (15-5) can be written as

yd,W |4+ cos

@ tan ‘| Aine.f-W
(P= 15-7
i 2 cos 8 — f, tana 23 ( )
The torque T,, required to lower load W is obtained in the same manner
as Tp. However, in this instance, both the applied force F and the friction
force F, act in a direction opposite to that shown in Figure 15-10. The reader
should also note that the frictional torque of the thrust surface is positive
because it always opposes the effort of the applied torque. The result of this
procedure leads to the equation for T,, the torque required to lower the load,
namely

_ d,W E — cos 0, tan | A dine.S-W (15-8)

T, =
z 2 cos 6, + f, tan a 2
Equations (15-7) and (15-8) can also be expressed in terms of the lead / by
replacing tan « by equation (15-2). The respective torque equations for raising
or lowering a load are

ee d,W | fdmn + cos 6, releme) pa

c 15:
= 2 EB cos 6, re] % 2 Te

af. =
dnW
m
| fd,
sum
— Icos 0 n dnt.
mc/c Sil

ra EB cos 0, =i pee Cea

Equations (15-9) and (15-10) can also be written in terms of the number of
threads n, and the screw pitch p, by noting that / = np.
For applications where a ball or roller thrust bearing is used in place ofa
thrust collar,f,can be considered as zero. The term d,,.f.W/2 can therefore
be eliminated from any of the above torque equations.

SECTION 15-4
Overhauling
In Figure 15-10, if the force F were removed and the helix angle « gradually
increased to a value such that the load W would slide down the inclined plane,
the screw would be performing an action called overhauling. In other words,
the raised load W would lower itself causing the screw to rotate. In such a
situation, an outside torque would be required not only to raise a load, but
also to maintain the load in its raised position. It can be shown that the torque
 762
Chapter 15: Power Screws

To ofa screw which is overhauling is

T dy W ta Tecticos
OF tana a [WV
{that is, Tg < 0)
ane cos 6, + f, tan 9)
(15-11)
In equation (15-11), if we neglect the collar friction (that is, because ofa thrust
roller bearing) and make 7, = 0, we can solve for a helix angle «. This helix
angle would then be the one that would permit the screw to overhaul. Thus,
from equation (15-11), since d,,W/2 and cos 6, + f, tan ~ cannot be zero, we
have

(BlnGy = Js (15-12)
cos 6,
If the parameters in equation (15-12) are such that f,/cos 6, > tana, then
the screw will not overhaul. A screw and nut (also a worm and wheel—see
Chapter 11) having this property are said to be self-locking. This is a desirable
characteristic in many applications (for example, a nonreversible drive).

SECTION 15-5
Screw Efficiency
The efficiency of ascrew may be defined as the ratio of the torque required
to raise a load W without friction to the torque required to raise the load W
with friction. Applying this definition to equation (15-5), we have for no
friction
WwW
(A=
R ) an a

Dividing this equation by equation (15-5) and cancelling W/2 from the
numerator and denominator, the efficiency of a screw is expressed as

ne d,, tan o (15-13

ar f, + cos 6, tan « ay, 1)
™| cos 6, — f, tan a ca

If the collar friction is negligible, the efficiency is

cos 6, — f, tana
= (15-14)
cos 6, + f,cot «

Representative curves showing how the efficiency varies with the helix
angle for an Acme screw are shown in Figure 15-11. The curves in this figure
are based on a screw without collar friction.
 763
Section 15-5: Screw Efficiency

100

80
Acme thread
= Ao

40
%
efficiency,
Screw

20

eae mantle =
0 10 20 30 40 50 60 70 80 90
Helix angle, deg.

Figure 15-11 Typical efficiency curves for Acme thread screws.

At best, the values for the coefficients offriction for various screw materials
are somewhat wide in range due to such factors as lubrication, surface finish,
tolerance, alignment, and load distribution. Table 15-4 contains values for
the coefficients of frictionf,andf,and is based on the assumption that the
screw assembly has been “‘well run-in.”’ Starting friction can be taken as 1.33
times the values listed in Table 15-4. In selecting friction values from Table
15-4, the lower number should be used for good workmanship, the higher
number for poor workmanship, and some in between number for other
gradations of work quality.
In some applications, particularly when a power screw is used as a linear
actuator (for example, instruments, motion control, mechanical computing
devices), it is desirable to eliminate or minimize backlash between the nut and
screw. A common procedure used to achieve this condition is to split the nut
in a plane normal to its rotational axis and then bolt the two halves together.
Tightening the bolts brings the two halves closer together axially and thereby
eliminates the clearance between the nut and the screw. This technique is also
used to eliminate increased backlash caused by normal wear over an extended
time period.

Table 15-4 Coefficients of Friction £ and f,

Screw Material Cast Iron

Steel (dry) 0.15-0.25 0.15-0.23 0.15-0.19 | 0.15-0.25

Steel (lubricated) 0.11-0.17 0.10-0.16 0.10-0.15 | 0.11-0.17

Bronze 0.08-0.12 0.04-0.60 0.06-0.09

 764
Chapter 15: Power Screws

Example 15-1. An Acme thread automobile screw jack is made of cold rolled 1045
steel having a diameter of 14 in. The thread is cut for general purpose applica-
tion, but the workmanship is of poor quality. When used to raise an auto-
mobile, the base of the screw is supported by a steel collar having a mean
diameter of 14 in. The nut that does the lifting is made of the same material
as the screw. To raise a 5000-lb car to change a tire, what would be the required
torque and the screw efficiency?
Solution: For a 13-in. diameter screw, we obtain from Table 15-1 the helix
angle « = 3° 33’, and basic thread height, h = 0.100000
in. From Table 15.1,
the coefficient of friction for a dry screw and nut made of steel and having
poor workmanship is f, = f, = 0.25. The coefficient of friction for starting is
1.33(0.25) = 0.33. For an Acme thread 0 = 14° 30’ (see Figure 15-1).
Using equation (15-6), we determine tan 0,,. Thus
tan 6, = cos «tan 8 = cos 3° 33’ tan 14° 30’ = (0.9981)(0.2586) = 0.2581
Therefore 6, = 14° 24.43’. Knowing the basic thread height and the basic
major diameter, we calculate the mean screw diameter
dm = 1.1250 — 0.1000 = 1.0250 in.
Applying equation (15-5) to initially calculate the starting torque, we have
7, (1.0250)(5000/4)[ 0.33 + cos 14° 24.43’ tan 3° 337] __1.5(0:33)(5000/4)
LET 2 cos 14° 24.43’ — 0.33 tan 3° 33’ 2
= 573.5 |b-in.
The running torque required to raise the car is based upon the coefficient of
friction f, = 0.25. Therefore, using the above calculation with the only change
being the friction coefficient, we find that
Tz = 443.25 in.-Ib
We see from these calculations that it requires a 29.3% higher torque to
start the screw in motion than it does to keep the screw in motion. We also
see that, if we assume a reasonable crank arm length, say 18 in., the operator
is required to exert a starting force of 31.9 lb (that is, 573.5/18). Not too many
people can continuously exert this force. Consequently, the screw jack will
have to be lubricated or have an improved quality of manufacture or a longer
crank arm.
The screw efficiency is determined from equation (15-13) as follows:
1.0250)(0.0620
eo ee ae
0.25 + 0.9685(0.0620)
1.025 of eae= ee a 5(0.
8.97%
This screw efficiency, which is relatively poor, can be markedly improved
by reducing the friction coefficient, increasing the helix angle, or by making
both changes. @

Example 15-2. In Example 15-1, will the screw overhaul after the load W has been
raised?
Solution: For overhauling to occur, it is necessary for the lowering torque,
equation (15-8), to be negative. Substituting the appropriate numerical values
 765
Section 15-6: Stress Considerations for Power Screws

of Example 15-1 into equation (15-8), we have

_ 1.025(5000)[ 0.25 — 0.9685(0.0620)] — 1.5(0.25)(5000)
aie 2 0.9685 + 0.25(0.0625) 2
1433 |b-in.
Since T;, is positive, the screw will not overhaul. @

Exampie 15-3. What helix angle would be required in order that the screw of Example
15-1 would just begin to overhaul? What would be the efficiency of a screw
with this helix angle?
Solution: For incipient overhauling to occur, T;, in equation (15-8) must
zero. Therefore, we have

0 = in| — Cos 6, tan | ue hp

cos 6, + f, tan o
Solving for tan

tan. a =
Is + fe COS On(4me/4m)
cos 6, = f.f(dine!
In)
Substituting the numerical values results in
_ 0.25 + 0.25(cos 14° 24.43')(1.5/1.025)
t os = 0.68
Bn cos42 0445" 109502515095)
a = 34.6°
Thus, for overhauling to occur, the helix angle must be at least slightly larger
than 34.6°.
Using equation (15-13) to determine screw efficiency, we have
1.02 165
a 0.25
COS ' 543’ aetan 34.6°34. a ied
ae E 14° 2433" — 0.25 tan ae if
= 45.5%

SECTION 15-6
Stress Considerations for Power Screws
In analyzing the strength of a power screw and nut, or any screw or nut,
the designer should be aware of the limitations of his calculations. We do
not mean to imply by this statement that the analytical expression used to
define a particular stress is no longer applicable to a screw. We do wish,
instead, to emphasize that the stress “picture” of the thread engagement
between the screw and nut is not completely understood. As an illustration,
it is commonly accepted that the load carried by a screw and nut is uniformly
distributed throughout the thread engagement. This is not, in fact, the case.
Goodier [5] showed that, due to deflection, only the first one or two threads
in engagement carry the major portion of the load. The remaining threads
carry lesser or greater portions of the load depending upon the elastic-
plastic deformation of the first couple of threads.
 766
Chapter 15: Power Screws

Goodier’s analysis also obviates the assumption that the bearing stress
is uniformly distributed throughout the thread engagement. However proper
lubrication can be a factor in support of this hypothesis.
Again citing Goodier’s analysis, the tensile stress of the screw across the
root area is greater at that point where the load ‘‘flows”’ from the screw
to the nut than it is at any other root area.
Lastly, we note that additional factors such as fillet radii, surface finish,
class fit, and so on, also have significant affects on the actual stress values.
Thus, the analytical relations that follow, although not exact, provide
the designer with some simple tools to help him achieve a reasonable
design—particularly when conservative design stresses are used.
The main stress considerations in designing a power screw are (1) bearing
pressure, (2) bending stress, (3) shearing stress, (4) tensile or compressive
stress, (5) combined stress, and (6) buckling stress.

Bearing pressure
The bearing pressure is the crushing stress between the surface of the
screw thread and the contacting surface of the nut. The relationship for
this stress is
Ww
OR (15-15)
— nd,,/n
where

oO, = bearing pressure (that is, crushing stress of projected thread area), psi
W = load, lb
dj, = Mean screw thread diameter, in.
h = depth of thread, in.
n = number of threads in engagement

Table 15-5 lists the design bearing pressures for some representative applica-
tions, materials, and speeds.

Table 15-5 Design Bearing Pressures for Screws

Type of Material Design Bearing Rubbing Speed at the

Mean Diameter of the
Service Pressure, psi
Thread

Hand press Bronze 2500-3500 | Low speed, well lubricated

Jackscrew Cast iron 1800-2500 | Low speed, not over 8 fpm

Jackscrew Bronze 1600-2500 | Low speed, not over 10 fpm

Hoisting screw ; Cast iron 600-1000 | Medium speed, 20-40 fpm

Hoisting screw Bronze 800-1400 | Medium speed, 20-40 fpm

Lead screw Bronze 150-240 High speed, 50 fpm and over

 767
Section 15-6: Stress Considerations for Power Screws

Bending stress
The bending stress in the thread is determined by. treating it as a short
cantilever beam with a built-in end at the root diameter (see Figure 15-12).
The load W is assumed to be uniformly distributed over the mean screw
diameter (that is, the load W acts at a point which is one half the distance
of the thread depth, h).
The cross-section of the developed thread at the root is a rectangle with
a beam depth of b and a beam width of zd,,n. We thus have a section modulus
for the thread of
I (xd,,n)b?
ee
C eee raaa (15-16 )

The maximum bending moment for this cantilever beam is

Wh
M = — (15-17)
2
The bending stress equation is
MC
6, = — (15-18)
I
Substituting equations (15-16) and (15-17) into (15-18), we arrive at

3Wh
6, = —— (15-19)
td,,nb?

where o, = the maximum bending stress in pounds per square inch, and all
other terms are as earlier defined.

Shearing stress
Both the threads on the screw and the threads on the nut experience a
transverse shearing stress due to bending. For a rectangular cross section,
the transverse shearing stress according to Section 5.6 in Chapter 5 (for
z= 0) 1S

Figure 15-12 Screw thread loaded as a cantilever beam.

 768
Chapter 15: Power Screws

where A is the cross-sectional area at the built-in end of the beam. This area
for the screw is zd,nb, and for the nut, the area is zd,nb. Thus, the transverse
shearing stress for the screw is
3W
5 2nd,nb ( )

The transverse shearing stress for the nut is

3W
= 15-21
‘ 2nd,nb (

where t is the transverse shearing stress in pounds per square inch, d, is

the root diameter of the screw, d, is the major diameter of the screw, and
the remaining terms are as earlier defined.

Tensile or compressive stress

Depending upon the particular design, a power screw may be subjected
to a tensile or compressive stress by the load W. The area used to calculate
the tensile or compressive stress is called the “tensile stress area” in the
various screw thread tables (see references [1] and [4]). This stress area is
slightly larger than the area that would be calculated by using the root
diameter. This larger area is determined by a circle having a diameter equal
to the average of the root and pitch diameters. The basis for using the
tensile stress area is found in the experimental results which have demon-
strated that a threaded tensile test specimen is stronger than a plain specimen
(that is, no threads) whose diameter is the same as the root diameter of the
threaded specimen.
We can now write the expression for direct tensile or compressive stress
as
W
Orore — v7 (15-22)

where
O;orc = the tensile or compressive stress, psi
nid, + d,\?\,
Ame | ine
4 2
d, = root diameter, in.
d, = pitch diameter, in.

Should the designer wish to be more conservative in his calculations, he

would be justified, then, in using d, only in calculating A.
 769
Section 15-6: Stress Considerations for Power Screws

Combined stress
When the unsupported length of the screw is so short that column action
can be ignored, the screw is then treated purely as a compression member.
In such an instance, the area of the screw which is at the root diameter is
subjected to a biaxial stress. Specifically, we have the combined stress
effects of equation (15-22) and the shearing stress caused by the applied
torque. Recall from Chapter 5 that the combined biaxial stress equation
employing the maximum shear theory of failure is

meee) (G) 2) en () (15-23)

The shearing stress t, caused by the applied torque is

= TG s/2- 1d,/2), 16T

J n(d*/32) x3 ieee
where T = the applied torque in pounds-inches. Substituting equations
(15-22) and (15-24) into equation (15-23), we get

w\? 167\2
Tmax = 7A Sts se, (15-25)

Because column action plays no role when the screw is subjected to a

tensile stress, equation (15-25) always applies when the screw is under
combined tension and torsion, regardless of the screw length.

Buckling stress
When the unsupported screw length is equal to or greater than 8 times
the root diameter, the screw must be treated as a column. As the reader
may recall, analysis of column stability leads to the Euler column formula,
which is restricted to long columns only. For ductile materials, short columns
are defined as those having a slenderness ratio* up to and including 100,
whereas columns having a slenderness ratio greater than 100 are considered
as long columns.* Short columns are assumed to fail due to compression
and bending on the concave side. Long columns fail due to buckling (that is,
instability).
There are a number of different short column formulas that extend the
Euler equation into the short column region. Many of these are intended
for structural applications, and are included as part of the requirements of
different codes (for example, building codes, and so on). In machine design,
a convenient short column formula, known as the Ritter equation, is often
? Slenderness ratio is defined as L/k, where L is the column length between supports, and k is
the least radius of gyration.
3 Some specifications use a slenderness ratio greater than 120 to define long columns. This
limit is particularly used for structural members.
 770
Chapter 15: Power Screws

used. This equation is a modification of the more widely known Gordon—

Rankine formula. The Ritter equation is

«= Al!+ (G)eel i
2

where

o. = the stress induced on the concave side of the column, psi. The
stress value should not exceed the design compressive stress of the
material.
S,, = the yield point of the material, psi.
A = cross-sectional area of the column, in.”
L = column length between supports, in.
k = least radius of gyration = ,/1/A, in. -
P = column load, lb.
E = modulus of elasticity, psi
K = factor depending on the end supports of the column; 0.25 for one
end fixed, and the other end free; 1 for both ends pin-connected;
2 for one end fixed, and the other end pin-connected; and 4 for
both ends fixed
Equation (15-26) is applicable only when the load is purely axial. If,
however, the load P is eccentric by a distance e from the column centroid,
the additional stress caused by the bending moment P, must be added to
equation (15-26). The resulting combined maximum stress equation becomes

CS

Te
Ayah Es aan k?
k] n*KE_
(15-27)
-

where c is the distance from the centroid to the outermost fiber of the column
in inches.
Although the use of long, slender screws should be avoided, it is sometimes
not possible to do so. In such cases, Euler’s column formula should be used.
As previously derived in Section 5.10, Euler’s column formula is
z Kr? AE
Pe als
where P., is the critical buckling load in pounds, and all the other terms are
as defined above.

SECTION 15-7

Ball Screws
A ball screw is a linear actuator that transmits a force or motion with
minimum friction. A basic ball screw assembly is shown in Figure 15-13.
As can be observed from this figure, there is a circular groove cut to proper
conformity with the balls. The groove has a helix angle that matches the
 771
Section 15-7: Ball Screws

Figure 15-13 Ball screw assembly showing the component parts. [Courtesy
Saginaw Steering Gear Division, General Motors Corporation. ]

helix angle of the ball nut. When the screw and nut rotate relative to each
other, the balls are diverted from one end of the ball nut and are carried
by ball guides to the opposite end of the ball nut. Such recirculation permits
unrestricted travel of the nut relative to the screw.
Because the balls roll, as in a ball bearing, the frictional losses are minimal.
This characteristic is one of the important advantages that ball screws
have over threaded screws. Efficiencies of 90% or higher are possible with
ball screws over a wide range of helix angles when converting rotary to
axial motion because of minimum friction (see Figure 15-14). When a ball
screw is used to convert axial to rotary motion, efficiencies as high as 80%
can be expected, even when the helix angle is so small as to cause threaded
power screws to be self-locking.

Ball-bearing screw

Figure 15-14 Comparative efficiencies of an

Acme screw and a Ball screw. [Courtesy
Saginaw Steering Gear Division, General
Screw
%
efficiency,
Motors Corporation.]

Helix angle, deg

 Viz
Chapter 15: Power Screws

Comparing ball screws with threaded power screws we can cite the
following advantages and disadvantages :

Advantages
. High efficiency—approximately 90% or higher.
. Predictable life expectancy.
. Precise positioning and position repeatability permitted.
. No tendency to “‘stick-slip’’ as with screws.
. Negligible heat expansion due to friction.
. Easily preloaded to. eliminate backlash at virtually no frictional
increase.
. Because starting torque is low, smaller motors can be used.
. Smooth movement and control over its entire stroke.
. The overall silhouette of the ball nut is much smailer because the balls
can withstand higher loads than screw threads.

Disadvantages
Us Requires lubrication to realize its maximum useful life.
2. Ball screws, because of their high efficiency, tend to overhaul. Thus, if
self-locking is required, an outside brake must be added.
The inclusion of dirt or foreign matter in ball screws can reduce or
even end their life.
Because the load carrying capacity of a ball screw is greater than an
equivalently sized power screw, it is not as stiff as a power screw.
Consequently, the deflection and critical speed of a ball screw can
become significant factors for certain applications.
The uses of ball screws are as varied as threaded power screws. Typical
of such uses are
1. Machine tools—numerical control positioning, feed control, table
drives, and slides.
2: Airplanes—flap and slat drives, landing gear actuation, canopy
actuation, air inlet controls, and thrust reverses.
Weapon systems—elevation actuation on cannons, missile and rocket
launches, and leveling jacks.
Materials handling—balance hoists, conveyor drop sections, die tables,
and crane boom actuators.
Miscellaneous—jacks, antenna leg actuators, hospital bed articulation,
valve operators, instrument drives, and antenna drives.

PROBLEMS
1. A 13-in. diameter square thread screw 3 ft long is used as an automobile bumper
jack. When in use, the base of the screw is supported by a 24-in. diameter collar
that rests on a recessed metal plate. The screw, nut, collar, and metal plate are
all made of steel and operate without lubrication. If the workmanship is considered
 V7.
Problems

poor, calculate (a) the torque required to lift 1500 lb, (b) the length of crank required
if 20 lb is considered a reasonable force that can be exerted by an operator, (c) the
torque required to lower 1500 lb, (d) the efficiency of the screw when lifting the
load, (e) the efficiency of the screw when lowering the load, (f) the torque required
to lift the load if a thrust ball bearing were used in place of the plain collar, and
(g) the efficiency of the screw when lifting the load if a thrust ball bearing is used.
2. Solve Problem 1 for an Acme thread having the same outer diameter.
3. (a) For a power screw with a ball bearing thrust collar (that is, little or no collar
friction) determine the expression for the maximum efficiency when raising a load.
(b) Assuming a coefficient of moving friction of 0.15, what is the maximum efficiency
for a square threaded screw? (c) For the same friction coefficient, what is the maxi-
mum efficiency for an Acme thread? As a check, your results should coincide
with the value from Figure 15-11 for the same helix angle.
4. For the screw of Problem 1, determine (a) the minimum length of nut engagement
required if the permissible bearing pressure is 1800 psi,* (b) the factor of safety of
the threads in bending if all of the components are made of AISI 1030 cold drawn
steel, (c) the factor of safety of the nut and screw in shear, and (d) the factor of
safety of the screw in buckling. Assume that the bumper is supported by the nut
extension at a distance of 4 in. from the centroidal axis of the screw.
5. A sluice gate of a dam is raised and lowered by two 3-in. diameter modified square
threads, that are run by a gear reducer, motor, and brake combination. The
nuts, which are attached to the gate, are made of bronze and the screws are made
of steel. Because each screw is supported by a thrust ball bearing, collar friction
can be neglected. The workmanship of the assembly is considered to be very
good and requires only a moderate amount of grease for lubrication. If the gate,
which weighs 50 tons, is to be raised or lowered at 2} ft/min, determine (a) the
revolutions per minute of the screws, (b) the motor horsepower required to lift
the gate assuming that the overall mechanical efficiency of the motor and reducer
Isto Uae
6. A toggle device for a baling machine is driven by a gear reduction motor as shown
in the Problem Figure 6. The two cast iron non-rotating nuts move axially in
opposite directions against two 6000 lb forces. The opposite directional motion of
the nuts is due to a left and right hand thread cut in the screw (see figure). The
screw, which is made of steel, is cut to a standard 23 in. nominal outside diameter
square thread and rotates at 80rpm. The workmanship is considered excellent
and the assembly will be lubricated. Calculate the motor horsepower required if
the mechanical efficiency of the gear reduction motor is 85 per cent. Because ball
bearings are used at the three bearing supports, the frictional torque at these
points may be neglected.
R.H. thread

L.H. thread

6000 Ib 6000 Ib
Problem Figure 6

Power in

4 See Table 15-5. Although a steel screw and steel nut combination is not listed, it is assumed
that a steel screw and cast iron nut combination permit comparable pressures.
 774
Chapter 15: Power Screws

ze Show that a screw just on the verge of overhauling when lifting a load has an
efficiency of about 50%.
8. A vertically mounted linear power actuator has a single thteaded Acme screw with
a nominal outside diameter of 2 in. A split nut (2 in. long) which is kept from rotating
moves along the axis of the screw at a rate of 5 ft/min while raising a 6-ton load.
The screw is supported on the bottom end by a thrust ball bearing with an I.D. of
1.3780 in. and an O.D. of 1.457 in. and on the other by a deep groove ball bearing.
The screw is made of steel and the nut of cast iron. The workmanship is considered
to be of the highest quality. If the screw and nut are lubricated, determine the
following :(Note: The coefficients of friction for roller bearings are listed in Table
9-4). (a) The starting torque. (b) The running horsepower. (c) The running efficiency.
(d) The thread bearing pressure. The length of the nut is 2 in.
. Repeat Problem 8 for a double threaded stub Acme thread.
. A 2-in. outer diameter modified square thread screw is suggested for use in a
power press which is to exert a maximum force of 6 tons. The screw is to be
made of AISI 1118 carburized steel and the nut, which is to be locked into the
upper press platen, is to be made of ASTM class 50 cast iron. The workmanship
is expected to be of the highest quality. To minimize torque requirements, a
thrust ball bearing is to act as a collar in supporting the screw. Both the screw
and nut will be lubricated. Determine (a) the running torque required to act against
the maximum load, (b) the efficiency of the screw system, (c) the height of the nut
required for a thread bearing pressure of 1000 psi, (d) the number of threads on
the nut, (e) the bending stress of the thread and the factor of safety, (f) the shear
stress of the thread, and nut and their factors of safety, and (g) the combined stress
of the screw and the factor of safety based on the maximum distortion failure theory,
and the maximum shear failure theory.
11. A vertical actuator is used to raise and lower a platform of a paper feeder. A 1-in.
diameter modified square thread is cut in the screw which is supported on the
lower end by a thrust ball bearing. Both the screw and the nut are made of steel
and are lubricated. The workmanship 1s considered to be of average quality. When
the platform rises with a full paper load, the total weight is 5000 Ib. When the empty
platform is lowered, it weighs 1500 lb. Determine the following mean and varying
stresses : (a) the thread bearing pressure for a 15-in. long nut, (b) the bending stress
of the thread, (c) the shearing stress of the thread, and (d) the combined stress of
the screw. No column action is considered in this problem since the screw is
assumed to be short.
If the platform rises at the rate of 2 ft/min, what size gear motor drive (that is,
horsepower) would be required if the mechanical efficiency of the drive is 85°%?
12. Problem Figure 12 shows a schematic drawing of a screw jack which is widely
used in construction, in rigging machinery, and in equipment for many applications
requiring the movement of heavy equipment. Design a screw jack similar to that
which is shown, capable of lifting 15,000 1b to a maximum height of 20 in. The
screw is to have a square thread and is to be made of AISI 1040 cold-drawn steel.
The base is to be made of ASTM class 40 cast iron. The threads, which are cut
in the base, will run lubricated with the screw threads. The quality of manufacture
and workmanship is to be assumed as being average. Because infrequent use will
be made of the jack, the bearing pressures listed in Table 15-5 can be doubled
(for example, in this case use 3600 psi to determine the nut height h).
If 50 1b is considered a reasonable force that a man can exert, what should the
length of the turning lever be? What is its diameter if it is made of AISI 1040 cold
drawn steel?
In determining the length of the turning lever, keep in mind that the starting
friction is one third greater than the running friction (see Section 15.5).
 775
References

Load

Cap for the load

Thrust ball bearing

G oe GY Turning lever

Problem Figure 12
Base

[1] Acme Threads—ANSI Standard B 1.5—1972. Stub Acme Threads—ANSI Stan-

dard B 1.8—1972. Buttress Threads—ANSI Standard B 1.9—1972.
[2] Harold A. Rothbart (ed.): Mechanical Design and Systems Handbook. McGraw-Hill
Book Co., New York, 1964.
[3] Colin Carmechael (ed.): Kent’s Mechanical Engineers’ Handbook, Design and
Production Volume, 12th ed. John Wiley & Sons, Inc., New York, 1961.
[4] National Bureau of Standards Handbook H28, Part II. U.S. Department of Com-
merce, Washington, D.C., 1957.
[5] J. N. Goodier: The distribution of load on the thread of screws. J. Appl. Mech.,
Trans. ASME, 62, 1940.
 Fasteners
SYMBOLS

A = cross-sectional area P = allowable load

C = stiffness coefficient p = pressure
d = diameter t = thickness
de; = effective hole diameter w = width
E = modulus of elasticity S = allowable stress
F = applied force 6 = deflection
F, = initial load o = actual stress
K = stress concentration factor t = actual shear stress
k = stiffness constant

In discussing fasteners or connectors, we will make a distinction between

permanent fasteners, such as rivets and weldments, and detachable fasteners,
such as screws, cotter pins, keys, and snap rings. The permanent fasteners
will be discussed first, and because weldments are discussed in Chapter 17,
we will start with a consideration of riveted connections.

SECTION 16-1
Axially Loaded Rivets
Although there are many ways of classifying rivets, the first distinction
we will make is between axially loaded and eccentrically loaded rivets. In an
axially loaded riveted connection, the line of action of the applied force
passes through the center of gravity of the rivet group, whereas for the
eccentrically loaded connection it does not. At the moment we will concern
ourselves only with the axial loaded problem. Further, although we will
consider rivets as the connector to be analyzed, it should be pointed out
that the analysis applies equally well to bolts. Finally, although it is true
that a bolt is a form of screw, we are considering the bolt, as used in these
applications, to be a permanent rather than a detachable fastener.

776
 I
Section 16-1: Axially Loaded Rivets

Figure 16-1 Steps followed in forming a rivet.

Rivets are used in many applications, such as for boilers, buildings,

bridges, ships, and so on. Because of the danger to human life that could
result from a failure of such applications, the design of riveted connections
are rigorously governed by construction codes such as American Institute of
Steel Construction (AISC) Code, American Railway Engineering Association
Code, and Boiler Construction Code of the American Society of Mechanical
Engineers. Clearly, the code applicable to the particular problem on hand
should be followed. When rivets are used in a machine design, there is no
code, which means the designer is left to his own resources with regard to
method of analysis, materials, and so on. It is not feasible to discuss each of
the codes with regard to rivet analysis; therefore, to illustrate one approach
to the problem, we will in general, follow the AISC structural code.
Before proceeding further, it would be desirable to define a rivet. It may
be described as a cylindrical body, known as the shank, with a rounded
end called the head. When used to join two plates together, the rivet is
heated to a red glow, and the shank is placed in the hole that has been
made in the plates. If the rivet is heated before being placed in the hole, it

Figure 16-2 Single riveted lap joint.

 778
Chapter 16: Fasteners

is known as a hot-driven rivet, whereas if it is not heated, it is known as a

cold driven rivet. The preformed head is then supported, as shown in Figure
16-1, while a hammer or some other means of applying pressure is used to
form the shank into another head. Upon completion of this operation, the
rivet will have two heads and look something like the one shown in Figure
16-2. Thus, the purpose of the rivet is to join two plates together while
ensuring proper strength and tightness. The holes in the plates are either
punched or drilled.

SECTION 16-2
Methods of Rivet Failure
In analyzing the various methods by which a riveted connection may fail,
we will first consider the stress situations that exist for a single rivet. Let us
assume that the two plates to be joined overlap each other and are held by
a single rivet as shown in Figure 16-2.
If we assume that the forces act along the center of each plate, it is clear
that the loading is eccentric and an unbalanced moment Ft exists. Figure
16-3 is an exaggerated drawing of the effects of this unbalanced moment.
The effect of the moment is to put a crushing or bearing load on the corners
of the plate, subject the rivet shank to a tensile stress, and cause a shear
stress in the rivet heads. It should also be clear that, since a moment exists,
there will be a flexural or bending stress Mc/I present. However, because of
a lack of information concerning how the load is applied, and therefore
doubt about what the magnitude of the moment is, the usual procedure is
to ignore the bending stress and to compensate for its presence by a larger
factor of safety.
The rivet itself may fail because of the shear stress that is developed in it.
The rivet in Figure 16-4 is subjected to single shear (one cross-sectional
shearing area), whereas the one shown in Figure 16-5 is subjected to double
shear (two cross-sectional shearing areas). The way failure will occur for
each of these methods 1s also illustrated in each figure.
The shearing stress present in the rivet is given by equation (16-1).

Pad (16-1
nd?/4 a
where d = diameter of rivet, and F = force applied. If the rivet is in double
shear, the area md?/4 is simply multiplied by 2. If a connection has more

Figure 16-3 An exaggerated drawing of the effect of the unbalanced moment

acting on a single riveted lap joint.
 779
Section 16-2: Methods of Rivet Failure

Figure 16-4 Rivet in single shear.

than one rivet, the usual assumption made is that each rivet takes an equal
share of the load. The actual shear stress is now compared to the allowable
shear stress. Clearly, the allowable must be equal to or greater than the
actual.
Another possible method of failure is a tensile rupture of either plate.
Figure 16-6 shows this situation. The tensile stress can be calculated from
equation (16-2).
F
0,
s (w — dere)t Ca
where w is the width of plate, d.,, is the effective diameter of the hole, and
t is the thickness of the plate.
An explanation of effective diameter is now in order. Rivet holes are
usually either drilled or punched. When the holes are drilled, they are usually
made in. larger than the rivet diameter to ensure that the hot rivet will
easily fit into the hole. On the other hand, when the holes are punched,
some damage occurs to the material around the edges of the hole, and this

ee a
(me

Gene
eas
a
Figure 16-5 Shearing failure in double shear.

Figure 16-6 Tensile failure of plate.

 780
Chapter 16: Fasteners

Initial position Final position

of rivet of rivet

Figure 16-7 Compressive or crushing failure of plate commonly known as a

bearing failure.

damaged material is unable to resist the load. As a result, an additional

jis in. is added to the rivet diameter in order to obtain the effective diameter.
To sum up, the effective diameter for a drilled hole is 7 in. larger than the
rivet diameter, and the effective diameter of a punched hole is in. larger
than the rivet diameter.
The third method by which failure may occur is by a crushing of the plate
or the rivet. This compressive failure is commonly referred to as a bearing
failure. Figure 16-7 depicts this type of failure. The usual assumption made
in calculating the bearing stress is that the load is distributed over the
projected area of the rivet. The bearing stress is thus given by
le
=F (16-3)
where d = diameter of rivet and t = thickness of plate.
Once again, for those problems involving connections containing more
than one rivet, the usual assumption made is that each rivet takes an equal
share of the load when bearing and shear failure are considered. The
assumption is reasonable because yielding of rivets which are slightly
overloaded will tend to distribute the load equally.
Another mode of failure is the shearing of the edge of the plate, as shown
in Figure 16-8. This failure is usually designated as a shear tearout. As can
be seen from the figure, shear tearout is dependent on how far a rivet is
located from the edge of the plate. The shear tearout stress is given by

18
ae (16-4)
where a = closest distance from rivet to the edge of plate and t = thickness
of plate.
The structural code specifies minimum edge distances, so that shear
tearout and tearing of the edge (Figure 16-9) do not have to be checked.
 781
Section 16-3: Lap and Butt Joints

Figure 16-8 Shearing tearout or shearing failure of edge of plate.

For example, the code would specify that the distance from the center of a
rivet to a plate edge parallel to the line of action of the force be at least 1.5
times the rivet diameter, whereas the distance from the center of the rivet
to the plate edge normal to the line of action of the force must be at least
2 times the rivet diameter. The 1.5 and 2 mentioned above are simply
average examples because the actual values as specified by the code depend
upon the magnitude of the rivet diameter.
To sum up, there are basically three modes of failure that must be
considered, shearing of the rivets, bearing failure of the plate or the rivet,
and tensile failure of the plate.

SECTION 16-3
Lap and Butt Joints
Axially loaded riveted joints are of two types, lap joints (Figure 16-10)
and butt joints (Figure 16-11).
A comparison of Figures 16-10 and 16-11 makes clear why the joints
are called lap and butt joints. In a lap joint, the two plates to be joined

mn

e ;

Figure 16-9 Failure due to edge of plate tearing out.

 782
Chapter 16: Fasteners

Figure 16-10 Riveted lap joint.

overlap each other, whereas in a butt joint the two plates to be joined (the
main plates) butt against each other. The minimum edge distances m men-
tioned earlier are shown in Figure 16-10.
Another important geometric property is the pitch, which is defined as
the distance between adjacent rivet centers. We usually speak of the pitch
for rivets along a line parallel to the edge of the plate, whereas the corre-
sponding distance along a line perpendicular to the edge of the plate is
known as the transverse pitch, p,. Both types of pitch are shown in Figure
16-10.

Figure 16-11 Riveted butt joint.

 783
Section 16-3: Lap and Butt Joints

It is useful, in comparing different joints, to have some idea about how

efficiently they connect the plates in question. The usual comparison used is
the efficiency of the joints as obtained from equation (16-5).

ed Pitowable max
i (16-5)
strength of plate with no hole

where Pajowablemax = the smallest of the allowable loads in shear, bearing,

and tension, and strength of plate with no hole = the tensile strength of the
plates to be joined if there were no holes in them.
Clearly, the most efficient joint possible would be one which will be as
strong in tension, shear, and bearing as the original plate to be joined is in
tension. This can never be achieved, because there must be at least one rivet
hole in the plate, and the joints’ allowable load in tension will thus always
be less than the strength of the plate with no holes.
It should be obvious at this point that most joints will have many rivets
in a variety of patterns. The analysis required for such joints is much more
complicated than that for a single rivet presented previously. We will
demonstrate the procedures that are followed in order to solve these more
complicated problems by means of the following examples. It is strongly
suggested that they be read over very carefully.

Example 16-1. Two 15-in. by 4-in. plates are to be connected by means of a lap
joint. The rivets to be used are {-in. diameter, and the working stresses are
as follows: shear, 13,500 psi; bearing, 27,000 psi; tension, 18,000 psi. Design
the connection.
Solution: The connection should be designed for maximum efficiency, and,
since it is usually assumed that the row of rivets immediately adjacent to the
load carries the full load, the maximum tensile load carried by such a row
occurs when only one rivet is in that row. The portion of the load carried by
an interior row can be determined from equation (16-6).

Te actual) = d Ji (16-6)
n
where Pactual) = Portion of externally applied load P acting on a particular
row; n = total number of rivets in the connection; and n’ = total number
of rivets in the rows between the row being checked in tension and the external
load. The use of this equation will be illustrated later on in this example.
The strength of the plate with one rivet is

F,, = Syatiow(W — dere)t

Since the rivets are %-in. diameter and the holes must be assumed to have
been punched, dir = 3 + § = Lin.
F,, = 18,000(15 — 1)(4) = 126,000 Ib
The strength of the joint in bearing and shear depends on the number of
rivets.
 784
Chapter 16: Fasteners

Strength of one rivet. In shear

T
Fs = 554’)

= 13,500 x =4 x (? = 81101b
In bearing
fp = S,td = 27000 xi xs
= 11,800 lb
Therefore rivets are weaker in shear and the number of rivets necessary is
126,000
Sor 15255

Thus 16 rivets are necessary in order to permit 126,0001b to be applied.

Clearly, with 16 rivets, the allowable load in bearing will be greater than
126,000 Ib.
The remaining problem now is to determine how the 16 rivets are to be
distributed. In doing this, it should be made clear that, for lap joints having
equal plate thicknesses, the last row of rivets (farthest from the load) in the
upper plate is the first row (row closest to the load) of the lower plate. Therefore,
the first and last row will have one rivet. The situation is as shown in the
figure below.

Let us next check the situation in row 2 if there are two rivets in that row.
Load on row 2. By equation (16-6)
rane eOe=1)
Be aciael) a n P= 16 126,000

= 118,000 lb
Strength of plate row 2
P, alow) = 18,000(15 — 2)5 = 117,000 Ib
The actual load is slightly higher than the allowable, but the difference is
small enough to be acceptable. Therefore, rows 2 and 6 will have 2 rivets.
Let us next check row 3 if three rivets are in it.
Load on row 3. By equation (16-6)
16 — 3
Peace = 76 126,000

= 102,500 lb
Strength of plate in row 3
P,yatlow) = 18,000(16 — 3)3 = 108,000 lb
Since 108,000 > 102,500 row 3 is satisfactory. Therefore, rows 3 and 5 have
three rivets.
 785
Section 16-3: Lap and Butt Joints

We have so far accounted for 12 rivets, let us therefore check row 4 with
four rivets in it.
16-6
P (actual) = 765 126,000 = 78,000 lb

Pallow) = 18,000(15 ma 45 => 99,000 Ib

Row 4 satisfactory.
We should also check the spacing requirements for row 4. Assume that
the minimum center-to-center distance between rivets is 3d (AISC code); as
mentioned previously the minimum edge distance is 2d. Therefore
required width = 3(3 x %) + 2(2 x %)
=O + 9 = y= ligin.
Since this is less than 15 in., the spacing requirement is met. The final rivet
pattern is thus satisfactory. ,
All that remains is to calculate the efficiency of the connection.
P.
= allow max

tensile strength of plate

126,000 _ 126,000
78000 x 15.4 1350007 °° ~

Example 16-2. A butt joint, whose rivet pattern is shown in Figure 16-12, is to be
used to form a circular steel boiler. The boiler is to have a 5-ft diameter and
be able to withstand an internal pressure p of 500 psi. Determine the required
main plate thickness as well as the efficiency of the joint. The rivets are to
be j-in. diameter, and the rivet holes are drilled to +2 in. in diameter. The
working stresses are as follows: tension, 20,000 psi; shear, 16,000 psi; bearing,
20,000 psi in single shear and 28,000 psi in double shear.
Solution: The rivet pattern shown is but a small section of a much longer
pattern. Clearly it would not be feasible to attempt to analyze the joint by
considering the whole length of rivets. The usual procedure is to determine
the minimum repeating section, which is the smallest length for which each
row of rivets repeats itself exactly. The rivets in row A and B repeat themselves
for a length equal to the distance between two adjacent rivets, but the rows C
and D do not. If the distance between adjacent rivets in row C were used as
the repeating length, rows A, B, and C would repeat their pattern, but D
would not. However, if the distance between adjacent rivets in row D, L,
were used, all four rows would repeat their pattern over this length. Thus,
L is the minimum repeating section for the rivet pattern shown. There are
eleven rivets in the repeating section, as can be clearly seen if section L to
the right of row A is considered.
Because the shear strength is independent of plate thickness, the maximum
allowable load the joint can resist is determined by the shearing of the rivets.
The rivets are in double shear so that 2 x 16,000 psi will be used as the
allowable stress.
Pp = Tt 2
S(allow) — NS scatow) 44

7)

11(2 x 16,000)5(7
= 212,000 Ib
 786
Chapter 16: Fasteners

Figure 16-12

The next step in our solution will be to determine how long the repeating
section will be. In Figure 16-13, a free body diagram of a length of the boiler L
and with the resulting cylinder slit in half lengthwise is shown. The force T
represents the tensile load acting in the boiler plates.
Summation of forces in the vertical direction yields
Dilie—op2 Rie
or
IP
~ pR

Figure 16-13
 787
Section 16-3: Lap and Butt Joints

Because the load cannot exceed 212,000 lb in shear, we will let T, the tensile
load, be 212,000 lb. Then
_ 212,000
L = 14.1in.
~ 500 x 3 x 12
In order to determine the thickness of the main plate, the larger of the
thicknesses as determined by bearing and tension is used.

Pratiow) = MSpattow) tp
212,000 = 117 28,000 x |x 7,
t, = 0.79 in.

Tension row D

P, = Syattow(W — Gers)tp
212,000 = 20,000(14.1 — 1a)
tp = 0.81 in. or fin.

Actual load acting on row C is

— 1
Pactual) = A J? 11 212,000

I| 193,000 Ib

Strength across row C is

P, = Stattow(W — dere )ty

= 20,000(14.1 — 2 x 4#3)33
= 200,000 Ib
Since 200,000 lb > 193,000 Ib, the plate is safe in tension across row C.
Actual load on row B

ile 8)
Jet(actual) = x 212,000 = 154,000 Ib

Strength across row B

P, = Statlow)(W = dere)ty

= 20,000(114.1 - 4 x #3) x #
= 168,000 Ib
Since 168,000 lb > 154,000 Ib, the plate is safe in tension across row B.
Row A will clearly be safe since the strength is the same as for row B, and
the actual load acting is less than the actual load on row B. Therefore, the
required main plate thickness is + in. The efficiency of the connection remains
to be calculated.
eae IPallow max
~ tensile strength of plate
. 212,000
20,000 x 14.1 x 2 =926% @
 788
Chapter 16: Fasteners

SECTION 16-4

Eccentrically Loaded Rivets

Up to this point in our discussion of riveted connections, we have con-
sidered only those problems in which the rivets were axially loaded (the
load passed through the center of gravity of the rivet group). However,
many examples exist for which the load does not pass through the center
of gravity of the rivet group. A typical problem encountered in structural
work is the situation that occurs when a horizontal beam is supported by a
vertical column by means of a gusset plate, see Figure 16-14. As is shown
in the figure, the left hand reaction on the beam is F, this force is transmitted
to the gusset plate, which in turn transmits the force to the column. It is
obvious from an inspection of Figure 16-15 that the rivets will be subjected
to a twisting moment as well as a direct shearing force.
Making use of the law of statics, which says that a force may be moved
to a parallel coplanar position by replacing it by an equivalent force and
couple, we replace the actual force F (solid line) by the equivalent force F
(dotted), acting through the center ofgravity of the rivet group, and equivalent
couple Fe.

Channel

4b
ot
43eT

Figure 16-14 Sketch of horizontal beam attached to a

vertical channel by means of a gusset plate.
 789
Section 16-4: Eccentrically Loaded Rivets

Figure 16-15 Free body diagram showing the twisting, direct, and resultant
forces acting on the rivets and due to the gusset plate.

If we assume that each rivet takes an equal share of the load (a reasonable
assumption, since the force now acts through the center of gravity of the
rivet group) we can draw a force vector through the center of each rivet,
acting in the same direction as F, and having a magnitude of F/n, where n
is equal to the number of rivets in the group. (For the group shown in
Figure 16-15, n = 9.)
The couple, Fe, which tries to rotate the plate about the center of gravity
of the rivet group, is prevented from doing so by the resistance of the rivets.
As a result, there are forces, T, acting through the center of each rivet, and
perpendicular to the distance d, drawn from the center of gravity to the
center of each rivet.
 790
Chapter 16: Fasteners

Let us turn our attention for a moment to the determination of the

magnitude of these turning forces. Assuming for the moment that the plate
can be rotated about the center of gravity through a small angle 0, it is
clear that the displacement a rivet would go through would be directly
proportional to its distance d. It follows that the unit strain resulting from
such a displacement would be directly proportional to d. Because we may
assume stress proportional to strain, the stress is also directly proportional
to d. Finally since the force T varies directly as the stress, it follows that the
force T varies linearly and directly as the distance d.
Clearly, the external moment Fe must be resisted by the sum ofthe internal
moments Td. In other words,
Fe = Id; ae Tod, =F ae St Tae. (16-7)

But since T is directly proportional to d it follows that

Te Dl ee
eS tie ee d,,
Therefore
d, as d
Do2
1 Ge
=
3}
T,— “Gh
yi
n
1 a8

Then substituting in equation (16-7), we obtain

d d d d,
Fe = Pe + Ty + Tee feet Tes ,

Beever
So ea Biase
ao daa Maen?
+ d5)
1

or
Fed,
Te 3 2 2
dj +d; +d,+>--+a@ 2

Similarly
F
ie ke
t= 1Ai
and, in general

Fed,
Pai :
eos
Be29
The final step is to find the resultant force on the rivets, in other words,
to combine algebraically or graphically the force vectors F/n and T.
Fortunately, it is not necessary to find the resultant on each of the rivets,
because the usual procedure is to make the rivet diameters the same. It is
 To
Section 16-4: Eccentrically Loaded Rivets

therefore only necessary to calculate the maximum resultant load. An

inspection of the vector force diagrams is usually sufficient to eliminate all
but two or three rivets as candidates for worst loaded rivet.
As an example, consider the rivet group shown in Figure 16-15. Obviously,
rivets 4 and 5 have smaller resultants than rivet 6; rivets 1 and 7, 2 and 8,
and 3 and 9 have equal respective resultants. Also, it is clear from the vector
diagrams that 3 and 9 have a larger resultant than the other pairs mentioned.
Therefore, all that is necessary is to calculate R; and R¢, in order to deter-
mine the maximum resultant load.
Although a shearing failure of the rivets is usually the only type of failure
that must be checked, in some problems the gusset plate thickness should
also be checked. In other words, the resistance to a bearing failure must be
determined.
Before ending this discussion of rivets with an example, it would be well
to reiterate a statement made earlier: If bolts or other removable fasteners
had been used instead of rivets, the analysis presented would be equally
valid.

Figure 16-16

oln
 792
Chapter 16: Fasteners

Example 16-3. Determine the maximum allowable load F that can be applied to the
rivet group shown in Figure 16-16 considering shear failure of the rivets only.
Diameter of rivets is gin. and allowable shear stress is 16,000 psi.

Solution: It is clear from an inspection of the force diagrams, that rivet A

has the greatest resultant load acting on it. It should be obvious that rivet B,
the rivet at the center of gravity, has no twisting force acting on it.

T=T=g—s
The eccentricity e can be found by considering triangle ABD since
ig:
Sas onl Sane

Therefore

paoceeees
ow o265F
The vector sum of T, and F/3 can be obtained either graphically or alge:
braically.
wat
<F,(vertical) = 0.265F + 0.333F x a

= 0.265F + 0.176F
= 0.441F

&F ,(horizontal)= 0.333F x wee

14.2
=O281F
Ry = \/(0.441F)?? + (0.281F)
= 0.523F

F (citowableinshear) = SSeatiow) * ; x (3)? = 9630 Ib

Therefore
0.523F = 9630
or
F = 18,400 lb maximum allowable load @

Example 16-4. A problem somewhat similar to eccentrically loaded rivets occurs

with eccentrically loaded bolted connections. The resulting stress in the bolts
is due to both shear and tension. To demonstrate an approach to the solution
of this type of problem, consider the following example.
Figure 16-17 shows a machine part that is to be bolted to a support by means
of two rows of bolts. Each row also has two bolts. If the bolts are sin—10
(j-in. outside diameter, 10 threads/in.), having a stress area of 0.334 in.2, an
allowable tensile stress S, of 20,000 psi, and an allowable shear stress, St of
11,000 psi, determine the maximum load F that can safely be applied.
 793
Section 16-6: Standardized Screw Threads

Solution: The force F produces a simple shear stress in the bolts, and a
reasonable assumption would be that each bolt takes an equal share of the
load. The load on each bolt is therefore F/4. The part of the bolt being subjected
to shear is unthreaded, therefore the shear area is (m/4)(3)* = 0.442 in.* for
direct shear,

7P_ = 044g
F/4 = 0505F

The effect of the moment F x 15 is to cause the machine part to pivot about
point A. Therefore the force produced in the bolts due to this moment will be
proportional to the distance from the center line of the bolts to A. The resulting
equation is
Mad
T, ==>
eae

Clearly the bolts in row 2 will have a greater load than those in row 1.

T= Md, _ F x 15 x (8 + 4)
2 (d? +d2) ~~ (2 bolts x (4)? 2 bolts. <(12)7]
= 0.562F

Therefore the tension force due to the moment on each bolt in row 2 is 0.562F.
The maximum tensile stress occurs in the threaded portion of the bolt, thus

The combined stress equations can now be used to determine the maximum
tensile and shear stresses

on om 2

Onmax) — 5) ar »

_ LOOF | ea 2
+ (0565F?
2 2

= 1.87F = Siatow)

20,000 = 1.87F

F = 10,700 lb

Tmax = x/(6,12)? + 7? = s/(1.69F/2)? + (0.565F)

=> 1.02F => S s(atlow)

11,000 = 1.02F

F = 10,000 lb

Therefore maximum allowable load = 10,000lb. @

 794
Chapter 16: Fasteners

Figure 16-17

SECTION 16-5
Screw Fasteners
The discussion presented in the earlier part of this chapter dealt with
permanent fasteners, rivets and bolts. Since most machine parts that require
connection must usually be made in such a way that they can easily be
disassembled as well as assembled, nonpermanent fasteners are required.
These so-called detachable fasteners come in a variety of types, and the
one chosen by the designer for a particular application depends primariiy
upon the major requirements of the problem. For example, a machine part
that is subjected to vibrations must be fastened by a screw and nut arrange-
ment that will not shake loose under vibration. The first, and perhaps the
most often used, fastener we will discuss is the screw.

SECTION 16-6
Standardized Screw Threads
The various types of threads used on screws have been standardized, and
it is important that the designer be aware of the various types available
and what their important characteristics are. Power screw threads are of
an entirely different type, and will not be discussed in this section.
One of the earliest types of screw threads was the V thread, Figure 16-18.
However, the sharpness of the crest made the thread very susceptible to

60°
Pitch (p)

Figure 16-18 V thread type of screw

thread.
 7)
Section 16-6: Standardized Screw Threads

60°
ceed Pitch (p)—=}

Figure 16-19 Sellers thread, in which the sharp crests and roots of the
the V thread are replaced by flat surfaces.

damage. In addition, the sharp root resulted in large stress concentrations.

It became obvious that the V thread would have to be modified in order
to relieve the conditions mentioned.
The Sellers thread, Figure 16-19, alleviated the problem by replacing
the sharp crests and roots with flat surfaces. Another solution was the
Whitworth thread, Figure 16-20, in which the crest and root were rounded.
The Sellers thread was formerly the United States standard, whereas the
Whitworth thread was the former British standard. Both countries now
have as their standard the unified thread, Figure 16-21.
The major or outside diameter is measured at the crests of the threads
for external threads and at the roots for internal threads. This is the diameter
used to designate a thread. As an example, a 3-10 screw has an outside
diameter of #in., and 10 threads/in. The pitch of a screw is defined as the
distance between corresponding points of adjacent threads. It is equal to
the reciprocal of the number of threads per inch. The 3-10 screw has a pitch
of 0.1 in.
The lead is the distance a screw moves axially during | revolution of the
screw. As was mentioned previously, in the section on worm gearing, the
lead of a single threaded screw is equal to its pitch, a double threaded screw
has a lead equal to twice its pitch, a triple threaded screw has a lead 3 times
its pitch, and so on.
The following threads are the most commonly used ones, and are thus
designated as “standard” threads.

Figure 16-20 Whitworth thread, which further modified the Vv

thread by rounding the crests and roots.
 796
Chapter 16: Fasteners

Min major dia. specified in

dimensional tables t SRS
8
S SS
a8H Internal s
S
aS
SS W Thread
SS
: SS
NN
SS

aces
SATS
RY
by
.
Pitch line
S _
SS
p

= S
ys
5
5H

eee ee

ex
° 22 RQ.
Nea
2
SSS “SS
S
ev
SS
\
H \ / MB /
a N 4 \ /
\l7/ 7

90° \

Axis of internal thread = 7

{

A
External a \
Thread fs
gZ
2 ZaRZ
2
_Pitch line ZN
>

2
D, fs—p
<A

H ZA <A,
|
4 Zeaie Max minor dia. at Z)
Y intersection of root Vv Minor dia.
contour and flank specified in dimensional
of thread tables

| Rounded root optional or

90° resulting from tool wear

{ Axis of external thread

Figure 16-21 Unified internal and external screw thread (maximum material
condition) design forms. [From American Standard Unified Screw Threads (ASA
B1.1—1960), with the permission of the publisher, The American Society of
Mechanical Engineers, New York.]}

1. Coarse (designated as UNC). Used for general assembly work,

especially where frequent assembly and disassembly is necessary. Should
not be used for designs in which vibration problems may be encountered.
They are also recommended for use in nonsteel metals. A good rule of thumb
for the designer to follow is ‘“‘always specify coarse threads unless a specific
factor of design indicates otherwise.”
2. Fine (designated as UNF). This series is widely used in automotive,
aircraft, and other applications where vibrations are likely to occur. They are
also recommended for use in designs involving holes tapped in steel.
 19)
Section 16-6: Standardized Screw Threads

3. Extra-Fine (designated as UNEF). Used in instrument applications,

particularly in aircraft work, and also for problems involving large vibrations
and shock. The screws in this series are usually made of a high-grade alloy
steel.
4. There are several standard constant pitch threads available. As the
name implies, these are threads that have the same pitch for all diameters.
As can be seen by looking at Table 16-1, the pitch for the threads discussed
previously increases as the diameter increases. As a result, it may be impossible
to tighten large bolts enough to induce the required initial tension.

Table 16-1 Tables of Basic Dimensions, Standard Series Threads

(a)—COARSE THREAD SERIES, UNC AND NC-—BASIC DIMENSIONS
a ; Lead f
Minor Minor Angle Section
Basic Basic é
J : iameter | Diameter . at Minor Tensile
Major Pitch pad at Basic
Sizes
:
Diameter,
*
Diameter,
Extemal | Internal, 5
Pitch
Diameter Stress
Threads, | Threads, Diameter, at oe
K,, r D-2h, In,
in, ines

1(0.073) 4 0.00218 0.00263

2(0.086) 4 0.00310 0.00370
3(0.099) 4 0.00406 0.00487
4(0.112) 4 0.00496 0.00604
5(0.125) 4 0.00672 0.00796
6(0.138) 4 0.00745 0.00909
8(0.164) 3 0.01196 0.0140
10(0.190) 4 0.01450 0.0175
12(0.216) 4 0.0206 0.0242
1/4 - 0.0269 0.0318
5/16 3 0.0454 0.0524
3/8 3 0.0678 0.0775
7/16 3 0.0933 0.1063
1/2 3 0.1257 0.1419
9/16 2 0.162 0.182
5/8 2 0.202 0.226
3/4 2 0.302 0.334
7/8 2 0.419 0.462
] 2 0.551 0.606
1-1/8 Z 0.693 0.763
1-1/4 2 0.890 0.969
1-3/8 2 1,054 1.155
1-1/2 2 1.294 1.405
1-3/4 2 1.74 1.90
2 2 2.30 2.50
2-1/4 1 3.02 Sao
2-1/2 1 3.72 4.00
2-3/4 i 4.62 4.93
3 1 5.62 5.97
3-1/4 1 6.72 7.10
3-1/2 1 71:92 8.33
3-3/4 1 9.21 9.66
4 1 10.61 11.08

SOURCE: Extracted from American Standard Unified Screw Threads (ASA B1.1-1960) with permission of the publisher The
American Society of Mechanical Engineers, New York.
 798
Chapter 16: Fasteners

Obviously this problem can be alleviated by holding the pitch constant,

as the diameter is increased. Some of the standard constant pitch threads
available are 8 pitch (8 UN), 12 pitch (12 UN), 16 pitch (16 UN), and 20 pitch
(20 UN). Some of the applications for which constant pitch threads are
useful are high pressure pipe flanges, cylinder-head studs, heavy machinery,
and so on.
In designating threads, the letter A is used for external threads, and B is
used for internal threads. Threads are further classified according to fit.
Class 1 fits have the widest tolerances and thus are the loosest fits, they are
used when rapid assembly. and disassembly is required. Class 2 fits are the
most widely used and are recommended for most applications, except when
contraindicated for some specific reason. Class 3, the final standard fit,
is the one having the tightest fit and is used for precision applications. A
variation of these fits can be obtained by using a different class of fit for
the external and internal threads.

Table 16-1 Tables of Basic Dimensions, Standard Series Threads

(b)—FINE THREAD SERIES, UNF AND NF-BASIC DIMENSIONS

Lead
Minor
; Minor
; Angle Section
;
Diameter | Diameter Aes at Minor Tensile
ae External | Internal Pitch Diameter Stress
Threads, | Threads,} Diameter ; at Area .
Ken D-2h, in.”
in.?

0(0.060) 4 0.00151 0.00180

1(0.073) 3 67 0.00237 0.00278
2(0.086) 3 45 0.00339 0.00394
30.099) a oe 0.00451 0.00523
4(0.112) 2 Gil 0.00566 0.00661
5(0.125) 2 26 0.00716 0.00830
6(0.138) 3 44 0.00874 0.01015
8(0.164) 5 of 0.01285 0.01474
10(0.190) 4 DI 0.0175 0.0200
12(0.216) yy 0.0226 0.0258
1/4 y 0.0326 0.0364
5/16 2 40 0.0524 0.0580
3/8 % ils 0.0809 0.0878
7/16 A iG 0.1090 0.1187
1/2 ES 7 0.1486 0.1599
9/16 il — 35 0.189 0.203
5/8 it ee 0.240 0.256
3/4 1 36 0.351 0.373
7/8 1 34 0.480 0.509
1 i sts 0.625 0.663
1-1/8 1 aS 0.812 0.856
1-1/4 elo 1.024 1.073
13/8 1 9 1.260 1.315
1-1/2 1 3 1.521 1.581
SOURCE: Extracted from American Standard Unified Screw Threads (ASA B1.1-1960) with permission of the publisher The
American Society of Mechanical Engineers, New York,
 799
Section 16-7: Various Types of Screws, Bolts, and Other Fasteners

Table 16-1 Tables of Basic Dimensions, Standard Series Threads

(c)—EXTRA-FINE THREAD SERIES, UNEF AND NEF—BASIC DIMENSIONS

Mi nor Miinor Rie ection

Secti
Basic : : Angle : /
Pitch Diameter | Diameter at Basic at Minor Tensile
hay Fe c External | Internal Pitch Diameter Stress
lameter,! Threads, , |Threads
reads, | Diameter, 2t Area
Kr D— 2h, in.?

12(0.216) 0.0270
1/4 0.0379
5/16 0.0625
3/8 0.0932
7/16 0.1274
1/2 0.170
9/16 0.214
5/8 0.268
11/16 0.329
3/4 0.386
13/16 0.458
7/8 0.536
15/16 0.620
1 0.711
1-1/16 0.799
1-1/8 0.901
1-3/16 1.009
1-1/4 1.123
1-5/16 1.244
1-3/8 1.370
1-7/16 1.503
1-1/2 1.64
1-9/16 1.79
1-5/8 1.94
1-11/16 2.10
SOURCE: Extracted from American Standard Unified Screw Threads (ASA B1.1-1960) with permission of the publisher The
American Society of Mechanical Engineers, New York.

Example 16-5. What type of screw does the following designation represent,
2 in.-10 UNC-3B-LH?
Solution: ¢in. outside diameter, 10 threads/in., unified coarse thread,
class 3 fit, internal thread, left hand thread. @

SECTION 16-7

Various Types of Screws, Bolts, and

Other Fasteners
The only difference between a screw and a bolt is that the bolt needs a
nut in order to be used as a fastener, whereas a screw fits into a threaded
hole. Some of the more common types of fasteners as well as some special
application types are shown in Figures 16-22 and 16-23.
 800
Chapter 16: Fasteners

Figure 16-22 Shock absorber bolt. A rubber cap is molded

over and under the ribbed head of this steel bolt. The ribs on
the top or under side of the head prevent the molded rubber
from turning or loosening. [Courtesy Elco Corporation,
Rockford, IIl.]

Figure 16-23 Some typical examples of stud bolts. A stud bolt has threads on
both ends and is particularly useful in applications where a plate is to be bolted to
a larger part and a through bolt cannot be used. A prime example for which a stud
is useful is the head on an automotive engine. [Courtesy Camcar Screw and Mfg.
Division.]

Nuts
Many varieties of nuts exist, both the common variety as well as ones for
special applications. It would be impossible to discuss all the types that have
been developed, much less to consider the many new types that are constantly
being developed. Some typical examples are shown in Figures 16-24 through
16-26.

Screws
Sems are preassembled fasteners that eliminate the need for hand assem-
bling of screws and washers. Some common types of sems are shown in
Figure 16-27.
 Figure 16-24 The self-locking fastener pictured is a
locknut that makes use of a nylon locking patch (the
lighter portion of the threads). This type of nut is useful
for those applications involving severe shock and
vibratory loads. [Courtesy Amerace Esna Corporation,
Union, N.J.]

Figure 16-25 A standard lock washer of the type

shown has maximum resistance to loosening since the
teeth are formed and twisted to bite deeply into both
nut and work surfaces. [Courtesy Shakeproof, Division
Illinois Tool Works, Inc.]

Figure 16-26 (a) The speed nut fastener shown is a

self-locking device that provides both an inward thread
lock and an arched spring lock. (b) The principle upon
which speed nut fasteners operate. [Courtesy Eaton,
Yale, and Towne.]

(a)
Prelocked position Double-locked position

Arched prongs Inward thread lock

Arched base Arched

spring lock
(b)

801
 802
Chapter 16: Fasteners

Type BA Type F Pilot point Taptite shown with

Dae fe a coned washer

Figure 16-27 Various types of preassembled fasteners known as Sems are

shown. [Courtesy Camcar Screw and Mfg. Division.]

Set screws are a type of semipermanent fastener used to prevent relative

motion between sliding surfaces. In general, they are useful in low torque
applications involving rotary motion. Set screws are made in a variety of
heads and points. Some of the common types of set screws are shown in
Figure 16-28.
It is extremely important for the designer to choose the proper type of
point. We will now briefly discuss some of the types and applications for
which they are used.
1. Cone Point. Used for designs where the parts to be joined will have
the same relative position with respect to each other. Standard cone angles
are 90 and 118 deg. The usual procedure is to make the cone angle 118 deg
when the set screw length is less than or equal to its nominal diameter,
whereas the angle is 90 deg when the length is greater than the nominal
diameter.
The following rules of thumb are suggested for all types of set screws.
The size of the set screw is usually taken as }-4 times the shaft diameter it is
to fasten. The minimum length of engagement for a set screw is about 1-14
times the diameter if the set screw is threaded in brass and 3-1 times for use
in steel.
The torque capacity can be increased by using two set screws. If they are
used side by side, the torque capacity is almost doubled, whereas if they are
placed 180 deg apart the capacity is increased by about one third.
2. Cup Point. Perhaps the most commonly used type of point because of
the ease and rapidity of assembly. It is rarely used in applications involving
hardened shafts.
 803
Section 16-7: Various Types of Screws, Bolts, and Other Fasteners

aw) <> <-> LSP

Sap —S Nin / Sy er
=
Flat Oval Cup Half Cone
point point point dog point

Figure 16-28 Common types of set screws. [Courtesy Amerace Esna

Corporation, Union, N.J.]

3. Flat Point. Useful for applications involving hardened shafts and where
frequent adjustment is needed.
4. Full Dog. Used for those designs that involve permanent relative
location of parts. It is necessary to drill a hole in the part that is to receive
the set screw.
5. Half Dog. Same as full dog except for length of insert.
6. Oval Point. Similar to cup point, except that a groove for spotting
the set screw is usually needed on the receiving part.
The tensilock screw, shown in Figure 16-29, has a rolled nylon washer
preassembled to a standard screw.

Rolled edge
prevents
outward flow
Controlled flow
forces nylon into
all openings

~
Before After
compression compression

(a) (b)

Figure 16-29 (a) This fastener consists of a rolled nylon washer, preassembled
to a standard screw. Because the outside rolled edge of the washer is stronger than
the inside edge, it resists outward flow during compression and the inside edge is
forced inward into the screw threads and into any space between screw shank and
hole. As a result the fastener seals, locks, isolates vibrations, is corrosion resistant,
and provides electrical insulation. (b) Sketch showing how fastener operates.
[Courtesy Nyltite Corporation, South Plainfield, N.J.]
 (a) (b)
Figure 16-30 Self-drilling fasteners, known as TEKS® are shown. They drill
their own hole, tap a mating thread, and then fasten, all in a single operation.
(Courtesy Shakeproof, Division Illinois Tool Works, Inc.]

Figure 16-31 A standard helicoil

insert used to prevent thread failures
caused by high tensile and torsional
stresses, eliminate corrosion and seizure,
and prevent stripping. [Courtesy Heli-
Coil Products, Division Mite
Corporation. ]

804
 805
Section 16-7: Various Types of Screws, Bolts, and Other Fasteners

Thread-cutting screws, as the name implies, cut their own threads, thus
eliminating the need for tapping. Basically, they cut threads because they
are slotted off center, presenting a sharp acute cutting edge to the area being
threaded. The screws shown in Figure 16-30 are typical and are used to cut
threads in (a) metal, (b) castings, (c) plastics, (d) sheet metal, (e) wood.

Figure 16-32 Various types of clamps. [Courtesy Wittek Manufacturing

Company. |}
 806
Chapter 16: Fasteners

Helicoil inserts are screw thread coils that are inserted into tapped holes
to salvage threads which have been damaged or stripped. They may also
be used to protect and strengthen tapped threads. A typical insert is shown
in Figure 16-31. When helical inserts are used to replace a damaged thread,
the following steps are generally followed.
1. A larger hole must be drilled.
2. This hole is then threaded by tapping.
3. The helicoil is inserted.

Other fasteners
Clamps are used to attach hoses, flexible plastic pipes, and ducts. Some
examples of the wide variety of clamps available are shown in Figure 16-32.
Spring clips are relatively light duty fasteners that can perform multiple
functions and are one piece devices that eliminate the need for handling
several smaller parts. There are no standard clips because most are designed
to meet specific requirements. Some typical spring clips are shown in Figure
16-33.
Pins are fastening devices, usually used in applications where the loading
is primarily shear. They are useful in such applications as a hinge, shear
pin, retaining shoulder, and so on. Some common examples of pins are
shown in Figure 16-34.
Retaining rings, sometimes called snap rings provide a removable shoulder
for accurately locating, retaining, or locking components on shafts and in
bores and housings. Some applications for which retaining rings are used
are shown in Figure 16-35.

(a) (b) (c)

Figure 16-33 (a) The “C” clip is used for attaching plastic or die cast knobs.
(b) A capacitor clip used as a fastener on electronic equipment. (c) A cable clip.
(Courtesy Eaton, Yale, and Towne.]
 807
Section 16-8: Stress Analysis of Screws and Bolts

Figure 16-34 Various types of

grooved pins. [Courtesy John Gillen
~ Corporation, W. Cicero, IIl.]

(a)

Type

Three tapered grooves along the full Three tapered grooves for one-half the Three straight grooves for the full
length of the pin. length of the pin. length of the pin with starting pilot.

Type Type Type

D E F

SS) ae
Three reverse tapered grooves for Three center grooves covering one- Three straight grooves for full length
one-half the length of the pin. half of the length of the pin. of pin with starting pilot on both ends.

Type Type Type

G H x)

eH (i (|—— })
Three half length tapered grooves Three half length reverse taper Three center grooves with annular
with annular groove at end. grooves with annular groove at end. grooves at each end.

(b)
SECTION 16-8
Stress Analysis of Screws and Bolts
In doing a stress analysis of a bolt or screw, the first step is to calculate
the required diameter by using the direct tension or compression formula,

od
where F is the load to be supported and A is the cross-sectional area based
upon the root diameter. If bending or column failure is possible, the diameter
should be checked for these conditions.
Another possible method of failure would be the threads stripping.
Figure 16-36 is an illustration of a nut and bolt being used to clamp two
plates together. It is clear from an inspection of the figure that the length
of engagement of the nut will affect the strength of threads in shear and
bearing.
 (b)

QHR2
 809
Section 16-8: Stress Analysis of Screws and Bolts

Figure 16-35 Opposite: (a) Representative ring types. “Internal” rings are used
in bores and housings; “external” rings on shafts, studs, bosses, and similar parts.
(b) Automotive rocker arm. Two different types of retaining rings solved a
critical problem for the manufacturer of this aluminum roller-tip needle bearing
rocker arm for racing cars. The rings reduce the size and weight of the rocker arms
and prevent costly field failures. The rocker arms oscillate through a 0:560-in.
range at a frequency of up to 5000 times per minute. Because of the oscillation, the
roller-tip must be secured in the housing with positive-locking fasteners that will
resist inertia and vibration forces.
(c) Truck transmission. Two-part interlocking retaining ring secures third-speed
gear on main shaft of truck transmission. Balanced to withstand high revolutions
per minute the ring is installed radially in a deep groove on the shaft and must
withstand heavy thrust loads created when the gear is engaged. The fastener also
must remain secure against relative rotation which occurs when gear is disengaged
and rotates at different speed from shaft. Ability to withstand wear and abrasion is
another quality standard for the application.
[Courtesy Waldes Kohinoor, Inc., Long Island City, N.Y.]

If we assume that the threads take an equal share of the stress, equations
(16-9), (16-10), and (16-11) can be used to calculate the shear stress on the
screw and nut threads and the bearing stress between the threads.

F
Tnut) = nd(h/2) (16-9)

F
Ciscewy = nd,(h/) (16-10)

= (16-11)
7» (n/4yd? — d?\(h/p)
where t = shear stress, o, = bearing stress, d = outside diameter of screw,
h = height of nut, d, = root diameter of screw, and p = pitch of screw.
Unfortunately, the threads do not take an equal share of the load. In
fact, a conservative estimate would be that one thread takes all the load.
However, yielding of the threads will allow some of the load to be transferred
to other threads. It would, therefore, be desirable to choose a soft nut
material so that the yielding and consequent spreading of the load can take
place.

V7

Figure 16-36 An illustration ofa

nut and bolt being used to clamp two
plates together.
 &10
Chapter 16: Fasteners

To sum up, because it is not clear how many of the threads are in engage-
ment and can share the load, the designer can either assume all threads
share the load equally (with this assumption, a large factor of safety should
be used) or all of the load is taken by one thread (a very conservative assump-
tion). In any event, a reasonable length of thread engagement can be approxi-
mated by the following rules of thumb.
1. For steel, the length should be at least equal to the outside diameter
of the thread.
2. For cast iron, brass, or bronze the thread engagement should be at
least 14 times the outside diameter of the thread.
3. For aluminum, zinc, or plastics it should be at least 2 times.

SECTION 16-9
Materials for Screws and Bolts
A designer has an almost endless choice when it comes to selecting the
material of which a fastener is to be made. It is therefore imperative for the
designer to choose the material to suit the particular application.
Some of the factors that must be evaluated are the following: What will
be the environment in which the fastener is to be used? Should the material
have special electrical or heat conductivity properties? Is the fastener to be
opened and closed often? Will it be subjected to vibrations? Is weight going
to be a factor? What is the life expectancy to be? And other similar questions.
The designer should also bear in mind that coatings may very often be
applied to more economical materials so that they may be used in special
applications, in place of a more expensive material. Table 16-2 lists the
properties of some commonly used steel fasteners, as well as nonferrous
fasteners. As an aid to the designer, the materials suggested in Machine
Design Fasteners issue, which are typically used for certain kinds offasteners,
are presented in Table 16-3.
Steel is the material of which most fasteners are made. Table 16-4 lists
some of the grades of SAE steel with the marking system used to identify
the various grades. Aluminum is often used in fasteners because of its
relative low weight, high electrical and heat conductivity, resistance to
corrosion, and good appearance. Brass is a relatively inexpensive material,
is quite malleable, can be given a very high finish, is nonmagnetic and is
resistant to corrosion. Copper is extremely malleable, has very high electrical
conductivity, and is very corrosion resistant. Nickel is corrosion resistant,
does not discolor, and has good strength at high temperatures. Stainless
steels are useful in applications where corrosion, temperature, and strength
may be a problem. Finally, there are some nonmetallic materials, such as
plastics, which are very corrosion resistant, are good electrical and heat
insulators, and are relatively light weight.
Table 16-2 Properties of Commonly Used Metal Fasteners
C1015-C1018 Steel—Used for machine screws and wood screws, and a variety of industrial gases and chemicals. It can be readily
and specials where material is specified as low carbon and is not cold worked and is non-magnetic and non-sparking.
heat treated. The guaranteed minimum tensile strength is 55,000 T305 Stainless Steel (18-8)—Has excellent corrosion resistance,
psi. Low carbon steel is also used for sheet metal and thread- heat resistance and high mechanical properties but has a fairly high
cutting screws where no mechanical properties are specified by rate of work hardening. This enables parts only with rather modest
the customer, but a case hardened surface is required. upsets to be made. This type has slightly better corrosion resistance
C1019-C1021 Steel (Same as C1022 Steel)—Most parts made from than 16-18 and is lower in cost,
these grades are sems and thread-cutting screws requiring high 16-18 Stainless Steel—Type 16-18 stainless is a modified austenitic
torque and toughness. These are killed steels in the medium low chrome-nickel steel that is particularly useful for cold-headed and
carbon group and have closely controlled mechanical properties upset parts because it work hardens much slower than any of the
such as tensile strength, hardenability, and torque. conventional T305 (18-8) analyses. Type 16-18 resists a wide
C1029 Steel—Used in applications similar to C1038 when better variety of organic and inorganic chemicals and food-stuffs. Its cor-
cold working properties are needed. rosion resistance compares to the 300 series.
C1038 Steel—This grade is considered a medium carbon steel and T410 Stainless Steel-The most widely used of all hardenable
is used on Grade 5 bolts and special parts requiring a minimum types of corrosion resisting steels. It is capable of attaining high
tensile strength of 120,000 psi. mechanical properties when heat treated. For maximum corrosion
€1062 Steel—A high carbon grade with limited cold workability. resistance, T410 should be used in the hardened condition. Since
It is used in applications where extremely high strength and high it is also ductile and workable, it is particularly suited to cold
hardness (SO Re and above) are required. heading. Not recommended for applications in which severe cor-
4037-H Alloy Steel—This grade is in the medium carbon, low rosion is encountered.
alloy group. Parts made of 4037-H have a high yield to tensile 7430 Stainless Steel—A ductile, non-hardening stainless steel ex-
strength ratio and very consistent properties, Used for Grade 7 hibiting mechanical properties higher than mild steel with good
and Grade 8 bolts and special applications requiring 133,000 psi resistance to general corrosion and oxidation at temperatures up
minimum tensile strength. to 1400°F. All forms are magnetic.
Brass—An alloy of copper and zinc, brass is a tough, rustless, non- Aluminum—While there are many grades of aluminum used, the
magnetic material possessing mechanical properties similar to low most common for us is 2024. It actually meets the tensile strength
carbon steel. It can be easily formed into intricate shapes and is of mild steel at only .the weight, 2024 is fabricated in the H-13
relatively inexpensive compared to other copper base alloys. temper but heat treated to T4 temper in finished stage.
Copper—Copper in various grades such as Electrolytic Tough Pitch,
Nickel +400 (Monel)—Monel is approximately 5 Nickel and 3
Oxygen Free and Tellurium bearing is used extensively in archi-
tectural, automotive, electrical, and hardware applications. It can copper with a small amount of iron. Monel makes an excellent
be easily formed by many methods and has relatively high cor- fastener material because of its high strength, resistance to heat
rosion resistance. and corrosion and good cold workability. Monel has wide use in
Silicon Bronze—High tensile strength superior to mild steel. It has chemical, electrical, marine and construction applications.
corrosion resistance to salt and fresh water, atmospheric conditions

Gade |100 | 2024-74 6061-T6

Al, 99.0 min Cu, 3.84.9 Mg, 0.8-1.2
Mn, 0.3-0,9 Si, 0.4-0.8
Aluminum Alloys Composition p Mg, 1.2-1.8 Cu, 0.15-0.40
Al, Balance Cr, 0.15-0.35
Al, Balance

Physical properties
Density Ib/in.?
Melting temperature range, °F 935-1180 1080-1200
Ther cond. (77°F, ann)
Btu/hr/ft/°F/ft
Coef of therm. exp (68-212°F) per °F ISalE xl One 1216< 105° 13.0 x 10°*

Spec heat (212°F) Btu/Ib/°F

Elec res (68°F) microhm—cm annealed 3.45 (0), 5.75 (T4) 3.8 (0)
Mechanical properties
Tensile strength 13,000 55,000 (T4) 45,000 (T6)
Yield strength (0.2% elong) 5,000 50,000 (T4) 40,000 (T6)
Hardness (Rockwell) F20-25 B53-64 B60

Grade » Inconel

Ni, 77.0
Nickel Alloys Composition » Cr, 15.0
Fe, 7.0

Physical properties
Density Ib/in.? 0.319 0,304

Melting temp. Range, °F 2370-2460 2540-2600

Ther cond (80-212) BTU/hr/ft? /°F/ft 8.7
Coef of ther exp (77-212), per °F 7.8 X 10~6 6.4 X 1076
Spec ht (70-750°F) Btu/lb/°F 0,127 0.109
Elec res (68°F) microhm-cm 482 98.1

Magnetic Slightly magnetic |Nonmagnetic

Mechanical properties
Tensile strength 80,000 min. 80,000 min.

Yield strength (0.2% offset) 60,000 min. 45,000 min.

Hardness B90 B72 min.

811
 Table 16-2 (continued)
Grade

Steel Material »

Physical properties
Density Ib/in.? 0.283 0.283
Melting temp range, °F 2750-2775 2750-2775 2750-2775 2700-2750 2700-2750 —
Coef of ther exp (70-1200°F), per °F . 8.4X 10-6 8.4X 10-6 Ka
Spec ht, Btu/Ib/°F 0.10-0.11 0.10-0.11 0.10-0.11 0.10-0.11 0.10-0.11
Elec res (68 F), microhm-cm 14.3 14.3 19
Magnetic Yes Yes Yes e Ss

Mechanical properties
Sea
Tensile strength No req. $5,000 69,000 110,000 120,000 150,000
Yield strength 85,000 85,000 120,000
Brinell hardness 241 max 207-269 241-302 302-352
Rockwell hardness B100 max B9S-104 C23-32 C32-38

Stress Min. te tempering | Min. tempering

Other requirements
Relieved temp. 800°F | temp. 800°F
Copper
Commercial
Copper Alloys Grade D> (Electrolytic Yellow Brass Cartridge Brass
Bronze
Tough Pitch)
'

Physical properties Composition >

Cu, 99.9 min. Cu, 63-68.5 Cu, 68.5-71.5 Cu, 89-91
O02, 0.04 Zn, Bal Zn, Bal Zn, Bal

Density Ib/in.? 0.321-0,323 0.306 0.318

Melting temp range °F 1949-1981 a n % 1710 1870-1910
5
Ther cond (77°F, ann) Btu/hr/ft? /°F/ft 226 67 0 109
Coef of ther exp (68-572°F) per °F 9.8 X 10-6 11.3 X 10° 11.1
X 107% 10.2 X 107°
Spec heat (68°F) Btu/Ib/°F 0.092 0.09
Elec res (68°F )microhm-cm 1.71 64 62 3.9
Mechanical Properties
Tensile strength 35,000-45,000 60,000 min 60,000 min 45,000 min
Yield strength (0.5% elong) 10,000-35,000 40,000 min 40,000 min 40,000 ‘min
Rockwell hardness F40-80 BSS-65 BSS5-65 B40-S0

Copper Alloys Grade p> Naval Bronze Low Silicon High Silicon Phosphor
SS 3 3 is ® o Bronze A Bronze A
Cu, 62-65 Cu, 96 Min ; Cu, 95
Physical properties Composition p Sn, 0.5-1.0 Si, 0.8-2.0 mn ae pe
Sn, 3.5-5.8
Zn, Balance Zn, 1.5 Max i, 2.7-3.75 P, 0.03-0.35
Density Ib/in.*
Melting temp range °F
004 0.320
1750-1920
Ther cond (77°F, ann) Btu/hr/ft? /°F/ft
Coef of ther exp (68-572°F) per °F
re iS 47
9.9 X 10-6
Spec heat (68°F) Btu/Ib/°F Oo i=)oO 0.09
Elec res (68°F) microhm cm os h 9.6
Mechanical Properties
Tensile strength 50,000 min 70,000 min 70,000 min 70,000 min
Yield strength (0.5% elong) 30,000 min 35,000 min 38,000 min 40,000 min
Rockwell hardness B45-55 B65-75 B74-80 B70-80

812
 Table 16-2 (continued)
Grade” 430

C 0.12 max C 0.08 max C 0.08 max C0.15 max C 0.12 max
Approximate Mn 2.00 max Mn 0.50 max Mn 2,00 max Vem eaten Mn 1.00 max
Stainless Steel Chemical Si 1.00 max Si 0.50 max Si 1.00 max S100 max Si 1.00 max
Composition Cr 17.00-19.00 Cr 16.00-18.00 Cr 11.50-13.00 Cr 11.50-13.5
Cr 14.00-18.00
Ni 10.00-13.00 Ni 14.50-18.50 Ni 14.00-16.00

Physical properties
Density Ib/in.*

Mean coefficient of thermal expansion per °F X 107°

32-212°F

32-600°F
32-1000°F
32-1200°F 6.6
Melting point range 2700-2790°F 2600-2750°F
Magnetism (*in fully annealed condition) Magnetic
Mechanical properties

Brinell, Hardness
130-150 100-130 140-160
Annealed

Ultimate, Tensile strength, psi

Annealed 80,000-95,000 70,000-90,000 70,000-90,000 65 ,000-85 ,000 70,000-90,000
Cold worked 100,000-180,000 | 85,000-160,000 | 80,000-160,000 100,000 90,000-130,000
Yield, point, psi
Annealed 35,000-45 ,000 30,000-40,000 25,000-35 ,000 35,000-45 000 40,000-55 ,000
Cold worked 50,000-150,000 45 ,000-120,000 45 ,000-1 20,000 85,000
Elongation in 2 in.
Annealed
Heat resistance
Scaling temperature
Continuous service
:
1600°F 1600°F 1500°F
Intermittent service 1450°F 1450°F 1600°F
== 0h
Cold-heading properties Good

Corrosion resistance Very good resistance Similar to 30S Excellent resistance In both the hard- Excellent resistance
to atmosphere. Good to atmosphere and ened and unhard- to weather, water,
resistance to many weathering. Resis- ened condition it has good resistance (o
organic and inorgan- tance to organic and | very good resistance most chemicals and
ic chemicals. inorganic chemicals. to atmosphere, water, food products.
Superior to type 430 | carbonic acid, crude
but is not equal to oil, gasoline, blood,
that of type 304, 305 | perspiration, alcohol,
or 16-18. amonia, mercury,
soap, sugar solutions
and many other
reagents

SOURCE: Elco Industries, Rockford, Ill,

813
Table 16-3 Materials for Typical Applications

Material Application

Steel
SAE 1010 Useful for low strength require-
ments, such as carriage bolts and
machine screws

SAE 1018, 1020, 1021 Cap screws

SAE 1038 High strength cap screws and bolts

Aluminum

2024-T4 Machine screw nuts, cold formed

bolts, rivets

2011-T3 Screws and bolts

1100 Cold formed rivets

Brass

Yellow brass Cold-headed bolts, screws, rivets

Bronze
Commercial bronze Cold-headed bolts, screws, rivets

Naval bronze Hot-forged bolts, cold formed bolts,

nuts

Copper

High-silicon bronze Hot-forged bolts, nuts, set screws

Low-silicon bronze Cold-formed bolts, nuts, rivets,
screws

Nickel
Monel Bolts, nuts, screws, rivets
Nickel For applications involving contamination,
high and subzero temperature operation
Inconel For applications requiring resistance
to oxidation at high temperatures

Stainless steel
Martensite Used in magnetic applications, and
those requiring hardenability

Ferritic Used for economic reasons and for

those applications where extreme
corrosion resistance is not
necessary
Austenitic Best corrosion resistance

Non metallic
Nylon Used for applications requiring good
insulation and resistance to heat,
shock, and vibration, and chemical sol-
vents

Polyvinyl chloride Used for applications exposed to

outdoor weathering

Teflon Used when resistance to corrosion and

high temperature is important
SRS
FS a ED

814
 815
Section 16-10: Preload and Tightening Torque on Bolts

Table 16-4 Grade and Material Identification Markings Required by SAE and
GM Specifications

Is Mir’s Mechanical Properties Hardness

Identification Identifi- Nominal

Grade Fastener
Material Size Range, Remarks
Mark Designation Description
cation
Proof Yield Tensile
Symbol Load, Strength, | Strength, Brinell Rockwell
Required? psi min. psi min. psi

No Mark SAE grade 0 | Bolts and screws Steel Yes ;to | 5

————
ea
SAE grade | Bolts and screws | Carbon steel Yes Equivalent to
No Mark tto1s — | — | 55,000 |207 max. |B95 max. |ASTM A307,
GM 255-M Bolts and screws | Carbon steel |Optional grade A
ee rt Io

SAE grade 2 Bolts and screws | Carbon steel Yes fto ; $5,000 = 69,000 | 241 max. | B100 max.

No Mark iw over i to - 52,000 = 64,000 | 241 max. | B100 max.

GM 260-M Bolts and Screws | Carbon steel |Optional over :to 4 28,000 — 55,000 | 207 max.| B9S max.

: So
|
an atOin 85,000 === 110,000 | 207/269 | B95/104
SAE Grade 3 | Bolts and Screws Meaty, Pal Yes 4 1 7 5
carbon stee over 5 tog 80,000 == 100,000 | 207/269 | B9S/104

—! — oooh. - — |
1 3
SAE Grade 5 | Bolts and Screws Medium Yes 44 85,000 or 120,000 | 241/302 | C23/32
WA carbon steel, over 5to 1 78,000 | 81,000 | 115,000 | 235/302 | €22/32 Rene
heat treated 1
GM 280-M Bolts and screws Optional | over | to 15 74,000 | 77,000 | 105,000 | 223/285 C19/30

SAE grade S.1 | Sems , Yes

——— eee a
ea cated
3 and smalter | 85,000] — | 120,000 | 241/375 | €23/40
GM 275-M Sems Optional
SET
— ee

Medium 7to ; 110,000 =) 140,000 | 285/331 C30/36

— _— | SAE grade 6 Bolts and screws | carbon steel, Kes 5.3
heat treated over gto 105,000 —— 133,000 | 269/331 C28/36

aa a
A SAE grade 7 Bolts and screws | Low Yes Desee
rolled after
UN =i alloy steel, ito 15 105,000 |110,000 |133,000 |269/321 | €28/34 | jo
GM 290-M Bolts and screws ina: Web Optional treatment
er
ee ee

<aVA SAE grade 8 Bolts and screws | Low Yes Equivalent to

alloy steel, 4to 15 120,000 | 125,000 | 150,000 | 302/352 | C32/38 ASTM A354,
A S GM 300-M Bolts and screws | heat treated Optional grade BD
- am
Corrosion P it : 2
— GM 455-M Bolts and screws |octane stee}| OPtional 4 to l> 40,000 = 55,000 | 143 min. | B79 min.

SOURCE: Camear

SECTION 16-10

Preload and Tightening Torque on Bolts

In general, when bolts are used to connect two parts, the strength of the
joint depends, among other factors, on the preload applied. The benefits
of a preload are especially pronounced in applications involving fatigue
loading. The parts to be connected may or may not be separated by a
gasket. In this section we will consider the situation when no gasket is used.
Consider the bolt and nut shown in Figure 16-37. The parts are subjected
to an external tensile load F, and external shear load F,. F; is the initial
 &16
Chapter 16: Fasteners

Figure 16-37 The sketch shows two

plates being clamped together by a
nut and bolt. Shown are the preload, F;,
external tensile load, F,, and shear
load F,.

Parts to be connected

tensile load placed in the bolt by tightening the nut before the external
forces act. Clearly, the preload produces compression between the parts and
thus creates sufficient friction between the parts to resist the shear load.
Obviously, without this frictional resistance the bolt would be subjected
to shear and bearing loads in addition to the tensile loads.
The most important factor that determines the preload induced in a bolt
is the torque applied to tighten the bolt. There are several methods commonly
used to apply a predetermined torque. The torque may be applied manually
by means of a wrench which has a dial attachment that indicates the
magnitude of torque being applied. Pneumatic air wrenches are also widely
used. In this method the air pressure is cut off automatically when the desired
torque is obtained. Another method is to tighten the nut by hand and
then to use a wrench to give the nut a predetermined number of turns.
Unfortunately, a wide range of torques can result from the use of the
techniques described above.
Empirical equation (16-12) can be used to give reasonably accurate results
for the relationship between Vz~ and applied torque.

ak ?\
T = cdF, \\ (16-12)
~

where ( ae
| q :
T = applied torque, in b x
c = torque coefficient ANXN —
d = nominal bolt diameter, in=
F, = initial preload, lb
The value for the torque coefficient for dry surfaces and unlubricated bolts
is taken as 0.2, and for lubricated bolts a value of 0.15 is suggested.
 817
Section 16-11: Elastic Analysis and Fatigue Loading

If a more exact value is required, equation (16-13) can be used to calculate

the torque coefficient.
_ 4,,[ UW,sec d + tan i d.
(16-13)
~ 2d\1 — p, sec ¢ tan Oy
where

d,, = mean bolt diameter, in.

= nominal bolt diameter, in.
ie; ie
= mean bearing face or collar diameter, in.
Ht, = bolt thread coefficient of friction
L,. = coefficient of friction at bearing face of bolt or nut
o@ = thread half angle, deg
w = thread helix angle, deg
A comparison of this equation with the power screw equation, (15-5), will
immediately make obvious that they are essentially the same.
The determination of the exact tightening torque required in a particular
problem can probably best be determined experimentally. In other words,
a prototype can be built and accurate torque testing equipment used on it.

SECTION 16-11 \
Elastic Analysis and Fatigue Loading
o fatigue loads,
ily applied load is

defined in the same way as the spring constant is for springs. Namely, it is
the ratio of the force applied to the deflection caused by the applied force.
In equation form
1a da AE
(16-14)
ne a
where

k = stiffness constant, lb/in.

A = cross-sectional area, in.”
E = modulus of elasticity, lb/in.*
! = length, in.

Referring to the situation shown in Figure 16-37, let us assume that the
preload, or initial tensile load F;, has been applied. Both the bolt and the
parts will undergo a deformation due to this load, the bolt increasing in
length while the parts contract (see Figure 16-38). At this point the external
load F, is applied, as shown. The result is an increase in the length of the
 818
Chapter 16: Fasteners

Figure 16-38 (a) The plates to be

clamped by the nut and bolt are shown.
There is no preload, no stress in plates
or bolt. The length of the plates is L
and the length of the bolt extending .
beyond the nut is L’. (b) After the preload
is applied (by turning the nut-against
the plate), the length of the plates has
been reduced by L, and the portion of
the bolt extending beyond the nut has
been increased by L,. L, is greater than
L,. Clearly, the plates are subjected to a
compressive stress while the bolt is
subjected to a tensile stress.

bolt equal to F,,/k, and a decrease in deformation of the parts equal to

F,,/k,. These two changes in deformation must be equal, unless the parts
have separated. Clearly, if the external force is large enough to separate
the parts, ail of the load will be taken by the bolt. Assuming separation has
not occurred,
 819
Section 16-11: Elastic Analysis and Fatigue Loading

also

and similarly

or

\ F,=—-“*_+F, (16-15)

k
aN eS as WFl 16-1
—_— ke k (2 ( 6 6)

pressure vessels, gare hen a ARTE! cannot be used. Although the

maximum load on the bolt is greater because of the preload, the minimum
load is also greater. As a result, the average load is greater, but the range
component is reduced. Because the stress concentration factor is applied
to the range component, the reduced range stress results in a much better
stress situation. The average and range components of the bolt force are
obtained by substituting
F,,,,) and Fgange) in place
of F, in equation (16-15).

Kr
Fiymax) =
ee
eet) sul

lah nee k,F.

b* e(min) a F,

Oe Shp
F eelivenan Si Foconiny) KPa el F.

ae 2 ike
F Eran Fytnin) KF ec)
Xba) = 7;
 820
Chapter 16: Fasteners

Fyay and F,,, are now used to calculate o, and o,,. The fluctuating load
analysis previously discussed is then used to complete the problem.
A final wor¢ caution: if the preload is allowed to become too large,
a detrimental stress situation may result.

Example 16-6. In order to demonstrate the importance of the preload in fatigue

loaded bolts, let us consider the following problem.
The bolt shown in Figure 16-39 is a -in. coarse threaded bolt. The parts it
connects are subjected to a load F, that fluctuates from 1000 to 8000 lb. The
bolt and the parts being connected are made of the same material. Assume the
bolt material has a yield strength of 70,000 psi and an endurance limit of
30,000 psi. Assume that the factor for the threads is 3.8
( p
(a) Is the bolt safe for continuous action? (b) If the preload is 10,000 lb is it
safe for continuous action? (c) If the preload is 16,000 Ib is the bolt safe for
continuous action?

Solution: Part a.
1000 + 8000
Fgay) = 5 = 4500 Ib
0 — 1
Re a = 3500 Ib
stress area = 0.226 in.”
4500 :
ben == 0.226 = 19,000 psi

3.8 x 3500 5
Ko, = 0226 => 58,800 ps1

ure | licates that failure will occur.

eo AE 0SLE

Figure 16-39
 821
Section 16-11: Elastic Analysis and Fatigue Loading

ea

Figure 16-40

15,000

35,000 Sie = 70,000 psi

+ 10,000 = 10,502 Ib
[031k
SS aa
2475
SS

i I
10,502
= 46,500 psi
Pov) = 0.96
e(r)
_ 0.31 x 3500 Zs
aXe 5 OOPS A

A plot on the stress diagram shows the bolt is now safe. In other words, with
nothing changed in the problem except the addition ofa preload, the bolt went
from an unsafe to a safe condition.
Part c.

FE.
KF.
— —_b’ e(av)
0.31 x 4500
ieee este heh Molise
eee a amoaiesoay
= 16,502 Ib
16,502 ay
Ee
oar) = 0.96
 822
Chapter 16: Fasteners

kee 0.31 x 3500

SOS - ak, - 031 = 39
ae
_ 3.8 x 390
Ko = 6550 psi
= ~0.206
A plot on the stress diagram now shows failure will occur. This demonstrates
that if the preload is made too large, failure will also occur. ©

SECTION 16-12

Analysis When Gaskets Are Used With the

Connected Parts
There are applications for which a sealing or gasketing material must be
placed between the parts to be connected. Figure 16-41 shows a bolted
connection with a gasket placed between the parts. If the gasket is made of
a soft material, a greater proportion of the externally applied load acts on
the bolt. Clearly,
The total load on the bolt can be calculated from equation (16-17).

Pea CP eae (16-17)

where C is a stiffness coefficient and can be determined by using equation
(16-18).
k
C=—— (16-18
kp tke See
where k, is a combined elastic constant depending on the different types of
material being clamped. The combined constant can be obtained from
equation (16-19).
1 od
=
us (16-19)

Figure 16-41 An example of two parts to be connected by a

nut and bolt and separated by a gasket.
 823
Problems

where k,, k,, and so on, are the elastic constants for the different materials
For example, in the connection shown in Figure 16-41, k, is found from

where k, is the elastic constant of the gasket and k, is the elastic constant
of the connected parts.
In many applications, the stiffness of the connected se is so great
compared to the stiffness of the gasket, that the k, car 2 (

PROBLEMS
1. Two a, plates are ie be joined by a lap joint. The plates are 15 in. wide
and in. thick. The rivets are Z in. diameter, and the working stresses are as follows:
tension, 18,000 psi; bearing, 27,000 psi; and shear 13,500 psi. Design a joint for
maximum possible efficiency.
2. Design a butt Jone having maximum efficiency, for two tension members, each
12 in. wide and3 in. thick. The rivet diameters and working stresses are the same
as in Problem I
3. A longitudinal lap joint is to be used to manufacture a boiler having 4-in. thick
plates. Two rows of rivets are to be used. The rivet diameters are }in., and the
rivets in each row are spaced 14 in. apart. Determine the efficiency of the joint if the
working stresses are as follows: tension, 18,000 psi; bearing, 24,000 psi: shear,
13,500 psi. What change would you make to improve the efficiency of the joint?

4. A boiler with ;:-in. thick plates is to be spliced by means of a lap joint. The #-in.
rivets are arranged in three rows with the rivets in the two outer rows spaced 4 in.
apart and those in the middle row 2 in. apart. Determine the efficiency of the joint
if the working stresses are as follows: tension, 16,000 psi; bearing, 24,000 psi; and
shear, 12,000 psi.

5. A boiler having 2-in. thick plates is to be spliced longitudinally by a double riveted

lap joint. The rivets, which are ; in. diameter, are spaced 25 in. apart in each row.
If the boiler is to contain steam at a pressure of 80 psi and have a diameter of 6 ft,
determine the tensile, bearing, and shearing stresses.
6. A double riveted, longitudinal lap joint is to be used to splice a boiler having ;;-in.
thick plates. The 3-in. diameter rivets are spaced 2 in. apart in each row. Determine
the efficiency ofthe joint if the working stresses are as follows: tension, 18,000 psi:
bearing, 24,000; and shear, 13,500 psi. The rivet holes have been drilled.
7. A gas tank is to be made of3-in. thick plates, formed into a cylinder 8 ft in diameter.
The rivet holes are drilled to accommodate 3-in. diameter rivets. The longitudinal
lap joint is to have three rows ofrivets, arranged so that the rivets in the outer rows
are spaced 4 in. apart and those in the inner row are 2in. apart. Determine the
efficiency if the working stresses are as follows: tension, 18,000 psi; bearing,
27,000 psi; and shear, 13,500 psi. What will be the magnitude of the maximum
allowable pressure?
8. Determine the maximum rivet load for the eccentrically loaded connection shown in
Problem Figure 8.
 824
Chapter 16: Fasteners
10,000 Ib

Problem Figure 8

. Acolumn supports an eccentric load by means of two plates attached to opposite

ends of flanges as shown in Problem Figure 9. For an allowable shear stress of
16,000 psi, what size rivet is required?
10. A 1-in. thick steel plate is attached to a column by the unsymmetrical riveted joint
as shown in Problem Figure 10. The rivets are 1 in. in diameter. Determine the
total stress in the worst stressed rivet under the eccentric loading.
11. In Problem Figure 11, a gusset plate is attached to a vertical column by four rivets.
Determine the required rivet diameter if the allowable shear stress is 15,000 psi.
12. In Problem Figure 12 the fitting is to be bolted to a plate by means of three equal
size bolts spaced as shown. All parts are made of medium carbon steel, quenched
and tempered, having a tensile strength of 125,000 psi and a yield strength of
90,000 psi. A minimum factor of safety of 1.75 is required. Determine bolt diameter
required.
13. In Problem Figure 13, three bolts (3-10) are used to connect the seat shown to a
vertical column. If the working stresses for the bolt are 20,000 psi for tension and
12,000 psi for shear, will the bolts be satisfactory?
14. A torque wrench reads 240 in.-Ib when it is used to tighten a 3-11 nut on a flange.
Assuming 20% of the input torque is needed to overcome the friction under the
head of the nut, what is the tensile stress in the bolt if the threads have a coefficient
of friction of 0.1.
15. A pipe flange is connected by the turned down bolt shown in Problem Figure 15.
The 7-10 bolt is turned down to its root diameter and, after tightening, its length is
increased by 0.005 in. What is the preload in the bolt?
 S2ZD
Problems

30,000 Ib

Bearing plate
between the
two gusset
plates

Two gusset plates

Problem Figure 9

16. For the bolted connection shown in Problem Figure 16, the bolt is 3-10. For a
preload of 1500 Ib and an external load of 3000 lb, what is the total force on the
bolt? The bolt and part are made of the same material, and the part has an area of
Les int2
17. For the bolt of Problem 16, assume the external load fluctuates between 4000 and
10,000 lb. The bolt and the parts are made of the same material, and the part has
an area of 1.3 in.? Use a factor of safety of 2,5, = 60,000 psi, and S,, = 100,000 psi.
(a) Is the bolt safe for continuous operation? (b) Ifa preload of 8000 Ib is applied, is
the bolt safe? (c) Is the bolt safe with a preload of 12,000 lb? (d) What is the value
for the minimum force in the part for the preload of part (b)?
 16 Kips

Problem Figure 10

Problem Figure 11

826
 F = 20,000 Ib

Problem Figure 12

Paine

2in.
Problem Figure 13

2in.

2in.

de
4 in,

a =
Sth

L 23 in. seal

827
 828
Chapter 16: Fasteners

Problem Figure 16

18. For a connection similar to that in Problem 16, the external force varies from
2000 to 3000 Ib. The bolt is -13 UNC, bolt and part are made of the same material,
S,, = 100,000 psi, S, = 60,000 psi. Assuming an initial load of 3600 Ib is applied
to the bolt and parts, determine the factor of safety for the bolt. The cross-sectional
area of the parts may be taken as 0.6 in.”
19. In Problem 18 determine the maximum permissible value of the preload, if the
factor of safety is to be 1.5. :
20. In Problem 17, determine the maximum and minimum values for the external load,
if the preload is 5000 lb and the range stress, ¢, = 5000 psi.
21. The bolt shown in the Problem Figure 16 is used to connect two plates of the same
thickness but made of different materials. The bolt is 5-13, one plate is made of a
material having a modulus of elasticity of 15,000,000 psi, and the material of the
other plate has a modulus of 30,000,000 psi. A preload of 6000 Ib is applied to the
bolt. If the external force varies from 0 to 3000 1b, and S,, = 100,000 psi, S, =
60,000 psi for the bolt, what will be the factor of safety?
22. A boiler drum has a diameter of 90 in. and has the longitudinal butt connection
shown in Problem Figure 22. The rivet diameters are { in., and the holes are drilled.
The design stresses are as follows: tension, 20,000 psi; shear, 12,000 psi; bearing,
24,000 psi. Determine the maximum internal pressure and the joint efficiency.
23. The connecting rod and bolt shown in Problem Figure 23 each have a yield point
stress of 90,000 psi and an endurance limit of 58,000 psi. The bolt is 3-24 UNF and
is drawn up with an initial tension of 3500 lb. Stress concentration factor for the
threads is 3.8 and the bolt load varies continuously from 0 to 2000 Ib. What is the
factor of safety for the bolt?

: NTT
MYON) a By,AO AQ PQQ A NW
ay, YY ay, Se Be,
CO|wW
clm
Mlw

Probiem Figure 22
 829
References

Problem Figure 23

24. Calculate the size of bolts (all bolts of the same size) required to safely support a
bracket to which is applied an eccentric load of one ton as shown in Problem
Figure 24. Siatiow) = 20,000 psi; Ssanow) = 12,000 psi.

Load

Problem Figure 24

REFERENCES
[1] Manual of Steel Construction Handbook. American Institute of Steel Construction,
Inc., New York, 1970. ;
[2] E. C. Harris: Elements of Structural Engineering. The Ronald Press, New York,
1954.
[3] L. E. Grinter: Elementary Structural Analysis and Design. Macmillan Publishing
Co., New York, 1965.
[4] American Standard Unified Screw Threads (ASA B1.1—1960) ASME, New York.
[5] J. E. Lothers: Design in Structural Steel. Prentice-Hall, Englewood Cliffs, N.J.,
1972:
Welds and Adhesive J oints
SYMBOLS

e = load eccentricity, in. S, = shear strength of adhesive, psi

h = nominal size of weld, in. S.yp = Shear yield strength, psi
I = moment of inertia of weld S,, =
yield strength, psi
group S.» =
working strength in shear, psi
I’ = moment of inertia based on 1-in. t,, =
throat dimension, in.
throat x =
horizontal distance to center of
J = polar moment of inertia of weld gravity, in.
group y = vertical distance to center of
J’ = polar moment of inerta based gravity, in.
on 1-in. throat t = shear stress, psi
N = factor of safety

Most weldments are produced by fusion processes. These involve establish-

ment of a metallurgical bond between two parts by melting together the
base metals or by melting the base metals with a filler metal. Important
factors in successful welding are the elimination of surface barriers to joining
and minimizing of damage (distortion and unfavorable heat treatment) to
areas adjacent to the weld.
The necessary welding temperature may be produced by a gas flame or
by an electric arc. Coated metal electrodes are used most commonly for
manual arc welding. Gas shielded metal arc welding is also common.
Electroslag welding, electron beam welding, and laser and plasma-arc
methods are used on a small scale for very special applications.
Thin metal parts may be joined by resistance welding, particularly spot
and seam welding. A spot weld is produced by a pair of electrodes that
apply pressure to either side of a lap joint and form a complete circuit
(Figure 17-1a). Resistance is a maximum between the parts, and heat genera-
tion at that point causes local fusion. Figure 17-1b shows a portable spot
welder.

830
 831
Chapter 17: Welds and Adhesive Joints

Force

Electrode

Weld
VLLLLLLLLLL
LL LLL gh LL
RK
AASAASS AAS SASASSSS SS

Workpieces

Force

(a) (b)
Figure 17-1 (a) Spot welding. (b) Portable spot welder. Electrode pressure is
provided by air cylinders. [Courtesy Falstrom Company, Passaic, N.J.]

High strength adhesives are also utilized to join machine elements and
in structural applications, particularly when the area to be bonded is
relatively large. Plastics may bejoined by various welding processes including
induction welding, heated tool welding, friction welding, and hot air welding.
Figure 17-2 shows a hot air type plastics welder.

Position of

Welder
Plastic
fs
welding rod

Area to be
preheated

Base material

Figure 17-2 Plastics welder. (a) A hot-air welder shown without welding tip.
Units of this type heat compressed air or inert gas to 400-900°F to weld
thermoplastics. (b) Joining plastic sheets using a hot-air welder with a round tip.
[Courtesy Kamweld Products Company, Inc., Norwood, Mass.]
 832
Chapter 17: Welds and Adhesive Joints

SECTION 17-1
Arc Welding
Among the many manufacturing processes described in Chapter 4, arc
welding is one of the most important. Common methods of arc welding
include manual arc welding using coated metal electrodes, metal—inert
gas arc welding and tungsten-inert gas welding.
Coated metal electrodes may be used in the field for manual arc welding
(Figure 17-3a). The coating is vaporized to provide a shielding gas which
prevents oxidation at the weld. The coating also acts as a flux and directs
the arc. The electrode itself becomes the filler metal in the joint. If postweld
heat treatment is used, the electrode composition should ordinarily be
similar to that of the base metal. Because the electrodes are consumed rapidly
in welding and must be replaced, a process using coated metal electrodes is
not easy to automate.

Electrode

Coating

Shielding gas

Workpieces Figure 17-3 (a) Arc welding. (b) Metal-inert

(a) gas (MIG) arc welding.
Electrode wire

Torch

Shielding gas
 833
Section 17-2: Design of Symmetrically Loaded Weldments

The use of inert shielding gases reduces the operator skill required and
makes automation possible. In metal-inert gas arc welding (MIG) a bare or
plated wire is continuously fed into the weld from a large spool (Figure
17-3b). The wire acts as electrode and becomes the filler in the joint. (It would
be impractical to feed a coated electrode in this way.) An inert gas is also fed
into the torch. Cooling water may also be supplied through the power cable
jacket and into the torch so that a lighter cable can be used. High production
rates and uniform quality welds are possible with MIG welding.
Tungsten-inert gas welding (TIG) utilizes a “‘nonconsumable”’ tungsten
electrode and usually argon or helium as the shielding gas. This process is
particularly useful for welding aluminum. A separate filler rod is melted into
the joint when required, but the electrode is not used as a filler.
The fundamental types of welded joints and the standard welding symbols
are given in Table D-1 of the appendix.

SECTION 17-2
Design of Symmetrically Loaded Weldments
Coated metal welding electrodes are designated according to tensile
strength of the completed weld and according to the intended welding
method. The last two digits represent the type of weld application and the
position of the weld. The two or three digits immediately following the prefix
E give approximate minimum tensile strength in thousands of pounds per
square inch. Most machinery welding is done with electrodes in the E60XX
group. These give minimum ultimate tensile strengths S, = 60,000 psi or
higher in the as-welded condition and minimum yield strengths S,,, = 50,000
psi. Electrodes in the E70XX group are also commonly available. They
provide 70,000 psi or higher minimum ultimate strengths and 60,000 psi
minimum yield strengths. High strength welds are made by electrodes in the
E100XX group, which form welds that self-temper on cooling, providing
ultimate strengths in excess of 100,000 psi.

Butt welds in tension

When designing welds, we should specify an electrode with properties
comparable to the base metal when possible. In the case of butt welds
(Figure 17-4) the joint strength should then approximate the strength of the
base metal, giving us an efficiency near 100 % for static loads. Thus for full
strength butt welds, the joint strength is

P = S,)Lt
(17-1)
IGN
for plate thickness t, weld length L, and safety factor N for static tensile loads
in the plane of the plates perpendicular to the weld. There are, of course,
certain side effects, principally due to the intense welding heat. These problems
 834
Chapter 17: Welds and Adhesive Joints

eee
AL
Tv
t

Figure 17-4 Butt weld in tension.

include residual stresses, distortion of the part, and cracking. Stress corrosion
and a lowered endurance limit may also appear after the part has been in use.
In addition, slag, voids, and other imperfections may occur in the weld.
Forfluctuating loads, a butt weld made with an electrode having properties
similar to the base metal will have a fatigue strength about 15% below the
fatigue strength of the base metal. Weld metal that extends above or below
the surface of the base metal of a butt joint is called “‘reinforcement.” This
is thought by some to increase the joint strength and compensate for weld
imperfections. Reinforcement is probably of little consequence for welds
subject to static loads but may be detrimental as a source of stress concentra-
tion (as noted in Chapter 3) if loads fluctuate. Thus reinforcement should be
kept to a minimum or ground down for weldments subject to fatigue loads.

Example 17-1. An E60XX group electrode is used to butt weld two ;-in. steel plates.
The yield point of the base metal is 50,000 psi. Find the strength of the joint for
a safety factor of N = 2. The plates are subject to a static tensile load perpen-
dicular to the joint.

Solution: A butt weld in j-in. plate can be made with or without edge
preparation. In either case, full strength should be developed and the weld will
not materially reduce the strength of the plates. Thus, we have
ike Spt _ 50,000(z)
N 2
= 6250 lb/in. of weld. @

Welds in shear
When a butt weld joining two plates is loaded in the plane of the plates
parallel to the weld, shear failure is likely to occur. For total weld length L,
 835
Section 17-2: Design of Symmetrically Loaded Weldments

plate thickness t and safety factor N, the allowable load will be

Set Di:
P=? _ (17-2)
N
for full strength butt welds in shear on a static basis. If the shear yield point is
not known, we may generally assume

Thus, for E60XX group electrodes with a yield point of 50,000 psi S,,,, 1s
approximately 33,000 psi. In every case, S,,, > 5S,,.
Figure 17-Sa shows a fillet weld in shear. If the weld cross section has the
form ofan equilateral triangle, Figure 17-5b, the probable line ofshear failure
will be the weld throat as shown. For nominal size (leg dimension) h, the
throat dimension is given by t,, = 0.707h. In most welds, the minimum failure
dimension will exceed 0.707 times the nominal size, but we will still base our
calculations on the throat dimension t,, = 0.707h (see Figure 17-5c).
For the total weld length L counting all welds, the joint strength for fillet
welds loaded symmetrically in shear as in Figure 17-5 is given by

pe 0.707hLS
=vP (17-3)
N

(a) Shear Loading

NG (c) Throat
(b) Shear failure at Dimension
throat of weld. 4
Ay
S

Figure 17-5 Fillet weld loaded in shear.

 836
Chapter 17: Welds and Adhesive Joints

Actual stress distribution in a weld of this type is quite complicated and

depends on the stiffness of the base metal, slight bending of the plates, and
other factors that we will neglect. The result is valid for design, however,
because weld strengths are based on tests on joints of this type.
If fillet welds are also placed perpendicular to the direction of loading (say,
along edges BC and AD ofthe plates in Figure 17-5a), both tension and shear
stresses will occur on the throat section of those welds. Equation (17-3) will
still give reasonable results for design purposes, however, if L is taken as the
total length of all welds.

SECTION 17-3
Eccentrically Loaded Joints
When practical, welds should be symmetric about the line of action of the
resultant load. If the load vector does not go through the center of the weld
group, we must consider twisting and/or bending as well as direct shear. '

Loading in the plane of the weld group

Figure 17-6 illustrates an eccentrically loaded joint. The exact stress
distribution is complicated and depends on the rigidity of the parts being
joined as well as the weld geometry, but the following method will give
reasonable engineering results.
Assume the parts are relatively rigid. Then, stress due to load eccentricity
is proportional to distance from the center of the weld group. As with
symmetrically loaded fillet welds, we approximate the stress area in shear by
the throat area. Shear stress in the weld at radius r is given by the vector sum

je Aire
iG => —+4>> —
(17-4)
A J

where

P = applied load
A = total throat area
T = torque, the product of applied load and its distance from the center
of gravity of the weld group
r = distance from the center of gravity of the weld group
J = polar moment ofinertia of the weld group about its center of gravity
(based on throat area)

The design of weldments calls for an even simpler procedure. Because weld
throat area will not, in general, be known at the start of the design process, we
temporarily specify a unit throat dimension. Equation (17-4) then becomes

1 Ane
(17-5)
/
16 =
Leys
 837
Section 17-3: Eccentrically Loaded Joints

Sin p late
8

Column

Figure 17-6 Eccentric loading. G is the center of gravity of the weld

group.

where L is the total length of the fillet weld and J’ is the polar moment of
inertia based on a 1-in. throat. Furthermore, when locating the center of
gravity and determining r and J’ the weld will be treated simply as one or
more lines. We will measure only to the near edge of the weld, ignoring the
weld thickness and ignoring the contribution of moment of inertia about the
weld’s own central axis. Once t’ is thus computed, we compare it with
allowable shear stress to specify true weld size. For working strength
Ss = Ssyp/N, the nominal weld size (leg dimension) is given by

oesOS, 17-6
76)
Ordinarily the nominal size h that we specify would be the nearest “round
number” size (4, 7, ¢in., and so on) equal to or greater than the computed
value. Using this method, all welds on the joint will have the same nominal
size.
If the weld group is symmetric, the location of the center of gravity will be
obvious. Otherwise we will select a convenient origin and compute horizontal
and vertical distances x and jy to the center of gravity. We treat the weld as a
series of lines of length L,, L,, and so on at the plate edges with their own
centers at horizontal distances x, , x,, and so on, from the origin. Horizontal
 838
Chapter 17: Welds and Adhesive Joints

distance to the center of gravity is given by

tpula
Earls (17-7)

Similarly, vertical distance is

eey= a (17-8)
17-8

where the subscript i = 1, 2,..., for n welds. Once the center of gravity. is
found, the polar moment of inertia may be calculated, treating each weld as
a line. The contribution of a single weld parallel to the x axis is

Di) am Li?

/
Iel —2
Lyi — 49) a= LX;

where x; and y; are the distances from the center ofgravity to the center ofthat
particular weld. We add the contributions ofall welds to obtain J’. and J, and
then obtain J’ = I, + J}.
In most cases, it is easiest to compute horizontal and vertical components
of shear stress and then combine them. We replace equation (17-5) by
equations (17-9), (17-10), and (17-11).

Pee eelia ;
Tyyoett
jie de (17-9 )

Py etePr,
7 =—t 17-10

Iai e GalGH (L711)

where P, and P,, are the vertical and horizontal components of the load and
ry and ry are the vertical and horizontal distances from the center of gravity
of the weld group to the point where we wish to determine stress. The plus
sign in equations (17-9) and 17-10) applies if the direct load component (the
first term) results in shear in the same sense as the torsion component. For a
point nearest to the line of action of the load, the plus sign applies. Shear
stress is ordinarily greatest at a point in the weld group nearest to the line of
action of the load. If the weld group is not symmetric, we may have to check
two or more of the points in the weld that are farthest from the center of
gravity.

Example 17-2. A bracket is to be joined to a column by fillet welds. The dimensions

of this weldment are shown in Figure 17-6. Find the required weld size (leg
dimension).
Given. P = 2000 Ib. Allowable shear stress in welds S,,, = 7000 psi.
 839
Section 17-3: Eccentrically Loaded Joints

Solution: The center of gravity is given by

ely
4x 20
xe SL, = alee = 0.8 in.

9 San Ore ey
MeL LIN. © AaOm ee Ae
from which the load eccentricity is
Cs
= ke SS Oi,
Based on a 1-in. throat, the moments of inertia are
3
fh = (18)? + 2D + 6G = 18) 39.65

3
I, = 6(0.8)* + - + 4(2 — 0.8)? = 14.93

and
J’ = 54,58

By inspection, either point B or point C is subject to the greatest stress. At

those points, shear stresses based on a 1-in. throat are:
py Ler, 2000) 20006)62) 5
Ger eo oe + RSRe 904 (vertical component)

AGB ee Per, — 2000(6)(1.8)

= 396 (horizontal component)
eT Sh 5458
t’ = [(t,)* + (ty)7]'/? = 987 (total shear stress based on 1-in. throat)
, P Pery — 2000 — 2000(6)(0.8)
Spies mye ES SSG
an (CX Pery — 2000(6)(4.2) _
a eae eS
ie == GP)

Point B governs.
pee ee 8
= 0.199 in. (leg dimension)
~ 0.707S,,_ 0.707(7000)
For convenience, we might specify 4 in. (leg dimension) fillet welds through-
out. @

Loads that do not lie in the plane of the weld group

A somewhat different treatment is indicated when the loading lies outside
the plane of the weld group. Consider the angle section of Figure 17-7 which
is welded to a column. Load P is applied at a distance e from the plane of the
welds. The actual stress distribution in the weld group is very complicated.
However, for design purposes, we may assume a linear distribution of shear
stress due to moment M = Pe and a uniform distribution of direct shear
stress.
 840
Chapter 17: Welds and Adhesive Joints

CCa a S a AN a

Figure 17-7 Loading that does not lie in the plane of

the welding group.

We begin by locating the center of gravity of the weld group. As in the

previous example, the weld may be treated as a line in these calculations.
Consider the case in which the plane of the weld group is perpendicular to the
plane determined by the line of action of the force and the weld group center
of gravity. In this case there is no tendency toward twisting in the weld group
plane. The weld is subject to direct shear stress P/A and a moment-induced
stress Mc/I,. Moment of inertia J, is computed with respect to an axis
perpendicular to the load line of action and through the center of gravity.
Distance c is measured from that axis to the farthest point on the weld. In
spite of the resemblance to bending of the moment-induced stress, we will
design on the basis of shear in the weld. Shear stress in the weld is approxi-
mated by the vector sum

ues
T= —H—
Ale 2h.
(17-12)
-

If the above equation is to be used for design purposes, we may treat the
weld as a line, but base the area and moment of inertia on a 1-in. throat
dimension. Analogous to equation (17-5), we obtain

i
P Me
T => — > —
(17-13)
IES Ae Sik
Based on the geometry and loading of Figure 17-7, we see that the direct
shear effect is downward and that the moment-induced stress along edge AB
is outward. The vector sum is simply
P 2 Me Da |3ly

v= |(F] SE | se
 84]
Section 17-4: Adhesive Joints

For a shear working stress S,,,, the required minimum leg dimension is given
by
,
tb
Fj e ees
0.7075,
Example 17-3. Referring to Figure 17-7, let us find the required weld size if dimension
AB = 3in., dimension BC = 4in., eccentricity e = 2.5in., and load P =
2000 Ib. Allow a working stress in shear of S,,, = 5000 psi.
Solution: Moment M = 2000(2.5) = 5000. The center of gravity is located
by inspection and the moment of inertia I’, is calculated with respect to an axis
2 in. below AB. Based on a 1-in. throat dimension,
I’, = 2(3)(2)? + 2(4)°/12 = 34.67
L = 2(3 + 4) = 14
2000\? — {5000(2)\? ]1/?
ae (Fer zs|34.67 Send
from which the required leg dimension is
= 3219
= 0.091 in.
~~ 0.707(5000)
(For convenience, we would specify 4-in. leg dimension throughout.) @

SECTION 17-4
Adhesive Joints
It is beyond the scope of this paragraph even to list the adhesives currently
available. A large number of engineering adhesives fall into the categories of
thermosetting and thermoplastic. The latter, of course, are limited by their
tendency to soften at elevated temperatures and may even tend to creep at
room temperature when subject to heavy loading. There are many other
adhesive cure mechanisms. Anaerobic curing materials, for example, remain
liquid in the presence of oxygen but harden in confined spaces such as on
bolt threads or in the space between a shaft and a hub.

Types of joints
Joint geometry is most important when we join relatively high strength
materials. Large bond areas are suggested, particularly in the joining of
metals. The lap joint (Figure 17-8a) is inexpensive because no preparation is
required except possibly surface cleaning. For an in-plane load P, we may
specify the bond area so that the total adhesive strength in shear is equal to
the plate strength in tension. Shear strengths of 2000-2800 psi were recorded
in laboratory tests of one commercial cyanoacrylate type adhesive bonding
steel to steel and aluminum to aluminum. Higher values were reported for
some epoxies. Recommended working strengths are much lower.
 842
Chapter 17: Welds and Adhesive Joints

(a)

rel (b)

Figure 17.8 Adhesive joints. (a) Lap joint. (b) Butt joint (generally
unsatisfactory). (c) Single reinforced butt joint. (d) Scarf joint.

The butt joint, Figure 17-8b, is generally avoided since adhesive bond area
is inadequate. In cases where the lap joint is not desirable, however, a rein-
forced butt joint, Figure 17-8c, may be used. A double reinforced joint
 843
Section 17-4: Adhesive Joints

utilizes a similar reinforcement plate on both sides of the joint. Another

alternative, the scarf joint, Figure 17-8d, provides the desired bond area. The
cost of machining for a scarf joint, however, is seldom justified, and with thin
materials, the machining of a scarf joint is highly impractical.

Stress distribution
Shear stress in the adhesive in a lap joint is not evenly distributed. The
actual stresses depend on the thickness and elasticity of the joined members
and the adhesive. We will investigate a metal-to-metal lap joint, Figure 17-9a,
in an attempt to approximate maximum adhesive stress. It will be assumed
that the adhesive remains elastic under load.
Let us assume that the adhesive is subject to shear strain y and the metal
subject to tensile strain ¢, and ¢, (for plates 1 and 2). Based on the geometry
of an element of the adhesive layer of thickness t,, Figure 17-9b, we have
d
(1+ 6,)dx +t, y + as| = t,y + (1 + &€,)
dx
dx

for small y, from which

dy (8, — &)
S =_() 17-15
dx t. (
Differentiating with respect to x, we obtain
diymew iad
fe ee) ae

Plate 2

ita t

ke + egloxe|
Y

Ske .
ka tb
eyhdx—o y THaxe

(b)
Figure 17-9 Analysis of an adhesive joint. (a) Metal-to-metal lap joint. (b) An
element of adhesive with thickness t, and unstrained length dx.
 844
Chapter 17: Welds and Adhesive Joints

If we equate the change in tensile force in the plate for a distance dx to the
shear force applied by the adhesive over the same distance, we obtain
d
E, bt, 1 dx = Gyb dx
dx

or

Sipe Dn (17-17)
ax EG,
Similarly,

aay api (17-18)

de iat
for adhesive shear modulus G and tensile modulus E, and E, for the plates.
Using equations (17-17) and (17-18) in equation (17-16), we obtain
d’y ;
ee Cee) (17-19)
ax “i
where
eae Mey
ta\E,t, Ent,
The solution to equation (17-19) is given by
y = Acosh Cx + Bsinh Cx (17-20)
where arbitrary constants A and B are found by substituting the appropriate
boundary values of ¢, and €, in equation (17-15).

An adhesive lap joint with both plates of the same

metal and thickness
Consider the joint of Figure 17-9 with plate thickness t and total force P
at each end. Let the coordinate x be measured from the midpoint of the joint.
By symmetry, constant B = 0 in equation (17-20). At x = L/2 we have
€, = P/(btE) and €, = 0. Using these values and equation (17-20) in equation
(17-15) we obtain
GE P
AGSinh = ————=—
ee) |2 | btt,E :
from which
Te P
btt,CE sinh(CL/2)

and shear stress in the adhesive

a GP cosh Cx £ CP cosh Cx
t =Gy (17-21)
~ bttzCE sinh(CL/2) 2b sinh(CL/2)
 845
Section 17-4: Adhesive Joints

where
DG 1/2
Gee (il (17-22)

It is important to note that G is the shear modulus ofthe adhesive and E is the
elastic modulus of the plates.
It might be necessary for the designer to obtain thickness t, and adhesive
properties experimentally if these values are not provided by the manufac-
turer. Maximum adhesive shear stress occurs at x = +L/2. It is
Cie
= (ye
“man ~ 9b tanh (CL/2) eS)
Equation (17-23) may be written in the form

[Rade
T(max) = HL. (17-24)

where K, is the stress-distribution factor. This number is analogous to a

stress concentration factor, except that stress gradients are not as sharp as in
typical cases of stress concentration.
Figure 17-10 shows relative shear stress distribution for the case where
C = 4/L. \n that case, maximum shear stress is
P

Tmax) = 2.08
bL
Since average shear stress is P/(bL), we have a stress-distribution factor of
2.08.
Let us define a working strength for an adhesive as S,,, = S,/N where S,
is the shear strength and N the factor of safety. Rewriting equation (17-23),
we see that joint strength is given by
mes 2bS,,, tanh CL/2
(17-25)
> G

Average
stress

= =i 0 L L
2 4 4 2
Figure 17-10 An example of shear stress distribution in the
adhesive joint shown in Figure 17-9 (where C = 4/L).
 846
Chapter 17: Welds and Adhesive Joints

LOOM tm

of
Percent
strength
attainable
maximum

0 1 D 3 4 5 2
Figure 17-11 Theoretical effect of overlap Lin a lap joint with both
plates of the same metal and thickness.

If only the overlap L is varied, strength is proportional to the hyperbolic

tangent, Figure 17-11. For small values of CL/2 (up to 0.3 or so) tanh CL/2
is approximately equal to CL/2 and joint strength approaches
P = DLS,, (17-26)
For large values of CL/2 (say, 2 or greater), tanh CL/2 approaches unity.
Thus, based on the stated assumptions, an increase in overlap is beneficial up
toa point. An overlap L = 1.1/C theoretically yields about 90% of maximum
attainable strength. Further increases in overlap are theoretically of little
value unless adhesive yielding tends to redistribute the stress.

Example 17-4. Find the required overlap L if we are required to join two aluminum
plates in a lap joint as in Figure 17-9.
Given: Plate thickness t = 0.020in; yield strength S,, = 15,000 psi for the
plates; shear strength of the adhesive S, = 1500 psi.
Solution: Because the adhesive shear modulus and thickness are not given,
we will estimate a stress distribution factor K, = 2. Setting tia.) = S, in
equation (17-24), we have joint strength P = bLS,/K, at failure. The strength
of the aluminum plate in tension is P = S,,bt at failure. Equating joint strength
and plate strength, we obtain a required overlap of
L=K,S,,/S,=04in. @
 847
Problems

PROBLEMS
1. Two ;%-in. thick steel plates are butt welded and loaded as in Figure 17-4. An
E60XX group electrode is used and the yield strength of the base metal is 50,000 psi.
Find the strength of the joint based on a safety factor of 3.
. Two #-in. thick steel plates are butt welded and loaded as in Figure 17-4. An E60XX
group electrode is used and the yield strength of the base metal is 50,000 psi. Find
joint strength using a safety factor of 2.5.
. A fillet weld is loaded in shear as in Figure 17-5. Find the load which may be safely
carried by j-in. nominal size welds with a total length of 8 in. if E60XX group
electrodes are used and the safety factor is 3.
. (a) Find the safe load if 2-in. welds are used in Problem 3. (b) What is the percent
increase in strength of the 3-in. weld over the 4 in. weld? (c) Find the percent increase
in weight of the weld.
. Az-in. nominal size fillet weld is loaded in shear as in Figure 17-5. Find the required
weld length for a load of 11,000 Ib. Allow 15,000 psi in shear.
- A 4000-lb load is applied as in Figure 17-5. Dimensions AB and CD total 7.5 in.
Find the required weld size. Use a shear yield point of 33,000 psi and a safety factor
OZ:
. Locate the center of gravity and calculate the polar moment of inertia of the weld
group in Problem Figure 7. Find the nominal weld size required based on an
allowable shear stress of 10,000 psi.

Equal fillet welds

Near side
Far side

Problem Figure 7

8. Design a welded joint similar to the one in Problem Figure 7 if the 3000-Ib load
is applied at a distance of 8 in. from the column. Allow 10,000 psi in shear. Let
dimension AB = 4in., dimension BC = 4 in. and eccentricity e = 6 in.
. Find the required weld size for the bracket in Figure 17-7 if a 5000-lb load is
applied. Assume E60XX type welding rod is used and apply a safety factor of 2.5.
Let AB = BC = 4 in. and e = 6in.
10. Find the required weld size for the bracket in Figure 17-7 if a 2000-Ib load is
applied with eccentricity e = 9.8in. Allow 8000 psi in shear and let dimension
AB = 3 in. and dimension BC = 4 in.
 848
Chapter 17: Welds and Adhesive Joints

11. An angle bracket is loaded as in Figure 17-7. Using AB = 3 in. and BC = 4 in. find
the required weld size for a 3000-lb load applied with 3.5-in. eccentricity. Allow
7000 psi in shear.
12. Find the load that would cause weld failure in the angle bracket of Figure 17-7
where AB = 3in., BC = 4in., and eccentricity is 4.5in. Use a 4-in. weld with a
shear yield point of 33,000 psi.
13. In Figure 17-7, let AB = 3in., BC = 4in., and e = 2.5in. Load P varies con-
tinuously from 2000 lb upward to 2000 lb downward. Find the required weld size
based on an allowable stress of 4000 psi in reversed shear.

14. The two brackets in Problem Figure 14 support a static load of 3000 lb/bracket.
Find the required weld size based on an allowable shear stress of 15,800 psi.

(Not to scale)

oe Fillet welds Near side

~
SP Far side

P \b/bracket Column

Problem Figure 14

15. Each weld in Problem Figure 15 supports a load of 2000 Ib. Find the required
weld size based on an allowable shear stress of 13,600 psi.

Fillet
weld
4in. both
sides

Load P on each weld

(2P \|b/bracket)
Problem Figure 15
 849
References

16. Load P in Problem Figure 15 varies continuously from 500 to 2500 lb/weld. Let
S.yp/N
Pi
= 13,600 psi and S,,/N = 6000 psi. Find the required weld size.
17. Two 0.010-in. thick steel plates are to be joined as in Figure 17-9. Find the required
overlap if the plates have a yield strength of 40,000 psi and the shear strength ofthe
adhesive is 1200 psi. Assume the stress distribution factor for the adhesive is 2.0.
18.. Find the joint strength of an adhesive joint with 5-in. overlap. Let the adhesive
shear strength be 1550 psi and use a safety factor of 1.5. Assume a stress distribution
factor of 2.0.
19. Find the strength of an adhesive joint with 2-in. overlap where adhesive strength is
1200 psi in shear. Use a safety factor of 2.5 and a stress distribution factor of 3.0.
20. Find the required overlap if two j¢-in. thick aluminum plates are joined as in
Figure 17-9. Use a yield strength of 18,000 psi for the aluminum and a shear
strength of 1600 psi for the adhesive where K, = 2.

REFERENCES
[1] J. M. Alexander and R. C. Brewer: Manufacturing Properties of Materials. Van
Nostrand-Reinhold Books, New York, 1963.
[2] E. P. DeGarmo: Materials and Processes in Manufacturing, 2nd ed. The Macmillan
Company, New York, 1962.
[3] R. A. Lindberg: Processes and Materials of Manufacture. Allyn and Bacon, Inc.,
Rockleigh, N.J., 1964.
[4] O.W. Blodgett: Design of Weldments. Lincoln Arc Welding Foundation, Cleveland,
Ohio, 1963.
[5] E. J. Bruno (ed.): Adhesives in Modern Manufacturing. Society of Manufacturing
Engineers, Dearborn, Mich., 1970.
[6] C. V. Cagle: Adhesive Bonding. McGraw-Hill Book Company, New York, 1968.
A. R. Pfluger and R. E. Lewis (eds.): Symposium on Weld Imperfections. Addison-
Wesley Publishing Company, Reading Mass., 1966.
J. A. Newman and F. J. Bockhoff: Welding of Plastics. Reinhold Publishing
Company, New York, 1959.
| A. L. Phillips (ed.): Welding Handbook (5 Sections). American Welding Society,
New York, 1968.
[10] Welding Technology, NASA SP-5918(01). National Aeronautical and Space
Administration, 1970.
[11] N. M. Bikales (ed.): Adhesion and Bonding. John Wiley & Sons, New York, 1971.
[12] F. Koenigsberger and J. R. Adair: Welding Technology. Hart Publishing Company,
New York, 1968.
 Axisymmetric Problems
in the Design of Machines
SYMBOLS

a = inner radius, in. U = radial interference, in.

b = outer radius, in. é, = radial strain, in./in.
E, = kinetic energy, in.-lb p = mass density, lb-sec?/in.*
J = mass moment of inertia, 6, = tangential stress, psi
Ib-sec?-in. o, = radial stress, psi
p = internal pressure, psi Tmax = Maximum shear stress, psi
q = external pressure, psi v = Poisson’s ratio
u = radial displacement, in. j@ = angular velocity, rad/sec

In the class of machine elements with axisymmetric geometry and axisym-

metric loading, the basic problem may be defined in terms of a single
coordinate—the radial coordinate. A number of important applications fall
into this category, including thick wall cylinders, interference fits, rotating
disks, and curved beams subjected to pure bending.

SECTION 18-1
The Thick Wall Cylinder
Radial stresses are necessarily very small in a thin wall pressure vessel.
However when wall thickness is of the same order of magnitude as the
vessel radii, we must consider both radial stress o, and tangential stress a.
When uniform internal and external pressure act on a circular cylindrical
vessel we have axisymmetry. Then, there is no variation in stress with the
coordinate 0. In order to find stress and displacement everywhere, we proceed
as follows.

850
 851
Section 18-1: The Thick Wall Cylinder

Consider an element of a thick wall cylinder, Figure 18-1 having unit

thickness in the axial direction. Equilibrium of forces in the radial direction
yields

d(ro,) dé
d0 dr — 259 dr Sin =10
dr

Recalling limit sin x = x, applying it to the infinitesimal angle, and simpli-

x= 0
fying the result, we obtain
d(ro,)
O69 = (18-1)
dr
For outward radial displacement u, the strain-displacement relationships in
polar coordinates are
du
ee dr ands) ee r (18-2)
and the stress-strain laws for a perfectly elastic material are
1
ees rads — VG»)
(s=3)
1
€9 — p% aes vo,)

where axial stress is neglected.

d
or dé + 7 (0,1).
d0dr

99 dr

o,r d@
(uniform)

q
(uniform)

Figure 18-1 Thick wall cylinder.

 852
Chapter 18: Axisymmetric Problems in the Design of Machines

With five equations and five unknowns, a solution will be attempted.

Because the boundary conditions (the internal and external pressure)
correspond to radial stresses, equations (18-1) through (18-3) will be combined
to form a single equation in o,. We begin by combining equations (18-2) and
(18-3) to obtain

and (18-4)

These two equations are combined to eliminate u by differentiating the

second. The result is an equation in stress only

5 + (1 + v)(o, — o,) =0

Finally, eliminating tangential stress through equation (18-1), the result is

do, if aS 0
(18-5)
dr> or dr
which is satisfied by

0, == GC; 3 =- 18-6
(18-6)
Arbitrary constants C, and C, are found by setting —o,(a) = p (the internal
pressure) and —o,(b) = q (the external pressure). Tangential stress o, is given
by equation (18-1) and displacement u by the second ofequations (18-4). The
resulting equations for stress and displacement are

=?
b?
aa Rees 1 PK Oe
a a,

oe Se pe a (18-7)

b2 2
a’p|— + 1— bg 1+]
Pp
Og = are) (18-8)

iedae
r(1 — v)(a*p — b*q) + (1 + v\(p — q) :
 853
Section 18-1: The Thick Wall Cylinder

Hollow cylinders subject to internal pressure

For internal pressure p and zero external pressure, equations (18-7) and
(18-8) reduce to

o, = — (18-10)
-1

and
b2

07 + ]

Ve (18-11)
By inspection, the normal stress of greatest magnitude is the tangential stress
at the inner surface.

b* + a
Ct) =P (18-12)
If the cylinder wall thickness is t, where t « a, the second of equations (18-11)
may be approximated by the familiar og = pa/t.
As a result of the symmetry, shear stress t,, = 0, and the principal stresses
are o, (radial), o, (tangential), and a, (axial), which is equal to zero. Maximum
shear stress is given by

|Oo = G,|
2

loo — 9,
to =the larsestor = (18-13)

Noting the sign of o, and a, in equations (18-10) and (18-11), we see that
Tmax
will be equal to

Og — GO,
Umax = )

for the case of internal pressure. Using equations (18-10) and (18-11), we have
a*b*p
Tmax = Gee) (18-14)

By inspection, maximax shear stress (maximum for any location and

orientation) occurs at the inner surface and is given by
b? O(a) +
Tmax(@) = 59 == OF Tmax(@) = —— (18-15)
 854
Chapter 18: Axisymmetric Problems in the Design of Machines

Example 18-1. Given: A pressure of 4000 psi in a pipe with 3-in. inside diameter and
0.5-in. wall. Neglect axial force. (a) Find maximum stress. (b) Find maximum
shear stress.
Solution: Part a. Using equation (18-12), we have maximum normal stress
given by
b? +S,
oo(a) = Pes + a? = 14,300 psi.
The approximation for thin wall cylinders gives a result about 16% low for
o, in this example of a thick wall cylinder.
Part b. Maximum shear stress occurs at the inner surface (r = a) where
0, = —p. It is

t max
= OP _ 9150 psi e

SECTION 18-2
Interference Fits
Loose, free, medium, and snug fits (classes 1 through 4 respectively) provide
for positive clearance between a shaft and hub. For a 1-in. diameter shaft,
diametral clearances range from 0 to 0.001 in. for a snug fit and from 0.003
to 0.009 in. for a loose fit. Force and shrink fits have negative clearance
(interference); a heavy force fit or shrink fit of a hub on a 1-in. diameter shaft
has about 0.001-in. diametral interference. Although the equations of the
preceding section are based on circular cylindrical bodies, they should give
reasonable results for bodies with irregular external surfaces, such as gears
and pulleys, where an effective outside diameter can be approximated.
Let us consider an interference fit between a solid shaft and a hub of outside
radius b and inside radius a. Subscripts h and s will refer to the hub and shaft
respectively. The resulting pressure p will subject the shaft to a uniform stress
6, = 6g = —p throughout, and a radial displacement at its surface.

Uu, SS
al rf v,)ap
E Ss
(18-16)
The shaft length and the axial length of the hub will, of course, be different
but nevertheless we will assume an axisymmetric stress field. Then, for the
hub, the internal displacement is

ap (b* + a?
uy, = ie Be ae + Vi (18-17)

Total radial interference U is made up of the two radius changes (Figure 18-2)

b* + a?
le eke sive l-—v
U =u, — u, = ap\—————__ + “ (18-18)
E, E KY
 SoD
Section 18-2: Interference Fits

Radial interference U
(exaggerated)

Figure 18-2 Interference fit.

An approximation of torque capacity and force required to assemble the

shaft and hub may be made on the basis of a coefficient of friction around
fe Oo.

Example 18-2. A 4-in. diameter, 2-in. face width, 20-tooth cast iron pinion is to
transmit a maximum torque of 1200 in.-lb at low speed. Find the required
radial interference on a l-in. diameter steel shaft and the stress in the gear due
to a press fit.
Solution: The shaft radius a = 0.5. Based on stub teeth, let us use the deden-
dum radius

Poisson’s ratio will be taken as v = 0.3 for both metals and Young’s modulus
E, = 30 x 10° for steel and E, = 15 x 10° for cast iron. For torque
T = 1200 in.-ib, required tangential force at the shaft is F = T/a = 2400 lb.
Based on a coefficient of frictionf = 0.1 and hub length L = 2, the same as
the face width, the tangential force due to friction (at incipient slipping)
F = 2naLfp = 2400 Ib
from which the minimum contact pressure is 3820 psi. If we include a 50%
margin of safety to insure transmission of power, then the contact pressure
will be specified as p = 5740 psi.
Using equation (18-18), the required radial interference is U = 0.00035 in.;
a nominal hub inside diameter 0.0007 less than the shaft diameter. Since
tolerances must be applied to both dimensions, selective assembly would
probably be necessary, that is, a large shaft would be assembled with a gear
 856
Chapter 18: Axisymmetric Problems in the Design of Machines

having a large hole, and so on. Assuming a radial interference that did not
exceed 0.0005 in., maximum possible contact pressure would be p = 8200 psi.
Using equation (18-11), the maximum tensile stress in the gear due to the
force fit occurs at the inner surface. At that point,
b? + a? '
6,(a) = i Bee 9570 psi

based on 0.0005-in. radial interference.

The maximum interference assumed above falls into the heavy force fit
category and is not generally recommended for cast iron. Based on the value
of oy, however, we might find the design conditions acceptable for high strength
castiron. @

SECTION 18-3
Stresses and Displacements in Rotating Disks
Considerable stress may be induced in bodies rotating at high speed.
Analysis of this effect is important because failure of rotating machine
members is particularly hazardous.
Consider a rotating disk with inside radius a and outside radius b (Figure
18-3). The force equilibrium equations of Section 18-1 did not include body
forces. If we include rotational effects, there is an additional force term
pwr? dr d@, where mass density is given by p(lb-sec/in.*) and angular
velocity is given by @ (rad/sec). For a rotating disk with no external loading
but with an internal pressure p (due to a force fit or shrink fit), equations
(18-9), (18-10), and (18-11) become .
b2

-l?, an 3 DD
o r amab2 = a yay al
8 Petey ca Ra
r (18-19)
b2

Ge +1
ie ; OY Pa ares 1+ 3v\ ,
Ol are 5st PO ee a MO si eS: (18-20)

a’p pt No) por [3+

ee | ae
# wea" Sac
a
|+ E | g |
(18-21)
«|? + P= + a2 (1 + v) — jhemaye
oaks P|

Ordinarily, inertial stress and displacement of a shaft would be neglected.

Thus, for a shaft, we have approximately

OR Ooh aD
(18-22)
—(1 — v)pr
“i=

E
 857
Section 18-3: Stresses and Displacements in Rotating Disks

Internal radius, a EIN

External radius,b

Inertia effect
pw2r2drde

ly) :

Figure 18-3 Inertia loading of a rotating disk.

We note, however, that the contact pressure p depends on rotation speed w.

For any value of «, the initial radial interference U equals the sum of the
expansion of the hub and the contraction of the shaft at radius a. Using
equations (18-21) and (18-22) for u, and distinguishing between the physical
properties of hub and shaft, we have initial radial interference
Deane
ag : l—y (ya (Seay) be
ies ap eSE, Oe “p ees
Ea + apo 2 4B, | :
(18-23)

which is required for a given contact pressure p at rotation speed @ rad/sec.

This equation is valid as long as a positive contact pressure is maintained.

Example 18-3. A 16-in. diameter steel flywheel is to be force fit on a 2-in. diameter
steel shaft. At maximum speed of 4000 rpm, a contact pressure of 1000 psi is
to be maintained. (a) Find required interference and maximum normal stress
at assembly. (b) Find speed at which the contact pressure becomes zero.

Solution: Part a. Substituting a = 1 in.,b = 8 in., v = 0.3, E = 30 x 10° psi

and p = (0.280 Ib/in.*)/(386 in./sec”) in equation (18-23), we obtain
U = 677 10"
2p 128 a 10% c57
For p = 1000 psi and w = (27/60)(4000) = 418 rad/sec, the required initial
interference is U = 0.000292 in.
Part b. Substituting this value of U and setting p equal to zero in the equation
relating U, p, and w’, we obtain
oe 0.000292
(62)
2s 0m
corresponding to 4560 rpm. At this speed, the force fit becomes completely
ineffective. @

Figure 18-4 shows a miniature turbine designed for high rotation speeds.
Due to the variation in rotor thickness, however, the procedure outlined
above would be modified if we were to find stresses in the rotor. Some
problems of this type are treated by Den Hartog [1].
 858
Chapter 18: Axisymmetric Problems in the Design of Machines

Figure 18-4 Cutaway view of a miniature high speed turbojet engine.

[Courtesy Lab Sciences, Inc., Boca Raton, Fla.]

Front fan rotors Axial compressor rotors Centrifugal compressor Ist stage turbine Exhaust

Air inlet Annular bypass duct Interior duct Combustion chamber 2nd and 3rd stage turbine

SECTION 18-4
Energy Storage in Flywheels
Heavy disks often serve as flywheels designed to store energy for the purpose
of maintaining reasonably constant speed in a machine in spite of variations
in input and output power. In a similar application the stored energy is used
in a short burst as in a punch press or shear.
Stored energy in a flywheel is in the form of kinetic energy
E, = 4J@?
where J is the mass moment of inertia and w is the speed in radians per second.
For a solid circular disk of radius b in. and weight W Ib, J = b?W/2g¢ lb-sec?-in.
where g = 386 in./sec?. In the case of a heavy rim of inside radius a, outside
radius b, and a light hub

For a conservative (loss free) system, this energy is available as work output

workj= EW or |m2 = 5Ja’

Example 18-4. A 10-in. diameter solid disk weighing 40 lb rotates at 3000 rpm. Its
energy is to be used to punch metal during one third revolution of the disk.
What is the average torque available?
Solution: We set
T60 = Jo”
where 6 = 2n/3 and J = b*W/(2g).
 859
Section 18-5: Design Based on Plastic Analysis

Using w = (27/60)(3000), we obtain T = 30,400 in.-Ib if the disk goes from

3000 to 0 rpm in one third revolution. The braking torque required to stop a
similar disk in one third revolution would have the same value. @

SECTION 18-5
Design Based on Plastic Analysis
Due to the fluctuating nature of the loading of most machine members,
plastic analysis is seldom appropriate to machine design. In certain cases,
however, a saving of material can be effected by designing on the basis of
yielding throughout a part, rather than basing the design on loads at which
yielding first occurs. The thick wall cylinder furnishes one such example.
We will base our analysis on the maximum shear theory. Suppose we
assume an idealized shear stress—shear strain relationship as in Figure 18-5.

The elastic region

In the elastic region, the equations of the first section of this chapter apply.
In particular we will utilize the equilibrium equation

oy =d(ro,) (18-25)
from which

do, 09-9,
(18-26)
dr r

and the equations for radial and tangential stress due to internal pressure p:

oO, co a b2 a aeazae (18-27 )

b2

a’p al + 1

OV) = ee (18-28)

For the case of internal pressure, maximum shear stress at radius r is given by

Og — GO,
T max a 2 (18-29)

In the elastic region, we may use equations (18-27) and (18-28) in equation
(18-29) to obtain

Umaxiam 52
2D 7[. \ (18-30)
 S60
Chapter 18: Axisymmetric Problems in the Design of Machines

Shear
stress
i
max

Shear yield Plastic region

point
;
syp
Figure 18-5 Idealized stress-strain relationship for
elastic-plastic material.

Shear
strain

The maximax shear stress occurs at the inner radius. It is

b*p
Tmax(4) = (18-31)
lik ae

When t,,,,(@) reaches the shear yield point S,,,,, we are at the transition between
the perfectly elastic case and the elastic-plastic case. That is, yielding begins
at this instant. The pressure corresponding to the beginning of yielding is,
from equation (18-31)

P= Ssyp pz (18-32)

The elastic-plastic region

At internal pressure p, obtained from equation (18-32), yielding has just
begun at the inner surface. If we continue to increase the internal pressure,
yielding will progress through the thickness of the cylinder until finally the
maximum shear stress at the outer surface reaches the shear yield point.
During the period ofyielding, the equilibrium relationships still hold, but the
stress-strain laws are not the same as for the completely elastic region. The
equilibrium equation (18-26) for axisymmetric loading still applies

do, O6,—9,
dr r

but the maximum shear stress must not exceed the shear yield point

Tmax = ae a =D eyp (18-33)

 861
Section 18-5: Design Based on Plastic Analysis

Let internal pressure be slowly increased until yielding progresses just to

the outer surface. Then, at that time, the maximum shear stress will be a
constant throughout. From equation (18-33),

ee = Tee (18-34)
Using equation (18-33) in equation (18-26), we obtain the differential equation
governing the state of total yielding. This will be considered the failure
criterion for plastic design
do, 28
es 18-35
dr r ( )

The general solution of equation (18-35) is

O61
— 2S Syp —.(\nin+ C)

The arbitrary constant C is found by equating the external pressure to zero.

o,(b) = 0 = 2S,,,(In b + C)
from which C = —In band we may write
b
o, = —28 aaah) : (18-36)

for the completely plastic cylinder.

The pressure corresponding to plasticity throughout the cylinder is given
by the negative radial stress at the inner surface. For plastic design, the
failure pressure is
b
P2 = —a,(a) = 2S,,, In- (18-37)

Since we have used the maximum shear theory, the shear yield point should
be taken as half the tensile yield point. The allowable shear stress is
S Ss
Ssfattow) = a = ae (18-38)

for safety factor N. Using equation (18-37), the allowable internal pressure
for plastic design will then be
eye
P(atlow) = —In-
Neon (18-39 )

This result is reasonably valid if we use a ductile material and a generous

factor of safety so that expected pressures do not exceed p,, the limit for
complete elasticity. Then, in the event of a single overpressure up to p>,
failure will not occur. This criterion is unsatisfactory for repeated loading in
the vicinity of pressure p,. In the case ofa brittle material, plastic analysis is
 862
Chapter 18: Axisymmetric Problems in the Design of Machines

inappropriate. When the ultimate strength of a brittle material is reached at

any point, we expect a crack to form and propagate through the part,
possibly resulting in a catastrophic failure.

Example 18-5: Plastic Analysis. Consider a thick wall pipe where a = 1.5 and
b = 2. Find maximum allowable internal pressure, before yielding begins,
and on the basis of plastic analysis. The yield point of the material S,,, = 60,000
psi and we will use a factor of safety N = 4.
Solution: Based on yielding, maximum allowable pressure is given by
equating o,(a) to the working strength
b? + a? S
d,(a) = p = 3.57p=—
b? — a’ SOP N
from which the maximum allowable pressure based on the maximum stress
theory (in terms of the onset of yielding) is

P (allow) = 4200 psi

We will now base our results on the maximum shear theory and the onset of
yielding. Maximum shear stress occurs at the inner surface where

i pepe? = DDG

Maximum internal pressure is given by equating t max to the working strength

based on shear yielding

Syp

Saw = 2N
from which
S
DID = 2
°P = 5N
or
P(atlow) = 3280 psi

Now, suppose we assume an idealized stress-strain relationship, as in Figure

18-5, and design on a plastic basis. Using equation (18-39) we have
S b

SECTION 18-6
Initially Curved Members
When a machine member with considerable initial curvature is subjected
to pure bending (Figure 18-6), the form of the stress distribution resembles
that of the thick wall cylinder except that there is a logarithmic term. The
results for a curved rectangular bar (see Timoshenko and Goodier [2]) are
Cc
o,=C,+—¢+C,lnr

(18-40)
0g = Ci Tema Is C3(1 + Inr)
 863
Section 18-6: Initially Curved Members

Figure 18-6 Bending of an initially curved member.

This stress distribution is independent of the tangential coordinate 0, but,

unlike the above axisymmetric problems, there is a 6-dependent tangential
displacement.
For boundaries at r = a and r = b, as in Figure 18-6, we have

g(a) = o{b)= 0 (18-41)

Moment M at the ends yields the condition
b

Mi -c{ Oor dr (18-42)

Applying the boundary conditions, we obtain the arbitrary constants from

which tangential stress
22 b a
-am(—" In?+ In + aPin’ +B — a
r a b r
Og =
b 2)

1G 1d) —s4asbs (in

||

The condition of zero net load on the ends is identically satisfied. Maximum
tangential stress occurs at the inner surface and radial stress is nowhere as
great as this value. Substituting r = a, we have

~4m =e
sx No in?
a
O(a) = ake 5 ea ae aA | (18-43)
1G 19°)? = 4a>bs in4 |

A detailed treatment of curved flexural members is given by Seely and

Smith [3]. They examine I and T sections as well as circular and rectangular
sections.
 864
Chapter 18: Axisymmetric Problems in the Design of Machines

PROBLEMS
1. A steel cylinder with 1-in. internal radius and 3-in. external radius is subject to
10,000 psi internal pressure. Axial stress is zero. (a) Compute and plot tangential
stress at r = 1, 2, and 3. (b) Computeand plot radial stress at r = 1, 2, and 3.
(c) Compute and plot maximum shear stress at r = 1, 2, and 3. (d) Find factor of
safety N based on yielding due to normal stress where S,,, = 60,000 psi. (e) Find
N based on shear yielding where S,,, = S,,/2.
. Repeat Problem 1 for a cylinder with 1-in. internal radius and 2-in. external radius.
Plot values at r = 1, 1.5 and 2.
. A long steel cylinder with an inside diameter of 4 in. is to hold liquid at a static
pressure of 8000 psi. (a) Find the required thickness based on thin wall theory,
using a yield point of 68,000 psi and a safety factor of 4. (b) Find the maximum
tangential stress and the factor of safety based on the maximum normal stress
theory, using the thickness found above. (c) Find the maximum shear stress and
the factor of safety based on the maximum shear theory.
. Repeat Problem 3 for an internal pressure of 4000 psi.
. A 2-in. inside diameter cylinder is to be made of steel with a yield point of 75,000 psi.
Find wall thickness based on the maximum normal stress theory and a safety factor
of 3 for 5000 psi internal pressure.
. Repeat Problem 5 using the maximum shear theory.
. A long thick wall pressure vessel is subjected to internal and external pressure. It
is free to expand axially. Show that axial strain is constant, that is, independent of
the radial coordinate. (This condition justifies our neglecting axial stress except
that caused by pressure on the ends of the vessel.)
. A thick wall cylinder is subjected to internal pressure p. Plot o¢/p, ¢,/p, and Tnax/P
versus r for the elastic case; b/a = 2.
. (a) Find allowable internal pressure p for a 6-in. O.D. 4-in. I.D. steel vessel based
on the maximum normal stress theory (that is, o, shall not exceed S,,/N) where
S,» = 60,000 psi and N = 3. (b) Repeat using the maximum shear theory.
10. Find (a) interface pressure and (b) maximum stress due to an interference fit between
a 1.5-in. diameter solid steel shaft and a 6-in. O.D. steel disk where radial inter-
ference = 0.0005 in.
11. A 2;-in. long by 3-in. outside diameter hub is pressed on a 14-in. diameter steel
shaft. (a) Find tangential stress in the hub if
interface pressure is 25,000 psi. (b) Find
the radial interference.
12. A 5-in. O.D. aluminum disk with 1.5-in. axial length is press fit on a 2-in. diameter
steel shaft. Radial interference is 0.0012 in. Find low speed torque capacity for this
fit based on a coefficient of friction of 0.1. Plot radial, tangential, and shear stress
versus radius for the disk (when torsion free).
13. A 10-in. diameter hub is shrunk on a 2-in. diameter shaft producing a pressure of
10,000 psi where the surfaces contact. (a) Find maximum normal stress in the hub.
Where does it occur? (b) Find maximum shear stress in the hub. Where does it
occur?
14. An 8-in. diameter steel disk is force fit on a 1.5-in. diameter steel shaft. (a) Find the
initial contact pressure required if contact pressure is to be 500 psi at 3000 rpm.
(b) Find maximum stress at zero speed.
 865
References

15. Solve Problem 14 if contact pressure is to be 200 psi at 2500 rpm.

16. A 6-in. diameter steel disk with 2-in. axial thickness is force fit on a 1-in. diameter
steel shaft. The radial interference is 0.0005 in. At what speed in revolutions per
minute will the fit become completely ineffective?
17. Repeat Problem 16 for a 4.5-in. diameter disk.
18. The speed of disks A and B in Problem Figure 18 must be reduced from 2000 rpm
to 1000 rpm in 3 rev. Find the average braking torque required. The disks are made
of steel with a density of 0.28 Ib/in.? and rotate as a unit.
2 in.

Problem Figure 18

7 in. dia.
8 in. dia.

19. Find the average braking torque required to stop disks A and B in Problem Figure
18 in 10 rev if their initial velocity is 6000 rpm and they are made of aluminum of
density 0.10 Ib/in.?
20. A thick wall cylinder is subjected to an internal pressure p. Plot o,/p, o,/p, and
Tmax/P versus r for the fully plastic case; b/a = 2; p is the failure pressure.
21. Design a pressure vessel using (a) thin wall theory, (b) maximum normal stress
theory for the elastic case, (c) maximum shear theory for the elastic case, and
(d) maximum shear theory for the plastic case. The yield point in tension is 120,000
psi; the shear yield point is 60,000 psi; internal pressure p is 20,000 psi; the inside
diameter of the vessel is 6 in.; and a factor of safety of 3 is used.
22. Acurved beam is subject to a bending moment M. Show the effect of initial curvature
by plotting

A on hes
oO

where
6M
C6; =
°~ ib — 5ay
represents maximum bending stress in a straight beam.

REFERENCES
[1] J. P. Den Hartog: Advanced Strength of Materials. McGraw-Hill Book Company,
New York, 1952.
[2] S. Timoshenko and J. N. Goodier: Theory of Elasticity, 2nd ed. McGraw-Hill Book
Company, New York, 1951.
 866
Chapter 18: Axisymmetric Problems in the Design of Machines

[3] F. B. Seely and J. O. Smith: Advanced Mechanics of Materials, 2nd ed. McGraw-
Hill Book Company, New York, 1952.
[4] J. H. Potter (ed.): Handbook of the Engineering Sciences, Vol. 2, Sect. 13. Van
Nostrand-Reinhold Books, New York, 1967.
[5] C.-T. Wang: Applied Elasticity. McGraw-Hill Book Company, New York, 1953.
[6] st International Conference on Pressure Vessel Technology Proceedings, (3 parts).
American Society of Mechanical Engineers, New York, 1969.
[7] R. W. Nichols (ed.): Pressure Vessel Engineering Technology, American Elsevier
Publishing Co., New York, 1971.
Material Properties

867
Table A-1 Mechanical Properties of Some Cast Ferrous Materials

i
: & pe, en bo m
¢ : a A Z ze) d=)
cas Zs € re s = 3 2 < ayes
i Os m § S oe = aa a = ge
Se8 282 5 Sez
ao 2
a B R2 =ise)
3 83< ence
a &e Sed
zg ba
ie
cs
AB.
= (= 3s €& Z ~
e206 = 2 SS < BS ve 2 ie a
5 3 = 6 Ss 39 - S so
Zz Zz n& EY ae
aS) 3=| £22 mo 32 30
= 6.
2 a2 a. oO
ie} oO Pal fea & i= n a
= & = & =

Gray Cast Irons

ASTM, Class 20 12 X 106

ASTM, Class 25 13X 106
ASTM, Class 30 15 X 106
ASTM, Class 40 17 X 106
ASTM, Class 50 19 X 106
ASTM, Class 60 20 X 10°

32510 (ASTM A47) 25 X 106 25 X 106

35018 (ASTM A47) 25 X 106 25 X 108
45010 (ASTM A220) 26 X 106 23.2 X 108
50007 (ASTM A220) 26.5 X 106 23.2 X 10
60003 (ASTM A220) 27 X 108 23.2 X 106
80002 (ASTM A220) 27 X 106 23.2 X 106

60-40-18 un
23.5 X 106
or 60-45-12 \A536
80-55-06 Ga ) 23.5 X 108
100-70-032 ( ” ) 23.5 X 106
120-90-022 ( ” ) 23.5 X 106
Heat Resistant 18.5 X 106

SOURCE: Compiled from /973 Materials Selector, Reinhold Publishing Co., New York; Tool Engineers’ Handbook, 2nd. ed., McGraw-Hill Book Co., New York, and
ASME Handbook-Metal Properties, McGraw-Hill Book Co., New York.
* Obtained by heat treatment involving normalizing or quenching and tempering. € Normalized.
> Below 8% total all content. 8 Wear resistant.
> Normalized and tempered. “ Corrosion resistant.
Quenched and tempered. ' Heat and corrosion resistant.
> Annealed, 1 2000F, Water quenched,

868
Table A-1 (continued)

ksi BHN ksi

Number
or Hardness
Specification Tension,
in
Identification Yield
Strength, Elongation
in ft-lb
(Charpy),
of
Elast.
Modulus Strength,
Tensile Endurance
Limit, of Compression,
in
Modulus
Elast. Shear
Strength, Impact
Strength,
Strength,
Compressive

Cast Alloy Steels>

Class 65,000 30 X 108

Class 80,000 30 X 106
Class 105,000 30 X 108
Class 150,000 30 X 108
Class 200,000 30 X 108

Class 60,000° 30 X 10%

Class 70,000° 30 X 106
Class 85,000° 30 X 106
Class 100,000¢

Ni-Hard Type 28 25 X 106

Duriron® 23 X 106
Ni-Resist Type 1! 16 X 106

29X 108
28X 10
28X 106
24 X 108

869
Table A-2. Mechanical Properties of Plain Carbon and Alloy Steels
(based on a 1 in. diameter specimen)

Machin-
Tensile Yield Elongat. |Reduction Hardness, | ability
Condition Strength, | Strength, | in 2 in., in Area, BHN
(Based on
ksi ksi %
1112 = 100)

HR & turned
CD

SOURCE: ASME Handbook-Material Properties, McGraw-Hill Book Co., 1954; Ryerson Data Book, Joseph T. Ryerson
and Sons, Inc., 1965.
NOTE: HR = hot rolled, HRA = hot rolled annealed, CD = cold drawn, CDA =cold drawn annealed, HRN = hot rolled
normalized, A = annealed, N = normalized.

870
Table A-2 (continued)

Machin-
Tensile Yield Elongat. es
Condition Strength, ] Strength, | in 2 in., Hardness, | ability
ksi : BHN _ | (Based on
1112 = 100)

871
 | poos AoA plO@-LLZ (Gr, || GSE || ali || BBFC LO-87l
0} pooy
: poos Aran | p76 1-19 18-6 | €9-8S] 87-L | 089-881 | Pe-LEz
0} poo’
Lees: pbLI-ZS€ 06-01] OL-%Z] 87-9 | O9-8E1 | LE-8SI
“pu
aq sn | 677-10h €S-0€ | 97-01 | LOZ} 061-881
‘puuy
aq ysnyy | PECT-BBE Z-O1| IS-SE} ve-z1 | OL-~hI} LIT-O61
pocc-LS§ -LSZ "p[uuy
:-
u£90P
EOI-LS
901-097 | SI1-<z¢ | o09Is 09-Oh]
tpl]
€01-091
pCIZ-IZE
87-71
89-ZII
cde
Seiad
isn
aq
ae
(3
|
p6CC-LC9 es
pc9c-109 COV-88Z | OET-SOE |uSS%O Foy py=ws poo
| —PCOI-IZE ES-9I} €9-%b] OF-O1 | 19-801] 96-€F1
pSST-SSs PII-0Sz | €z1-78z | 0S98 bood
vc-109
pl €€-61
89-8p]
ZL-9E}
68-€11
79-98
BOI-OLT
oyPEST
“792
SII-STE
x0S19
ney
|
pI hc-109 ZOL-OSZ | 9IT-ZIE | OSIS Boas .
- 2
LII-8bz | 8zi-sog | OSIy | ornmeg | ° 6L1-S6 y 00I-
: IL-8h
Be- eee |e

PpouspseH
pC9T-8LS

98[991$
plvc-8LS 001-0b7 611-067 Ovls
pLOC-VES

IdA*[20X9
8-8C7 VII-8L7 01S

0}
p£6C-SSS

0} "p,.3
OE
1-877 -V8TCHI Over (,$90°)

*[20X9
pSEC-8LS
19)
OOI-1S7 LI1-06¢ 1b
0b

“IoA
“p,3
(,SP0')
pecc-SSs C6-6C CI
1-087 Orie
pStC-8L$ DES 00
1-C87¢ Orel
pLIC-S6P £6-0€7 ST1-OS7 (0£98
(,9~0")
(29)
pcOc-19v 68-L6I 86-VET TL OE 100d
(,,840')
|
790,
suluspieyySopely ‘s[9a1g AOT[V S[90}§
poezlingie ‘uogseD g ule

yyguans joedwy
(leulg)
SuruTyoRy

(1124490)

arg “(Poz1)

jo uononpoy

ul uoNesuClg
z =

alo
es
@
o=
==
3)°
ic 5 xa is) som ra
ste) N 2) $
ze z = g. ae 28 Self ears
nb 5 3 Y Pn = 3 ee Sj Bn ad],
Zz xe
2 2D 58 =
2
=e as adi S) 2 Se a8 eg
: a 3 ISIV ce S a =e =5 Xs 3 2 2 a 3 ISIV
5
ssoupiey ar)
ge
og
a
> ssoupiey = =
eS[991S apeiy
Huluapsey pue
Bulziunques jo
jed1ueyds|] saijedoig BIqeZE-Yy
 poyouanb

way) 0}
“4.0551 :(P2}BOI
poyeayos
pure

0} JoyeM
“Ao00E PUL
pareayas ‘pajoooASIMIOYIO
ul 1-00
“ALOSEpayouanb
jod
“4,0S9T 18 SSATUN)
posodura) ‘1y
18
‘4, B40}
pasadusa)
12 ¢/ SMOT]OJ
poztpeusou
¢] ALOOLI
ay)
0}
pue
SE
pur
payrayas Po}ed1]
2
payouonb
ee [to peztinqgied

(dio
ur 1LOY
YSO1D
“4,
payouanb
19}eM 0S9] 1B

JaaIg
Wayatyieg
PUP :UONIPUOD
(}U9}U0D
1B
“4 O9OID “4QSS1
poztpeusou
OSp]
UOQIeD
BUIMOT[OJ
0} ‘OSOID OF
pareoyas
st payeoyor
UI_ “pa

BUISeOIOUT
‘Saljiadodg

O0/D
:SMol[OJ
94)
“AOOLT UI
“A,00€‘J
SPeLIa}eUL
JO
SE OS9T
payeasy OPI
18 1-006
I®
AAU.

poztinqies paztewsou UI
10J
ROY38 podULIIL)
posoduia} ‘ATaAToodso1
UsAPOPJ291SPUD

ou
ore s[eLIDVeW,
st s[eIIeW
BT uoy) pp/
TTD A,00p
[J
“AOOET
puke
pue “A,OOET-OOp “1OOET
10}eM Pur
S|LE}OP1AJO1O}

LI
TIOOF A,00E
UT O}
A,00p
“A,OSEpayousnb
8 OOF
JO
poradusa}
Ipe“e9€

“YIOA 10.J) VIOUI

Woy sainjesoduia}
Wor
1#
“E9E ENE

“6tb SIb

LAKE ME
sSOPPIHH

“41,0091
parodusa)
soinjerodus) soinjeisdus)
SIp ‘SIP

pur
SUIZLINGIeD4y 3

[lo Surioduia}
puke 93
ut
pareayal
“OD MON

payouanb
Sulodw9} poyouonb Sutsadui9}
Burystgng

ye
“4 paure}qo
OOLT
sayem
ye
‘4 Jo
00ST
afqeutejqo aduvs
‘sjaa}¢9

syouelmW ‘4ojIa;ag ployuley

“ASTPI18
sanyea
pezewsou
01 0}
(penujuos)

pareayas spuodsasi09
“pojou sonjea juasaidas
0}
AOTTY

poyeayas
a1aym Jo st ‘4,SLST
O€O/D
a8ues soin3ty
sanyea
‘SEL LEI
‘9bT OSI

SVEN SUE

jd90xa
“payooo
“1a]eM
‘SO 79

0} “A,00E1 OM Jo
P8I “LOL | SO

paztjeusou
puocdsass0o
‘spunol aduLY
yod SpuNos UL
1-00
Wor EL6]

‘ry payouanb
“JOSIJO
Gli We ||

“spunos
“EOI ZO |

“ELI 891|
‘O8I O81|

“ul-]g 12
st “ul-S*()
:-qOunos payidwop

103 sadups
posoduia}
(punos
10j AgSLOT %T"O
“Ul-¢°Q *45$/
eqey¢-y

satjiedoid
“TPT 10}
“LOZ

yout Ve satjiadoid ST
“€8
08
“LLZ
87
‘6Il
SII |‘88I
“88€
Ive
‘6rI
OZI
LOI 38 1D
voy) YIBUAS 10j
C/7) 0}
PaZUNgieD
10J sarjrodo1g
pue pareayay
adesaAy
yd2oxq
fio [¢7 PIA 9100
tft 1D
“A,
‘4OS9T
“A,00E1-00p
“4
S/S
“4,OS9I
“A,O0ET-00h
7£7/D
pur
payouanb
SLpT
660/D
porodwiay
pareayar
st
pozifeulsiou
posoduray
pozeulsou
uoyy
ur
[to
pue
“4
poreoyar
1-00
00E
voy)
ul
jlo
18
0}
1I12 “A,0SS1-SSPI
payouonb
ut
pajeyse
[lo
pue
Udy) ¢]
*porodusa}
» g 5 3 g 4 ‘40091
$7
payeayad
‘4
payouanb
ul
jlo
pur
vay)
“poxeduia}
paztewsou
0}
Ve
i “40ST
“A,0SST
pareayey
OF |“ASLPT
Ppyy
ysnomy
“pauapiey
5*s}[NSO1
BPZUIS
jeOY payouanb
UT
payeayay
O}
“19}@M OF “AGSLST
poreoyey
OF yy yPAZMPUWION
OF “AOSST
“AQSTIT
‘40591 Poztfeulson
32
P2ZtEWION
pareayal
Ie

873
Table A-4 Mechanical Properties of Some Heat-Treated Steels at Various
Tempering Temperatures

f
=z |g
=]
o' c
S$ S %
#/s*| 5 %2in., %
2in.,
°F.
Temp.,
Tempering
3 § °F.
Temp.,
Tempering Yield
Point, Elongation
in
°F.
Temp.,
Tempering Yield
Point Elongation
in
Yield
Point, 5o a] Tensile
Strength Strength,
Tensile
Strength,
Tensile 4 of
Reduction
Area, of
Reduction
Area,

C1040 quenched in oil at 1575°F C3140 quenched in oil at 1525°F C5150 quenched in oil at 1525°F

SOURCE: Data compiled from Modern Steel and Their Properties, 7th ed,, Bethlehem Steel Corp.
NOTES: (a) All the data in this table are based on single heat results,
(b) 1-in. rounds, were treated for 1040,1060, 1095, 1137, 1144 and 9255S steels, 0.565 in, rounds were treated for 1340, 4027, 6150 and 8750 steels. 0.530+in, rounds were
treated for 3140, 4063, 4130, 4140, 4340, 5150, 8630 and 8650 steels. However, for testing, all specimans were turned down to 0,505 in, dia.
(c) Since every grade of steel ranges in composition and the results may vary from heat to heat and so can cause significant differences in the properties attainable by thermal
treatment. Hence the properties listed in this table should not be considered as maximum, minimum or average values for a particular application of the grades involved.

874
Table A-5 Influence of Mass on the Mechanical Properties of Heat Treated Steels

- Tensile a Ae: Reduction Brinell Tensile Elongation Reduction Brinell

Strength, in ei Hardness Strength, in 2 in., in Area, Hardness
psi. psi. %

104,750 146,750 131,500

96,250 141,000 121,500
92,250 128,000 100,500
90,000 125,000 81,500

142,000 147,000 141,500

139,750 141,250 129,500
134,500 133,750 116,500
130,000 121,500 94,500

C1141 oil quenched and tempered at 1200°F C8650 oil quenched and tempered at 1000°F

87,400 177,500 168,750

69,600 172,500 159,750
65,300 165,250 148,500
60,300 143,250 113,000

118,000 194,000 163,000

112,000 188,000 156,000
107,000 145,000 114,000
105,000 136,000 105,000

C9255 oil quenched and tempered at 1000° F

133,000 122,500 170,000 146,500 331

128,000 113,250 164,250 133,750 321
114,500 91,500 154,750 102,500 302
101,500 77,500 149,000 94,000 293

145,000 218,000 198,000

139,000 216,000 197,000
134,750 211,000 192,000
124,000 200,000 149,000

SOURCE: Data compiled from Modern Steels and Their Properties, 7th ed.
Bethlehem Steel Corp. except for C2340, C3140 and C9840 which is from
the International Nickel Co., Inc,

875
 yeayysod
UTIM
quaT[aoxq

qua|[2oxq

YIM -oid

aIQeploA\
-a1d IO} “WlOda1
spoou—s1e

jeoId 91e9
YIM JuaT[aoxq

pood
AMIGEP IOAN

quatjaoxg
yuatjaoxg

qualjaoxg

yeaywy
poyesu

Ie
ney Tey

rey
J suorneoaid
JON
"WOdal
jeay
w
yeoujsod ‘ey
ysod
-ue ey
UOoIsNy
SUIP[aM ynq
aievo
9q
udye}
JsNU

NH@ uo paseg)
Ayiqeuryoey,

“VMWuyT

isd
aoueinpug

29 Pouspsey
‘yiduass yoeduty
gry (poz)

'pS61 OA
pajeouuy

MON
poreduiay

“OUT OD
Y pouopsey
ssoupsey [jug

Yoo
NHd

|
OSI ssl
S]A91S

MMIN OP
payeouuy
1yYBNOIAA Ssajule1S

‘Sadadodg s/Dlay
09 c09
“Baly JO "par

SY
€91

pojeouuy

YOOYPUDL] TWSV LOK

poraduiay
3 pouapiey
BWOS

suo
ZulSul
pajesuuy
MON “02D Surysyqng ployuloy ‘voJrajag sjoLialoW £26 [ Wiosy poyid
jedIUeYD9W) Saladolg

“YBa

posedwiay
3 pouapsepy
jo

1S
Play

pajzauuy

poseduay
FW pouepsey

“Wauans
Is}

“NH#
“UMPIP
pajeauuy
ontuejsny
BIGeL9-Y

OCCT-O0T
BIE. SEM OD
SOTE

Pjoo
JOEb
LE Bre

oIIsuaj
?-ADUNOS
OAV

pue
9}
“OLE

O19
cOE
por

PauApPseH
O€r pajRauUY
ICE

se
Och

Ob
lev

OIE cOr OlP +} SOv bP tunumtul

ssonyea
viv
SIP
II

p“SONIPA
ABLIOAY 5 |,
5paradway
poo
“UMRIP
pur

876
 ‘yRunes WoW Batyejos ase | pue qd ‘D ‘WV -ALON

877
*ROLAWY JO "OD UINUILN| y“yooypuDFy

220K
OMY
MAN “OD Turysygnd ployulay ‘assy 40199/9g S}aUaloy

UL
[BANIINAIG BOI VOD [CIOW SPLOUADY WL ‘yoo vivg t
puUroAd|g “oD Burysyqnd uous oy “MIsag Yanp-anssy aouasafay Spolay 0L6] Wor) poliduios sem MEG -ADUNOS

Chee

0-6€0L 0-SLOL
1S)
“yypesoute 10 syed jeanjon4yig

EAE
suoleoydde ajissiw
pur s1uasoA1o papjamy
a

Os
OU -9909
SAINIONAYS PapjaM 404

scar

(ae
suoisni}X9 pur ssulgio4 > = Srl ep

cl

0
sasn [esnjoayoue

9€
ce
“ourmpiey ‘saurpies ‘adig iE el 0 -€909
*sdg[ie4} {sued “YoY ‘s[les adpug
OP
Sb
cl
SMR

S6
OE
val
q
‘sa8uey ‘adid ‘sasn auteur *soinjonsis
0-1909
881
Sc

qd

0€
Cl
Aynp Aavay ‘*),ba uoneyodsuesy

6
S Lo 6c | 8IH-

Vv Vv Vv
BOG Eee
sapling
[imac Cae cae
“$agn) Pa[loo SaRApseY

Ul vl ot

Of Iv Is
= = =
a<t<
‘suoneoydde [einjoanyoie 9 cc SE lates

qa
‘s]Isua}n Suryooo ‘saouryddy Of 9 81 0 -SO0S
-jddv jesnjoayyoue “Buidid ‘syur) advsoys v LC 6C 81H-

ss
Ol
ol
Vv
eseecsee.
ma <
*sjassaa ounssaid *)8a [roles ‘ouem

aed
eae

<<
0€ 9 91 0 -800€

Vv

87
I
L
-piey slapping ‘sjaued yons) “ysoM JONG

Ocl Sel
yIL
SL
98.L-

dq dq
ley se

Ip =
c =
JJesdIIe “s[aoyM

aa

AA
syird Jean jonsys

ae
07 Ly 89 val
yon} ‘s}onpoid ouryoeut Mag
€l gt LY c Il Lc 0 -¥707
suojsid ‘spray sapulAd 9L-8 107
sjonposd auryoew Ma19g
*souinsy YONI)
“SQUN JONAS LLANE “SHoqui aut
[EAN joNsys “SBUN Ly [eI ‘sprEg
Jaaoys Jamod *sdurdsoy Ainp AAcay

EL-110C
sjonposd auryoew Masog
“sjisuain 3UL4009
TYIOM [PAU aAOYs “s1OJI9Lo1
‘juawidinba jeormayo ‘syur) ades0js
pure ainssoid *suaBuryoxe icay se iS €l 0 -OOll

“MYL YO /I
¥ wn
al] 6 3

poy “eq “ul an

‘jsisay UOISOLIOD
g = mo = = 5

“ul Z ul “Buol|q

m
SuIyIeID ssa1js
|| ce

BUIyIOM PIO)
6/3 32

Ayypiqeuryoey,
sada L.

SUIPIAA\ SED
$

“pley [feu
s A

SUIpIOM vy
QO 2 (S Ss oe

OUBISISAY
SUIPIOA
~~ S = —

Surzeg
a
suoivoyddy awog @o aS = a e ae go pur

NH

Ys
~ S Sea aa So

%
ga mates Ss. 33 S
Aoulv
a = = a
SKS 2
SI
2|% 2 5 S
Zila = s =o

ays
: g

a)
sAo}|W Wnulwniy 1YBNOIA\A BWOS jo saledoig jediueyoay) Z-Y aIGeL
 Tr
AVTEQEPlaAy
OUPISISAY

Me
av
<

Se

een

Y pareary
uonnjos
yi8uasg

1ealfy1a}
ISy
seays uoNnjos

pasy ®P parrary
‘dutog plata “TS

uonn{og

“]ea1] ea
1S uOIN|OS
gia

siupyeo
pasy 7 porary
uonnjos “Ayoanjoadsou
Sy
pug

pow
“yeo1] yeaH
uolinjos
SAO||V

908

uRUIOD
SSOUPIE}Y
WNUIWN]|Y

09
jung

uonnjos
“pose
PUR PueS OF JOJo1 AO) UDAIT OUR SaN|EA OM) 9194 Mp

Aypmioyijue

year] Way
uolnjos
a'pS
20L
| ‘$9

yng
payrasy

posy % powary
ut “WZ
,uonesu0]q

uolnjog
ph’
BWOS ISeD

pasy » payeary
%

uonnjog
OAV,
:
jedlUeYyo\ saliadolg

:
%T-0)

FY parraiy
JO

uonnjog
“(1asjjo
1sx

“dulde
PI?! 4

“VROIL |eaY
juswia}ras)

uonnjog
pur
Land
SpmsaIDW ““oJIa/aG PloyUlay
a

UoIn[os
“yiBuans

pady
Y payeary
uoNn[og
pucoas aneAq
apisuay
1S)

1894] 19H
£76

uolnjos
9'pOE
Woy 94)/
aS

palidwoo

urse jeIoryiyreAq pauteygo anjea Isdt45

87
21qe18-V

Heq sem

s8uljsep
SSE
d-0r
ssunseg

:AOUNOS
Sold

Sol ep PIT0¢7
9SE
ssuljseg
‘ysisoy
“dwe]

*],ua95
asoding
ysiyq

UOISOIIOD
“}SISOY
=
=5
q‘s1e)
‘pood
g‘poo#
‘juajjooxe
ygy
Jood
Asan
=
*sajaAo
Ol
BUNEIONg
Jo
38a}
OF
X
Weaq
yout
“uauitsads
C/
82}
1p
“Pip

878
 0X9
Isd 901X6I

901 XSI

91X91

901XSI

901XSI

90IX8I
“uols Ud} UL
AWONSEA JO “Pow

Surpjan ooue
“sisay ving

Pare eee
SUPP
ary uoqied

Supa
Surutog
auajA}aoeAXQ

Sumeig
AOULV J9AT'S

Suuapjos

gAiqeuryoryy

“1G Va) + Sa) eo)

on + + isa)
IeaYyS
Is}
pejrouuy
taal
a] :
wo
a
1YBNosN 48ddoOD SAO|IY

a
~
a lo] S >
oo CA) ce)
co mn Ey
1aANpOY
ssoupsey

[4 [24 =4
is)
re)
ee
ce = Pie aisha ie

Pel RoUUy
SBWOS

ZUL
esUO|y
“ul
vor
‘oO wo foal (oa)
jedUeyDE!)y saivadolg

Oo + wow wn
jo

Is}

PleH JIPH
p
“wsuans

poyeauuy

purH
“yySuang

Bunds
W 294-] Uo pas!
apisua
1S}

89
yp

‘jayotu-oidng)

“JOATIS-[O4OIN
Surq}N9-2014]
91qe16-V

(ezu0l1q
ZyUNW) (jeJowW

pea]-luni

LOb sy
(aZU0Iq

YStH) uOdTTS
“disaig)“prey
‘sseiq (ZOL

Sth bbb ‘fbb ‘Chb

po

(gi-s¢
(sseiq Ajtperwpy )

(sst1q juoudojaaag
(sseiq

(¥q

“99%
(“OE

unt] soydsoyg)
ezuoi
eVdO

otLI
O87

SIL
ss9

asSouRSUPI)
Ly9
aspuqieD)

pue (ES
IS-[V)

(ezuo0iq

Arag SOF[RARN)
Ore

OLL
o9E
097

SLO
6£9
pay)

adcy
JOqUINN (raddoo :AOUNOS
IS

) HOP ezu01g
(Vy 1addop,
)

Por,
diss,
pos
*A[UO
“Ajaarjoadsa
pure

879
 F1geL OL-YW jed1ueysoy] saiadoig
jo awos ised saddozaseg sdoj|y

aatlssaiduiog

880
yisgueans

A
>

Bsa

=Pce
BSS
B® oe.

*,

Fw
==

(001 =sst
oe peardA
sas

qllaawe
NTs} ss
os of

3aPe

«
%
= >5

3 3
& 25
ae

% =
2D
Py 2re)=2
=

0 > oa s= n

apesy
<< a=

Is
aoe

Dn ped

Ssme
2.5
@s

S = oeo Z.

i)
o ae3
Ss oe
-— & Ss
oe
es ssSe

Ss
5= == 3 5 =f
Ss

a a<
3=
5

& ai

a
1g 8utjyno
ees

s=
Fs
“18S “Ul-010'0
198 “Ul-001°0

3 =)

pue ‘uontsoduiog
‘ssuleag “sSulYsng UO}SId “sSUL
duind ‘sarpoq *saajva

ULL
‘s1a][aduil

dl
WUed}S ‘SSUI}}IY*S1b95

“azuolq
papers]
uly ‘s}jog ‘synu ‘sivas ‘saayea duund
‘ezuoiq
QZ ‘suoysid ainssaidedid ‘ssunqy
uotlsuedxa
*squlof

auTYyoRp
[O0} ‘sSulIeaq

Ysty
popes]
[vious Sulivaq
"sash

UN} JE
“9ZUOIg
aAjPA

pol Vr
popes
“Sapo

MOTadid
ainssaid

‘sseiq
‘ssuryy

*sqied

indg ‘sivaspliqa3 ‘syed

surquinyd

Yysiy
Malos UMOP ‘s]nuU
“sSulle9q

yysuais
08
“ssP1q
MO][OA
winuruiny
‘eZU0IG
V6 auryory‘syred duund sioqjadut
pue ‘ssurseo‘savas Surpor
[[!tu
‘sduleaq ‘sraysem jeortuayo jueyd
y urlwun ‘yusudmbasue ‘ssajjadoid
duind
‘ssurseo ‘sdurjqyysureyopuke “syOOY

6
papesy jeyolu

‘9ZU01q
yy saute ‘ssugyy ainjuiny
pue
sseiq
pure
SUIPpINg
“WII) SNOLPA dIBMPILY
‘ezuoiq
VII

UODI[IS
sseiq
*syied

‘ssuLIeag
‘sivas JOYIOI “SUP
pue ‘ezuoiq
ae]
uit ‘sroypad
aayea ‘swiajs ‘sdojfadord

:ADUNOS
req stem patduioos
Wo 24) £76] SjuLdaIDW ‘vOJJa/ag PloyUlay BULYStPgnY
“OD MON YIOAPUR DUI
Y ANS Y-YOOUPUOH
[OIA Og ‘SaITdad IINH-MCLOOW
YOK “OD MAN MOK “PSO
000£¢
34
g SPIEpue ozuo1g }OSuy ainiysuy
 881
*payval) }Ray UOTNJOS,

‘sapaAo
“sao. gl ‘ysa] wiraq BUNeIOUD

[Ix
“prof
L0l
v/y
y=
*paystuly ssayun 10},eM IPs AQ PEXIPNVg
*S2JAD | “PLO] P2S49AA4 [PIXE
*soqn} = J ‘sing = eg ‘suoIsniyxa = ¥ ‘sdusvo pjow juourWied = 9 ‘spor = y ‘sduljseo pues = - ‘sayejd = g ‘sjaays = §
*Q[QEIOAR jSC9] = D ‘d[QLIOAR] Sso] = g ‘9/QrIOAL] SOW = Y :SALON
OA MON “OD SULYsIPgng PloyUldyY ‘4012a]ag SpoidaiDW £26 241 Wor payiduios sem BIeq *AOUNOS
VEIZV Se aUleS W d SL-VCE-ZH
s}UdUOdWIOD afissItwW puke }JeIOITY Vv Vv 9L-VIEMH

OL6ZV
VVEIZV
VE9ZV SB awes Vv Vv VOOINV

ules
Se
6)
VEOZW
VEIZV se
sues Vv qd
*serouleo ‘*},ba Bul|pueyY *s],jeU! ‘sjoo} puey ‘sjied ‘yoew Sureooidi9e1
VI8ZV
‘syied yoni] 2 Ojne ‘},ba aduRUpIO ‘sjaoym ‘sjled ayIsstul “jJeIOIy Wi qd
A-ATEZV se auesg Vv dq SL-V0O9AZ
A-A1EZYV se
ates Vv d SL-VO8ZV
AATEZV Se 9wUeS Vv € A-VI9ZV
*sJaoyM “BUT[pUeY ‘s].JeW ‘s[OO} [Jam [TO ‘sjred autyorul **},ba oTUOI}OaIa Vv Vv A-AIEZV
‘aqejd usojjed Arpuno}j ‘sainjxty 3? sail ‘sauTyOeU doIJJO ‘Safisstul “}yeIOIW *
pTH-AIEZV se owes W Vv vCH-VIEMH
*ajeyd uraqjed Aipunoy ‘soin}xtj 3 Saif ‘soUTYOeUL JdTJJO ‘SafIsstul “YyeIOITY Vv V bTH-F1EZV
g Se] 2s = g
3
om
= ae< g3
z S| Oe 3
ae
oS.
Lee
a
me
eS
B
@8| ce
Belz Sun
&
= o
: oe
Sie mae
|e
suoneoyddy jeordé1 a SS a oe B/43 se = a2 = |23 2] 22 adA] WLISV
aS Se ie ree Do @ a @ € se a
Se
S S a
Selbas
S| 26
s5 | 2|°2|
a
#8]
Ss
2= ==:27AV[QEPIOA
sAoj|V
wniseuBe-: pue yseg
1YyHnosjy JO
jedIUeYDa|] Salwedolg BWOS 21qeL LL-Yy
882
BIGeL ZL-y jediueyseyW) salyadoig
jo awos81q Buiysed sjevajyeyy

y, jeordA
sasq

1S

Is}

eae]
ssoupsey]
QOUL}SISOY
UOISOLIOD,

Isy ‘y}BUIIS
aAISSAIGWIO)

aourinpug

isd ‘uotsuay ut
Ayyiqeuryorw,

“yyBueNs 1e9ys
(001 = [918
Piru uo paseg)
qyay “(Adreyp)

“YIBUoNS PIA

“se]q JO snjnpop
Z Ul uOIesUuOTy
Isy “Y3uINS

“ysuons ajisuay
untsauseyy
Age
WLSV) <(16ZV:-
puz ve
pue

WLSV) (€16ZV ‘sjied

‘syied

ajisstwi
qjesioile

a
‘serauieo

aanowojny
‘ssursnoy

o1u0399[9
‘saoueidde

Iulz
Aoy[®

“sjoo}puey

+
saAnouloIny
‘sjied sdryJO ‘"},ba *sKO}
VOrOV

(1IXX
WLSV)
ouueg
se POVW[ pure osje dorp Jouwey
“SoIp

oulz
Aol[e
(AXX WLSV)

VIpOV
uty Aojfe
[
‘ssule9g
aIp SSuljseo
10} ALlep “AIOUTYORU
WLSV) “(19-€74
jejuap ‘saouerjdde
[voIsIns *s]UDWINIJSUT
WLSV) (7S-SOId

AUOUI}UY-peoT-ulL,
8 ACTIV
10}
PUL

YsI]
SpLO]

puke

ssuLteag

1e9
a}eI9POW

WLSV)
SurUTW

‘spaads

:AOUNOS
BEG sem pofiduioo

(19-€7€
Woy 24) £L6[ SppalDW "ojIa/ag ployuloy Buryst[gng
“OD MAN “WIOK

WLSV)
g0le safaAo
Aq Buljr}01
wieaq }9]
“AJOUIYIEU

(7S-SOId
POD,
UOISSIWSURI]

doueysisos
0} ‘oroydsowlye
poyxorye
Aq jes Jayem ssajun *poysiuyy
“speusnof

5 UaT]2OXA douK)sIso1
0} [eins pue Aj10 ‘seraydsourje
joy Advos ‘19,em Buyjuttd
yu‘s YOU} 201] ‘quUa]AY Uoged “OPHO]Yousja}
AIp SueuTUUNTTE
sed pue AIp pue plow 901} “sUOQIeOOIPAY
We] aduUNASISAI
0} ound [AYJepuke [AYO ‘sjoyooye “uLs99Aj4
sJajEMpur Wnajosed *s}ONpoid
Joog BOURISISAI
0} ‘UNed}s [eUNUL‘sflo JuO4)s splokpuL soseqpuke SOIN}XIW

Jo
JO
PUL
SISISS UO!JEPIxoO sjonpoid
Jo ‘s}uLdIIqn]
poo} ‘s}onpoid
199q pur PI}CUOGILD “S9BB19ANqG

*19}@M
systsay., UOEPIXO
Aq [eWOU *s}UROTIQN]

JOYOO]E
dUTIIDATI
 B1qGeL EL-y jeolUeYDeEW) saluedoug
jo @WOS |@49IN SAOIIVPUE JA49IN 8SEq o-gSAOl|

[eae
yeodéy suoneoyddy

isd

isd
Is

1S}
ssoup1e}
1-33 “(POz1)

“uoIsua] Ul
ysuens yoeduy

“yse[q JO snjnpop
“yIsuaNs PIA
Z ul uONesu0ly
Ayqeuryo

‘yBuans ajisuay,
“y}8u9s)§ doueInpug
ATEQEPIOA
SUIIOJ a[QuIeAy

ry
[ayxouemng
[O€ OF X 901 UOISOIIOD yuLIsIso1
syed YIM YSTY Y{BUdI}S
PUP ‘ssoUpIeY
sutids satjyiedoid
0} 4,009
eee oe badges
[auo =>] UoIsoIOZ yueysIsa1
syed YIM AHoNp A070}
pue [euoIsusUTTP“AyIgeys
oAleA, ‘seas ‘ssoury duind ‘spoi ‘ssurysngq‘se[zZou aurqiny
[eu0p,SOS e ' ‘suipeyq Arpuney ‘souryoew
iaded[jiu pue flo ‘Jar “},ba poo}
| “Sul|puey‘ss9TIOq JUDTJOOXY JOULJSISAI
0} BUI[[PD
PUL UOISOJO
et ae
: jor ‘SouTsUd ‘SafISSIUL “SQOBUINY
[euoou
-X *9}9 aiayMysty “duia} y}sussjs
f ’ pur UOISOIIOD
*}SISII
J1¥ [BUA
Aoploysey
1d Le X 901 oUUeS
se [9UOdUT OSL-X

}@WIp/)
100 jor ‘souIsue “QUIGIN} ‘sayisstut
*99 osayMysry “dis} Yysuasjs
puR UOISOIIOD
“]SISO1
o1e ‘JEUIA

Aojedsem : sureS
se JOwWIPA
OOS

:qounos vied sem Wosy24)

pofidwod £26 sjouaINW ‘voj2a/ag PloyUly BurYs“OD
Igng MAN “OK
:Sa.LysON= ‘sadey
J =s ‘sdul3“Auy
10}= yususaaut‘SBurs q
Aj=eo ‘O11M ‘sirq
§ == ‘dis
y = ‘spor
J = Surqn
q =) ayeid
VL = your ‘sed ose-uo)s3unq YM
SuLpjam& sayy Teyaw O = ‘oud| sdajispure
g =A109eAx0 19ddoo Suize1q
IBLIDAV‘SONIVA
g SANTAaie 10] syetsajey)
}ew aavy usaq pajeou
purue ose Pouspsey
5 sanye
aseA 10J1 “A009
, “payeauuy
O.1OMOUS
S240,
paods
aovyingg ul tudy ym
ev ajduis jurod payuswaoApiqueJOO}
s
yy, pajreu
pue
u ‘pode
Adiey189}
o,}P Wools *dwi9}
IS4le aINdIy
st OJ Woods “dwa}pure ay} puooas
st je SNuIW “AQOTE
paseg,
uc , ISIV CILI@
= “001
asoul sanjea
oseye wool *“duta}10,4 19Y431y diva} ‘sanjeaJajor
0} WSY ‘yOoq
OApueH

oon
wyOT S3]9
1? A9, “AoOOST
Table A-14 Low-Temperature Properties of Nickel
and Nickel Alloys
Alloy and Form Temp Yield Tensile —Elongation Reduction Charpy
(F) Strength Strength — in 2in., of Area, V-Notch
at 0.2% (ksi ) vi a Impact
Offset Strength,
(ksi) (ft-lb)

Wrought nickel
Annealed bar Room 23.0 72.7 48.0 75.0 cr
100 25.3 81.4 58.0 76.0 =
Cold-drawn bar Room 97.4 103.4 16.3 66.9 =<
110 101.8 112.3 21.5 60.9 =

Berylco nickel 440

Annealed and Room 160.0 225.0 15.0 - 10 to 25
age hardened at —100 = — - = 10 to 23
950°F for 200 - - - - 10 to 22
1% br ~300 195.0 282.0 16.0 = 10 to 21
—425 204.2 303.0 16.0 — =

Monel 400
Annealed bar Room 31.3 78.7 Si. 75.0 189 to 216
-297 49.5 115.3 49.5 TE 184 to 212
Cold-drawn bar Room 93.7 103.8 19.0 71.0 181
—110 100.9 Tire 21.8 70.2 178

Monel K-500
Annealed bar, aged Room 100.0 140.0 25.0 40.0 =
—300 112.4 177.2 39.0 52.1 =
Cold-drawn bar, aged Room 120.0 160.0 20.0 35.0 ——
—300 160,2 202.0 27.0 47.3 =
Hastelloy B
Bar, heat treated Room 55.0 127.0 52.0 - =
at 2125°F —148 61.0 144.0 62.0 53.0 53
—321 84.0 173.0 58.0 46.0 53
Hastelloy C
Bar, heat treated Room $1.0 121.0 $0.0 = =
at'2225°F —148 79.0 143.0 25.0 26.0 =
—321 97.0 160.0 24.0 22.0 =
Hastelloy X
Plate, heat treated Room 47.0 104.5 46.2 = 54.1
at 2150°F —108 = 118.8 $1.0 = $0.6
—321 = 150.2 45.5 = 36.8
Waspaloy
Bar Room 122.0 195.0 28.0 28.0 14
—104 132.0 203.0 24.0 26.0 _
320 157.0 235.0 17.0 18.0 =
—423 163.0 243.0 18.0 14.0 —
René 41
Bar, solution Room 140.0 192.0 26.0 33.0 =
treated at 1975°F —100 142.0 200.0 29.5 34.0 =
—200 150.0 216.0 29.0 30.0 =
—300 160.0 235.0 28.5 27.0 =
—400 172.0 252.0 27.0 26.0 -
Sheet, 0.062-in. thick Room 159.0 204.0 16.0 - _
—100 174.0 227.0 16.0 =_ 7
—200 180.0 238.0 16.0 — _
—300 189.0 252.0 16.0 - —
—400 197.0 265.0 16.0 — -

Inconel 600
Annealed bar Room 36.8 93.8 37.3 64.1 236
—110 42.4 106.5 39.8 64.0 206
—310 - = = = 187
Cold-drawn bar Room 147.7 152.1 7.0 49.3 69
—110 154.9 163.9 9.8 $1.2 =
-315 = 182.0 10.0 49.5 =
Inconel X-750
Hot-rolled bar, aged Room 135.0 191.0 26.0 46.0 40
at 1300°F for 2 hr -110 140.0 202.0 30.5 48.0 38
-200 145.0 212.0 32.0 48.0 37
-300 149.0 225.0 33.0 47.0 35
—400 152.0 240.0 34.0 43.0 —
Annealed and Room 110.0 170.0 25.0 28.0 33
age hardened —=110 115.0 184.0 22.0 25.0 36
-200 118.0 192.0 20.5 23.0 37
—300 119.0 200.0 18.0 20.0 Sy
—~400 130.0 210.0 16.0 16.0 =
SOURCE: Reproduced from the /970 Merals Reference Issue, Machine Design by permission of the Penton Publishing Co,,
Clevelund, Ohio.

884
 885
87
se!
Sc
00

PEGS
Cae ee
(oa
‘saoejing “SIBUITea, ‘sapInydAjBA ‘ssuuesg‘saplg *3°9)
suoneoyddy uoyouy 10J MOT
speuaze
PEPE
es
EC
ee
ee

SSP[D-19}Ss9AOg
eee
sse[9-Axodg ae

59
Aoyye
DAd-SOADY
Peurpour
AoW ‘sou
soje1Ayng 48
-J20VY
aso]NI]J29

soue|AdoidAjog
Ayisuo p-ysty
saua|AyoidAjog

jorduir-yary
*sauatAyS
sjonq
‘s1autejUOD ‘sdulsnoy ‘spnosys 10j
syeliajeyy
w@ g e > i= is} a) ies} wa 0m 1 |) Et ~ ais 3
2 ol & 2 SNe & || ry Z\|2s4 off ea 2/8 > apes | <= <
= aj/e| >| 23 e|i 2] 2] 5 a/528| 8 o\e8 [n52| oF
28
s|alll SB
xSz/-sE|Se2|
=se(a22|2BIS +2 42
s2)-3
gs
8
g
eS Calas@ a7 z/
Ga =&|]
Bb
8
Se | SSl/SSsle5
a ene =
SerlRsei72
Flx 22
eel
2i*Xoe
al a &
Stale etsaeFy mo aweN
= = <S = EB) Spe og o- @ s see 2eeols & $$
be A ei aed * os s
uonsi4 ce =a wes [Ss 5S 3 ey aE aa 2° 2 2 duauey
fe aourisisoy
0} jo 3 | Silas
e & 8 ge “ |S 5Ff S iS
1UaI914J909 = vonsa1}9q gi
soise|q jedIdA| awOS jo Salueddsg §=GL-Y BIGeL
 *O.) Muyysyqnd UOIUDY OUL Sands AoUa4afay SADUOISO]/SINSPd [£61 24) WOI) prided ‘AOANOS

auajAuaydAjog
ssej9-Axodq

apixO
souaAdoidAjog
Oo
ee

|i
QUON

ddd ¥ dAL
qq

ea
PABESSAae
judWdinby jRUaYy] pur
peorWoY. Joy
speudieW
paytty qe)
S|

*salpoudyd

so}Ruo
-qird
hog
partly 40qu)
“ddl 12399V
speiaoy
SUO|AN
Ayisuap-yaiy
‘souajA yah [Og
“pany 49qy
“AAL 1229V
suljeoigny
‘[raoy -JJ2S

mm
|
sjonq
“siauteyuo> ‘spnorys
Sea less ee¢|==Bastees] a} eo = so) Sa=> Ge i]
= =|2¢ = z SS $i|
es nese
oles=

su aAjOS
| .f|€2s 982 a

SO
1ulog apg
es| < 73 e =
= Ss|/fSa5 Ca = & 99x S RP] -—- ZS aIS
|isd poz|isd ve Ze
ra a = =

886
a So | Sa ehtS Oe ae) baaed —Salee =|2 elas =
ts
; =
&| Ss
= Se SeZle cml az|%a3 awe
Ss SS|2Se172 8

aourisisay

yond
<El ey) | Sie) Fea 6% Ge Sosa e a)|S: ae Ss
= Saag ||ess 55 Sua olen | ad eal each ay Quauay

<
| Pee # S ES e =

do
jo JAOJ Sa : eed acl fe

0)
1UAIDI = &: uonrayed
(panuijuoDd) a1qeL GL-v
Table A-16 Miscellaneous Mechanical and Physical Properties of Various
Materials"

Modulus Shear Coefficient Thermal

of
Elasticity Modulus Poisson? Densit of ae 4 2
of i 2 's a TKarnal Conductivity Specific Melting
in Tension, Elasticity, Ratio Ib/in. Expansion, Btu/hr/ Heat Point#
psi psi in./in./°F ft? /°F/ft. | Btu/lb/°F

Gray See 5.5 X 106 C1.20

cast iron Table A-1 18.0X 106 c1.50| %27 | 0.260
Malleable
cast iron
See 12.5 X 106 0.27 | 0.264
Table A-1

Nodular or
ductile iron
8 A-l
Table 9.3
; X 106 0.29
: | 0.251

Cast alloy See 6

steels aie 11.3 X 10 0.33 | 0.283

See
Cast carbon
ue 11.3 X 10 is 0.33¢ 0.283
steels

Alloy cast See 11.3 X 106° 0.33° 0.266

irons Table A-1

Cast stainless See

steels
ea 11.5 X 10 6° 0.26 c seg)

Plain carbon
steels
30 X 106 11.5X 10° | 0.27-0.30 O28:

Alloy steels 30 x 106 11.5 X 106 | 0.27-0.30 0.280

Wrought
stainless steels
28.5 X 106 11.5 X 106 0.26 So i Oo=)

Wrought
aluminum alloys
10.3 X 106 3.8 X 106 0.36 | 0.097 13 X 10 -6

Cast
aluminum alloys
10.3 X 106 3.8 X 106 0.36 | 0.097 12.6X 10°

Wrought copper
alloys
15.5 X 106 5.8 X 106 0.33 | 0.305 | 11.0x 10-6 bes)

Cast copper
base alloys TableareA-10 5.4 X 106 0.33
ES | 0.303
: | 10.1
x 10-6

Nickel base
SB — 0.34 | 0304] 7.6x 10-6
alloys Table A-13

Perouen Cad ANH TG5

magnesium alloys
108 2.4 X 108 0.35 | 0.065 | 14.4x 10-6
SOURCE: Data was compiled from the /973 Materials Selector, Reinhold Publishing Co., New York, and Properties of Some
Metals and Alloys, International Nickel Co., Inc,
NOTE: The values listed aye average values except as indicated.
®These are max, values
bWhere two values are given, they are respectively the low and high values and are too wide for
a meaningful average. The reader should refer to the ASM Handbook, Vol | for detailed
and specific data.
These are approximate values,
4Refer to ASM Handbook, Vol. | for values of particular materials,
e f g
Ferritic Stainless Steel 5.6 X 1076 12.1 — 15.1 2790
Martensitic Stainless Steel 6.0 x 1076 11.7 — 21.2 2800
Austenitic Stainless Steel 9.4 x 107 8.0- 9.4 2650
Kat room temperature

887
 Table A-17 Standard Gauge Decimals (gauge and equivalent in decimals of an inch)

Aluminum Steel Strip & Steel Wire Aluminum Steel Strip & Steel Wire
& Brass Sheets Tubing ae & Brass Sheets Tubing
Ga. Brown & Mfrs, Birmingham Steel Music ; Brown & Mfrs. Birmingham Steel Music
No. Sharp Std.* or Stubs Wire Ga.* Wire F Sharp Std.” or Stubs Wire Ga.i Wire

SOURCE: Joseph T, Ryerson & Son, Inc,

“Replaces U.S, Standard (Revised) Gauge.
tReplaces Washburn and Moen Gauge,

888
 400

350

e)oOoO

ASTM -A
ksi
strength,
tensile
Ultimate Tempered valv
e Spring

150

100
0.010 0.015 0.020 0.030 0,040 0.050 0.100 0.150 0.200 0.300 0.400 0.500
Wire diameter, d, in.

Figure A-1 Minimum tensile strength of spring wires. (Courtesy of Associated Spring
Corp.)

889
ksi
correction,
Final
including
curvature
stress,

0 5 10 15 20 25 30 35 40 45 50 55 60
Initial stress, due to first load, corrected for curvature, ksi
For commercial spring materials with wire dias. up to 1/4 in. except as noted
Stress ranges may be increased approximately 30% for properly heated,
preset, shotpeened springs.
Materials preceded by an * are not ordinarily recommended for long
continued service under severe operating conditions

Figure A-2 Endurance-limit curves for compression springs. These

curves cover compression springs that must withstand unlimited fatigue
life or at least 10,000,000 deflections. These are based on range of stress
caused by the first or initial load and the stress caused by the final load.
(Courtesy of The American Society of Tool Engineers.)

890
 Rae
S

Stress
Concentration
Factors

891
 UNIFIED AND AMERICAN THREAD DESIGN

Lulz
Sanam CEE
eesOR TENSION
CUT

QUENCHED 38

DRAWN

KEYWAYS

PROFILE SLED-RUNNER

| PROFILE —_—i SLED-RUNNER

tg iceand ee TORSION

[ANNEALED]

Ky QUENCHED

DRAWN

S,, for bending or axial loading

veeS,, for torsional loading (tentative) Steel

Notch radius r, in,

 Figure B-1 Opposite, top and middle: Fatigue stress concentration factors: threads and
keyways. Note that these values are for K, and not K,. (By permission of Charles Lipson and
Robert C. Juvinall, Handbook of Stress and Strength, The Macmillan Co., New York, 1963.)

Figure B-2 Opposite, bottom: Notch-sensitivity curves for use with theoretical factors K,.
(Figures B-2 to B-10 inclusive are reprinted by permission of Robert C. Juvinall; Engineering
Considerations of Stress, Strain and Strength, McGraw-Hill Book Co., 1967. These reprinted
charts (with the exception of B-2 and B-3) are based on those found in Stress Concentration
Design Factors by R. E. Peterson, John Wiley and Sons, Inc, 1953. Fig. B-2 is based on
material found on p. 298 of Metal Fatigue by George Sines and J. L. Waisman (editors),
McGraw-Hill Book Co., 1959. The source material for Fig. B-3 can be found on p. 111 of
the same reference as that cited for Fig. B-1.)

Figure B-3 Below: Reduction of endurance strength due to surface finish—steel parts.

Hardness, Bhn

120 160 200 240 280 320 360 400 440 480 520

Mirror-polished ity

; Fine-ground or
commercially polished a

Surface
factor
Cre

ea
A
AEBS

° 60 80 100 120 140 160 180 200 220

Evil
240 260
Tensile strength S,,, ksi

893
 Figure B-4 Shaft with fillet (a)
bending; (b) axial load; (c) torsion.

894
Figure B-5 Grooved shaft (a) bending;
(b) axial load; (c) torsion.

895
 \ M M
hol J p p 4
L\ —~ te 1+)(}
ES La

eb Co alPe :
SS
Axial load:

2.2 4

\ a
Bending (in this plane):
_ Me. mM
Pnom ~ ~ 1D3/39 — dD2/g
A i = alee lense

rT ae il
Torsion:
ee eee fe jes ie
nom!) J 7b2/46—adD2/6

1.2
|

Figure B-6 Shaft with radial hole.

896
Figure B-7 Bar with shoulder fillet (a) bending; (b) axial
load.

897
 _ Mc _ 6M
Pnom is bh2

0 0.05 0.10 0.15 0.20 0.25 0.30

0 0.05 0.10 0.15 0.20 0.25 0.30

Figure B-8 Notched flat bar (a) bending; (b) tension.

898
 sail LU UE TS Aa

(b)

Figure B-9 Plate with central hole (a) bending; (b) axial
load.

899
 r/w = 0.050
r/w = 0.10
r/w = 0.20

ALSO ARGe ez Si lO 2,OF 22S 22S) OKA 215 26 ee oO) One O

W/w

Figure B-10 T-head, axial load.

900
 ach [ V1 — (r/c)? / |
Nominal
inal stress = GB)
t= (cle) (1 —V1 — (r/c)2 )

(Based on net section A-B)

0 0.1 0.2 0.3 0.4 0.5 0.6

Ratio r/c

Figure B-11 Tensioning of a plate with an eccentric hole.

901
 Tables of
Fits and Tolerances

903
 e1qeL4-9= Buluuny
pue Buipiig
su4

904
Ss81DOWT $8810OU7 Sse[DOU€ Ssk81DOUb

[BUIWION piepuels piepueis piepurys piepuris

aZIS ‘asuey Q 8 SW] Q 3 S}TWIT 2 8 S}TWITT 2g 8 SHWIT
‘Ul SB Ss BoB 2s
. OF| aon | weus| PE| aon | weys| & g 5E 10H | yeys
ee) Ba SH #3 9H 63 8H uy
10 cOt+ To— 10 $cOo+
| TO— £0 yOt+ = Si £0 90+ 0
O - ZI0 sv0 0 STO— ss0 0 £0-— S60 0 sso— ‘S|! 0 LO

0) a TO+ sto— STO sar sto— v0 SOo+ yO— LO+ y0-—

670
-Z1'0
$90
sco
ram
£10—
$0
0(EV
60—
70 Setar
|] — (EW (al) vot C0— $0 90+ SO= s0 60+
| SO—
-bz'0 Or'0 90 0 Sei0— SSO 0 S¢v0—| a 0 60-— Ore 0 je

: < SLO 0 Sv'0— $60 0 sso— al | 0 S|

SiO
80
r'0+
(0
£0
SO+
EO=
eal
ats
-IL°0 6I'I $60 0 sso— (GI 0 L0-— Dee 0 ot

vO—
90+
yo—
OT
v0
vot
On
Sh
iae
=611 L611 VI 0 L0— vl 0 8'0— OG 0 OC

v0 SO+ y0— v0 L0+ 0—Fr (G\\ Stt+ Ck

-L6"I
STE
L0-—
(GAl
60—
oT
=
Ve
0

eveWe ST 0 60— O07 0 bi ftks jf 0 ie

cy"“) 81 0 EC 0 [t=(ae (a4 0 (C=

fase 07 0 (A DY | 0 yI— o's 0 Sioa

ae Bey fic 0 rI-]|

| 6~ 0 [EO JESS 0 S¢—
 e1qey4-9 (panujuos)
Sse[D.OWS Sse[DOa9 ssejOOaL SselDOU8 SseOU6

jeurwioyy piepuris piepuris psepueys piepurig piepuris

azig ‘ofuey| 8S Suu] se sur] 2 2 SEE 8, 8 Suet 82, ST
i 532 abe 8 B32 BF
Fe (04 | yeys E: 5g 5g CE| PH | wes
J2AQ oL 8H La ie ae 4 11H

90 90+ 90 | o1+| 90—

| oF —|sz7+]|
or
Ps ees fe Deaf exere | anes \| el 8 leg6:5
ee 07 06 0 —| o9
SAG MAD Ol Oe Feeal + se —| os
ees Siz 0 0 —| TL

; 4 67 i 0 = || ia

eee of O'Or 0 SST 0 soI—]

* eee oF | S11 0 o'sI 0 ozI—|

ee ss S'€l $07 0 s€I—|

Ely
-STE steer KS OL ea + 06 | O'OI—
99 SSI O'bZ oO osI—|

3 Skt OL o'sI 0 0°87 o osi—|

ee On | [borcx | om
| fosGel
hase 98 SIZ 0 O'vE 0 o7z—|
Sunes Goma | [oeroe | ot| jottaLN
aa ool 0'Sz 1) O'8E oO o9z—]|
SHwTT
aIe UT SyypURsNOY)
Jo ue “You! syn]
10J soy pue yjeysare parjdde ATRorerqadye
0} AY} dISeq AZIS07 UTeIQO
aU) S}IUNT

905
JO ayy OZIs 10} ay} “s}Ied eeUT Pjog doRJ
St UT souEpsIOD"
YIM
V De UI99IS2
“SUS SIUL
eVep
sI OU ATo}0;duIOD
UMOYS ‘atay ay a}0/duI09
9192} sa030} B FeUIUIOU
azIs a8ueI
JO 0OZ “U; SJOQUIAS
‘SH ‘CB ‘o}A aK JOH PUe JyeYS suOIeUSISOp
pasn
Ul OW “WoysAS
 LL OIqeZ-2
«= adUeIea|Q J2euoNed07
.SH4

906
sse[9OTT SSF COT SSPE)OT€ SSEIDOT¥ OTSSFID
S
eUTUON, piepuris prepuris piepuris piepuris piepue}s
azisg “Buy Sita sur] Sita Sy] SPW

91°H

JO Sy]

JO SHUNT
JO Sy]

QOURIBATD

2ourles[D
QOURIEIID
weys

20uRlea[Q,
aouesea[Q
210H yeus
LH

weus
64
9u LH 98

2)
+
3

r)

SE

Ss

°
=
pot

1
MS
on
219HOTH al
IF

Lo

So
i)
oi
om
of
oo

w
Yo

S
w

ons
one
one
=

S
eel

N
coo

+ |
con |
oda
w

+ |

So
+ |
|

So
La)

+ |
|
ce4

|e eo COP aed 0 0+ sTOo Ssh|| SUD

~~
Se so o-| zo—-|| so o—|. S60

eri
oS Sr'0—

Te

ao
oe
Re

—
owe
=

nm}
ie)

=> ||
ioe)
ar ||
(av) = OS
| GU

+
S
N

Ss
ro)

S
2

it
s|Se|Sco

ard

Se
(GA!

iS

on
DiS 3:0

No

of’
of

iS

_
osc

w
NS

BES
=

So
©
om:
=

re)

|
oP qlee Wl [Pas 7
a:

ae ||
foe)

ae ||
=P ||
=P || ae ll

ar ||

LAO)
aa
2
: i ze 1) O=— || ee O= | wO= ei | = © || icv eee |p Se
1s0+ Ol 0 80+ 0+] 0 leat Oe 0 Ges|| =
=1L0 rt On 7:0) EI |S Sie % = @ || S0= Ss |)= oO || Ge=
904 Om moner One 0 Gye|) a 0 ort| 0+
=

~6U'I 6 O= wO=|| = @ || 0 Ge || O= || i= SS HS Se
on
tI
ou

tS
Seo
b- d—)}

+ |

0 LO+ ot Calc 0+ 0 8t+ + oO 0 Spt OF

ere, @halpe Daal |) 50a O24]
|| f= € @= || Git is Oren| =
oO 60+] 0+] 0 rIt 0+] 0 era:|] Oa 0 Oe || )
eeece Oe 0S49 Ca) = 0 |) B0= se 1 o | Pi Ss = || GS ; OR rit
WAL0 ere ees 0 91+ 0+] © | sex|| ®& er09+| 0+ 90 barSal | SAD
ae =|] © || wo || re = 0 | oT ie |) o= || on G= =| Gs oS S|
0 Nba(6 On-stie 0 Sit 0+ 0 87+ + oO 0 OL+ Oo 90 tts|] CY
Iya,Oe Oven|" 180 || OE = Ci=|| eS wa eae ae i Se KS OF sI—

Ses sei ie =|! O Go=|||| ae = O |} i= S = G¢@=||o

| ai @= | s=
 907
*uaysAs Og Ur pasn suoneusisap 1yeYs pur soy ore ‘919 ‘Sy ‘OH S|OQUIAS “UI YO Jo edUeL azIs [eUIWOU e Oo} S208 9[qQe} 9}9[dUIOD ay] “2194 UMOYs Aja}a|dwoo Jou st elep SIU, *s}UsUIseIZe Ogy
YIM doUEpIODOR UT d9RJ Pfog UT BIE “syed ay} Joy OZIS dy) Jo SITU] BYPULEIgO 0} OZIS oISeq ay} 0} Ayjeoresqosye pordde ose IeYys pue IoY Joy SIU] “You! Ue Jo SYyPURSNOY) UT ae syTUTT
09 OC | Oe TE (MW Or 07 SOo= 0 SLI CoS 0 TOL IP'Z1-S8°6
ap |
ae |

1 |
co 0

ee
|

oo
xilco
m1lat
oo

al aaa 8+ L Sita O8+ SY (KG os+ CC

ABIAR

0c ol (lar
lop)

nN

Ol 1782 o= | és o—| sei | ss— | 0 ssi | gp— eon ear,

|
ao
Pe

87+ (a6 Si 8I+ 91 it fe On OL+ Or o7— S¥+ 07

2

(= OR 89 = O= br (Al |) OQ) = 146 Oe 91 Si 0 S El a 0 60°L -€L'b

st St+ 81 Cla OT+ él Sa OI+ 8 Oar 9 Se— 09+ SE Ori Or+
So Or SG (We 6£ Gli= || O = We S8— 0 UL ELib -STE
ol (Gar Li bit IT {f= 6 + je os— S€+ vl
GE= = os Co O° — VE 0) WO) = SLI Om 0 09 STS -L61
A= 8I+ tl (0) CL ol = [5 ar 9 Or— oe+ (GNI
sc— |o-— | eo vole ia anes 0 re freee ore
ee eci@ vai 3 —| or+ 8 Se A Ge ie S74 | orl
(= O.= VE Si Oo = €T = oer Sr— 0 Or 6'l -IL‘0
(0) cls Ol [i Simei £ Kear SY SC O0T+ 80
¢23] OF — 87 (S| OF 0@ = £Ol Sis 0 (ONS, : :
9I— oO =— ST Ot LI — L8 OF — 0 87 eal Ope
ej a hone o— | sr —| 92 | ¥z- 0 fT ane
Ly O) = Gls Oss 99 = 0 61 ZI0 - 0
| 0 =
Ot s b+ 14 SG ar SG oT— OL+ £0
oi =
€1H e 71H ore 11H 6P OH 3 6H SL, 240
yeus 20H g 3101 E5 yeyus 2]0H g5 210 25 g5 yeys 30H g5
s Bo Ba Bo Bo Sie oP
Se 8 SHUT es, sur] 8 @ SUT a a 8g Seu 82 | ‘euey azis

(panunuods)
piepuris prepuris prepurys prepueis PrepUueys psepuris yeutuon
TT OT SSF OT OT Sse] 6 OTSSEID 8 OT Sse LOTSSPIOD 9 OT SSID,

e1qeL
Z-9
 aIqeyE-9 uolisues, |2uoNeo07
,SH4
sse1oLTT SED CLT s8se[DLT€ SSPIDLTV SseLT§

9
LT
Sse[D
TON, piepurys piepuris piepurys piepuels piepueis
azig ‘adury sywirq sywirq
STug| sytury Sy]

908
que Wd Nel Wd Wd Wd

ud
piepurys
3]°H 1yeUs 310 1yeUs 20H yeus 310H Weys 30H WeUs
0) OL 1H ost 8H ist LH or 8H Ly LH gu

SYIWT 319H
yeus
Lu

LH
oro—|°
| r0+| orot+ 90+| ZO+ so-| ro+ sot+]
eae 0507 0 oro 0 z0
SOO
stot|— 0 | szo+
==
vO+
Ol

SO
STOt
: sro—
| so+| svo+ LO+ stvo+| 90-| so+| 90+
STO+

(c0)

3
oy I S807 Cee EO — 0 | svo— wot| — 0 | eo+ TO+ SO+
80+
Of

e Zo-| 90+ 70+] €O—]| 6O+| €O+ sot| co—| wo+| Lo+ so—|]| 90+| sot
€0O+

°
|)
Gr

|
ee
| Rs.07 Olean|b Pid Giles
Oi

0 co-| | so+ ror| so+| o—| ro+ zo+]| o— rot}

Sse

TOT
co—| co+ cwo+] seo—]
| o1+ seo+|
| so—| cot| So+| SO—| O1+| g0+ 6O—]|| £O+| 60+
nN

i
(sae,
woot Qe 0. lseias ou FOr seca Po0s S|
Oe Or I |Fo0= Chen eLOes On OFC
=
O
OF LO+
rO+ (G\bae

Gort
szo—]°'*
| so+| szo+ vo—|| ci+ vo+]| vo—| 80+| 90+| 6o—|. ci+| cor| ri—| sot Ti]
+
|

nn|e scone |e ecco Rone PO | iro ort ai IFO Or! arin PO RO LOOe OPO
80+
SO+ tale

TOT
c0—|) ont| cot Wrso—| ote| so+| co=| ort| gor ri| [ore [iri l=| [Ore ele
=
~

a Oates|) c0—| ier lore (PSO oor Or Or ele FOSS| Oe (SOs Oe BOP
Oe tse
90+ (Peed)

SOs:
soe| ci cor] 90—]| sit 90+]| [eo 71— -|. soe| Pper= [sie ert ST]| CUS Sie
o
ai
|
||

mes pew cue tee coma aac Lok P90 teh Ot1 Cey OT 8 sO phon OFS ies Oe ee)
i= ise
LO+ ior

F0+
ro—| ri+ vo+] co—]|| c7z+| cot+| o1—| I+| ort| SI—| e+| Sit| GI—| vit| 6t+
~

ee
||

ea: ai) Lo yo eee) Port| co eres Ome|) boriy hic Oe erOt OS Py CTOs
80+ tea

One alae

yO+
so—| 91+ so+] so—|| sz+ got]| Vi— re| t= sce| Leis C=| (Ot Cet
a
Or+ i

oe)
|
OF Ot

eee Pay ed 0 SO | e+ 0 <0 |) oasis| ror yz4| Ors| torei ror OI | CTS
Catan

bast

yor
oO

90—-| sI+ 90+] Go—]|

| Sz+ Go+]| HI-| BI+| HI+| OZ—| SZ+| OT+| 9Z—| SI+| 97+
(GMa

Q
oe)
a

ee |e ceIe PONS 90|, her Soret |e lene Oe Orite Ic [cr Oc,I NCOn evo Cmee ET
Cota

Salet

vO+
90-| oz7+ o1—| oct.| ort| vI—| |oz+ Hit| Te—| OS| V+| 9e—| OCH| VE
+
oe)

a phe eel en ov+| — 0 o1—]| s+| — 0 | co+| sz+ | — o | zot+ 90+]| o— | vI+
vit Ain

es OT+

LO-| T7+ [Or estes| Ol = 19 eee TPO eaeiy CSR Cr LOS Cor| Oe
ioe)
ioe)

| 907 |

ter taal St | oO OI—]| OZ+| O- | TO+| ee+| O=| 7O+| 90+| O— | Fit
vit Se+

(Oe Gai

90+
s0—| sz+| sot TI-—|| oF+ I+]| SI—| szt+| SIt+| — [LT OFF| Le+| ve=| Str| vet
mM

eee |e Reed so-so

| Perr |For C=| vert Oye eee = CORO OF OS SEEN
Si Sir:

i Sivéae

LO+
Of
Si+

SyUTT
ore UT sy}pURSNoY)
JO UE “You! s}TWIT]10J efoy pue yyeYsaie parjdde Ajpeoresrqadye
0} BY) JIseq OZIS0} ULEIGO
DY} S}LWT]
JO DY} 9ZIS JO] oY) ‘sy1ed e}equl Plog adeUT adUepIOSOR
YIM Dy *s]UDWIODISE
SIYL SERIEP
JOU A[A10[dWIOD
UMOYS “a19Yy ay 2}2|dWIOD
91qQe} S9030} B [eUIWOU
ozZIs duesJO QOT “Ul ,j14,, sjUasosdal
94) WINLUTXELL dOUAJOJIA)UT
SnuIUL) ‘(sanjeA
puk ay) WNWIXeW sourse9]9
snjd) *(sanfea sJoqUIAS
‘7H‘sf "099 are

Ul
pue
OGY
pasn
sJoH
1JeYS
*wiajshs
suoljeusIsap
Table C-4 Interference Locational Fits*
ee Ces IN
Ces LN erie Class LN 2 Class LN 3

Nominal 3 Standard Standard 8 Standard

Size Range as Limits Limits S Limits
in. zs
ice e3S
5 2 Hole Shaft Hole Shaft 2 pee ou
Over To ta H6 ns H7 p6 =

0 oO 0 +0.25 +0.4 | +0.65 ese +0.75

. 0.45 = 0.65 = +0.4 Me +0.5
om af me ke +0.5 | +0.8 +0.9

+0.4 | +0.65 me +1.0 +12

+08
+ 0.8 ape alia fae +1.4
pes

0.71- 1.19 +0.5 | +1.0 +08 |+13 | o4 | +08 | +17

i : =O ||Os =) || $08 1.7 =@ || +417
1.19- 1.97 +0.6 | +1.1 +10} +16 | 0.4 +1.0 | +2.0
: j — 0 | +06 = @ |] Pie 2.0 = | 1A

ee eo 1.3 a eromaleris 029" i 2eeoi Roam eis 23

: : = (i) +0.7 2.1 = 0. || +E.4 2.3 —o | +16
ae ee +0.9 [+16 | 0.2 +14 [+24 | 06 +1.4 | +2.9
eee pO |tolOm| 25 |e Oe titled 29 =e |oar
0.2 | +10 |+19 | 02 | +16] +28 | 09 +3.5
ees eZ ne5

+1.2 | +2.2 +18 | +3.2 | 12 +42

(One HD 22 = 0 otI 4 mo

0.2 | +12 | +23 126 +34 | 15 | +2.0 | +47

*Limits are in thousandths of an inch, Limits for hole and shaft are applied algebraically to the basic size to obtain the limits of the size
for the parts, Data in bold face in accordance with ABC agreements, This data is not completely shown here. The complete table goes
to a nominal size range of 200 in. Symbols H7, p6, etc. are Hole and Shaft designations used in ABC system.

909
 e1qeLg-9 e104pue IUUYS,SHU4

910
ssoNaI Sse[ONaic Sse[DNe€ sse[DNAb Sse]Na§

jeurwioN, oo piepuris 8 pieputys Sane piepur}s Ce piepuris ie ay piepurig

azig ‘Bury $6 sur] BS SMW] qe SHU] Sg sur] 6 2 sywry
“un Pe oe ee ee 22
aE 20H | yeus @& S a el0H | yeys af aon | ueus ac joy | yeys
7240 OL = 911 = 9s = LH 9) 2 = 8H Lx
* | soo| szo+
| Sot| zo rot| Ss0t+ £0 4 el
0 — z10 0) = €O+ S80 = Q 90+ a! + 60
10 £0+ 0+9 c0 $O+ O'l+ : s0 Os We ial
—Z10 670 90 = bO+ Ol = LO0+ all i} (Or== ar Cll

citediy Slo = OQ 50+ vil oT O1+ 07 (i is Lal

ren ye 80 = ) ¢°O+ OM! = le (62) 81 7 SiG Oe ar CEO

70 vOr 60+ $0 LO+ oI+ L0 OOF +) St 80 Ol+ ae Sie
SEW
= WG) 60 Oy Ose 91 OF Glee 81 = oe Ae Sie = Silt
TO $O+ ee Us) 80+ 61+ 80 8107; lis@aetan|(e Ol G st Oe
We= C50) Eig = L0+ (= vit Ve Wi = 91 OE 0 CiGas
£0 S0+ (GAN 80+ 61+ 80 80+ Guar Ol 58075 Ses Si Glct SSG
=SEO 6 (|! = 80+ (== Pala Lig = 0) et 9 tie — GC Saat Ge OF; sie SG
ew) 90+ Sales Ol O'1+ 2 Oat Saat O1+ an ES vil SAG + Orv
ola RC ica = 60+ iG OFS OK Ve OF Sra Or OS Oe
v0 | 90+| vit act lett or Sc LS | VEAP,
vz | o1+ +] os
aera
Th Ua OR) Gn) fare [poe al encca Ueland)ole |S0-=-|-0's Opt
90 On Salar 80 (Nbr Lia a Sst Call GES £7C Gallats Cayman NE i+ COE
=/K5\ OH 81 = (ilar Lie (= 07+ (Cis On “aR SG (617 OF Sr SS (ES) (= + os
 L0 L'0+ 61+ OT (Or 6+
—96'7 STE Sl Ott
| Lest 87 tOVGE||Se EAP
ol On pit 67 = (4 i+8 Gilat
Ct INS = Sieve LY = O + Or G1E WS + O99
60 60+ pct rT Pit et V7Z Pt| + VY OMS
—SI'E POE ve Oe i+ 8 LE VAter Spl] OOS 4 i Gor + rs
= O T+8 vr Y= Se 6's Of + os v8 = +
SL! OL
60+ Nore Crit Palate OE+
aE Se SHE It | + 6v Ot Tee al]
97 = O'~+ 6€ Ok OO) 8s (Gre + v6
O'E+ ov (Qe + Ov 69 = + 09 +6 Y= + 08
Gil OI+ 6+ 61 oa + S VE CRG
| ot WEY vs
—E€L'b ZS'S 67 OF Ger. SY Onl| OSteG Sik SGP ME EP
= Q Sigs: 09 — () OS 08 Oe Omer Say Oo]
Cal OOI+
O1+ Ge VT Gr Sy OS+
—zsS"s 0:9 VE SPlice
|| ap OR) vs 91+ +]
(AS = Q SiGe os = O08 $6 ST+ Sets
O'r+ 09 O= Se ONS 08 = cr Oy oET = OQ OCI+
sl O1+ S°E+ 67 Siice SiS bY Chis|| sr OvG v9
—0€'9 60°L Sie — 0 87+ SS 9I+ +] 06 $6 SiC lise CMs
Of Spt OL O)= 0195: 06 Qi= + O08 Sie OQ= OCI+
81 (O\Gs Set (ENS Sit (GEhr cs Silct Stan CS CL
—60'L 88'L SE = @) Oe+ c9 I+ COI+] (CONN T+8 SSi+
Oe OS+ T8 Ot Oi COTS OS + 06 SSI Om
Sie OrI+
Git eb+ (AS Saat (GLEE
—ge'1 93°8 OS ee PAS
| sue nes 8 Sit|
ev ) Cet Ge) Oo CUlt GEL C+8 LI+
8
OS+ 78 Oo ae Ove.| GUE = OQ O'OI+ SLI OF O9I+
ie (EV y+ € (a7 i+S
—98°8 S86 (line CS) Sit
| + 76| COl
ev OF ne (Efe = ) eeSalct Gola CEI 87+ SLi+
09+ c6 = O) Oise GOD = OQ OTI+ SLI Q= O9I+
87 Woe (G\ b+6 Or OC+
—$8°6 €0'll Git OL OT+ TOI+
| OOL
6? = O Ob+ (Gul, = ) OT+| CEl+ OST oet+ 007+
09+ col O= + 06| CEL O= OTI+ 0:07 = OQ O8I+
87 Ti+ 60+ os Oi C8+ OL o7+| TOI+
| OCI
Ip'T1—€60
0'11 = Ov+ 78 O7%+ CSI+}] OLI Oe+ O'7T+
On OL+ TOl QO= + 06| TST O= OrI+ 0% oF 0'0t+
S}wry]
are UT syipuesnoY
jo Ue ) “Your syrwT]
404 BOY PUP VeYSore
eyep parjdde Aeoresqady
0} oY e dIseq dzISO} UTE}GO
ou
st Jou Aj2}2]dwWod
UMOYs “aay ay] a1a;du09
21qe1 sao80] © spun
Jo ay) ozIs 104 oY) “syied EVEUL Plog 29k}UT
leuruiou
azis aBuEI
Jo QQZ ‘ul SJOQUIAS
“LH “9s “919 Av aouepsoo—
YIMy8 Dg “S}uauda1
SIUL 3e
PIOH pue 1yeYS suoIeUsIsa
pasnUL p Jey “WaIsAS

911
ve

_—

|
ie

i]
Standard Welding Symbols
and International
Conversion Units

913
Table D-1 Standard welding symbols. (Courtesy of the American Welding Society.

Basic Weld Symbols and Their Location Significance ~

PLUG OR SPOT OR
FLASH OR
{ UPSET \ GROOVE
LOCATION
SIGNIFICANCE
FILLET
SLOT PROJECTION| © EAM
SQUARE v BEVEL U J

ARROW SIDE

OTHER SIDE

REE
NOT
BOTH SIDES USED

NO ARROW SIDE NOT USED

OR NoT NOT EXCEPT FOR NOT NOT NOT NOT
OTHER SIDE USED USED FLASH OR USED USED USED USED
SIGNIFICANCE UPSET WELDS

Supplementary Symbols Typical Welding Symbols

ALL FIELD weit tHRU CONTOUR BACK OR BACKING WELD SYMBOL
AROUND ees FLUSH CONVEX | CONCAVE |
ta Y ANY APPLICABLE SINGLE |
2» Eee Sao ae GROOVE WELO SYMBOL
|

Location of Elements of a Welding Symbol SURFACING WELD SYMBOL INDICATING BUILT-UP SURFACE
ORIENTATION, LOCATION

SIZE (HEIGHT OF he OI RE
OTHER eae
THAN SIZE ARE
OMISSION INDICATES NO SHOWN ON THE DRAWING
GROOVE ANGLE, INCLUDED SPECIFIC HEIGHT DESIRED
FINISH SYMBOL ANGLE OF COUNTERSINK
FOR PLUG WELDS =
CONTOUR SYMBOL
LENGTH OF WELD DOUBLE FILLET WELDING SYMBOL
ROOT OPENING; DEPTH OF FILLING
FOR PLUG AND SLOT WELOS PITCH (CENTER-TO-CENTER SIZE (LENGTH OF LEG) SES LENGTH.
SPACING) OF WELDS iT: OMISSION INDICATES
SIZE; SIZE OR STRENGTH: z 12 THAT WELD EXTENOS
FOR CERTAIN WELDS ARROW CONNECTING REF- SPECIFICATION, PROCESS |_1G) BETWEEN ABRUPT
ERENCE LINE TO ARROW SIDE OR OTHER REFERENCE CHANGES IN DIRECTION |
REFERENCE LINE OR ARROW-SIDE MEMBER OF ORAS DIMENSIONED
JOINT
CHAIN INTERMITTENT FILLET WELDING SYMBOL
y—-—— LENGTH OF INCREMENTS
n =~ 2-6
PITCH (DISTANCE BETWEEN
CENTERS) OF INCREMENTS
SIZE (LENGTH OF LEG) a
—(BOTH
STAGGERED INTERMITTENT FILLET WELDING SYMBOL
SIZE (LENGTH OF LEG) Saas PITCH (DISTANCE BETWEEN
SPECIFICATION, PROCESS, CENTERS) OF INCREMENTS
OR OTHER REFERENCE FIELD WELD SYMBOL
TAIL (MAY BE OMITTED WELD-ALL-AROUND SYMBOL
WHEN REFERENCE Sa OF INCREMENTS
NUMBER OF SPOT OR
1S NOT USED) PROJECTION WELDS
BASIC WELD SYMBOL SINGLE-V-GROOVE WELDING SYMBOL
OR DETAIL REFERENCE ELEMENTS IN THIS AREA
REMAIN AS SHOWN
WHEN TAIL AND ARROW SIZE (DEPTH OF CHAMFERING) i: 4%
ARE REVERSED CRLEHAMIEEMINCALOUR ON = 60°. ROOT OPENING
THICKNESS OF MEMBERS GROOVE ANGLE

-Supplementary Symbols Used with Welding Symbols

WELD-ALL-AROUND SYMBOL FIELD WELD SYMBOL MELT-THRU SYMBOL

WELD-ALL-AROUND SYMBOL MELT-THRU SYMBOL IS NOT

INDICATES THAT WELD EXTENDS FIELO WELD SYMBOL DIMENSIONED (EXCEPT HEIGHT)
COMPLETELY AROUND THE JOINT INDICATES THAT WELD IS
TO BE MADE AT A PLACE
OTHER THAN THAT OF
INITIAL. CONSTRUCTION oe
ANY APPLICABLE WELD SYMBOL

Basic Joints— Identification of Arrow Side and Other Side of Joint and
ARROW OF ARROW SIDE ARROW SIDE ARROW SIDE. a ARROW SIDE
WELDING SYMBOL OF JOINT OF JOINT OF JOINT ARROW OF OF JOINT
WELDING SYMBOL

OTHER SIDE OTHER SIDE

OF JOINT OF JOINT

OTHER SIDE \ OTHER SIDE

OF JOINT WELDING SYMBOL WELDING SYMBOL OF JOINT
BUTT JOINT CORNER JOINT TEE JOINT

DESIGNATION OF WELDING PROCESSES BY LETTERS

CAW : FW : Flash Weldini RB Resistance Brazing

cw Goa wowing se GMAW 2121212. 272Gas Metal-are Welding Projection Welding
08 Dip Brazing GTAW ........ Gas Tungsten-Arc Welding RPW
Diftusion WeldingWelding 18 » Induction Brazing RSEW Resistance-Seam Welding
OFw Electron Beam RS ooo... Infrared Braziny RSW Resistance-Spot Welding
caw SAW Submerged Arc Welding
Ew . Electrosiag Welding Ww Induction Welding
exw Explosion Welding Lew +s 1. Laser Beam Weldiny SMAW Shielded Metal-Arc Welding
Furnace Brazing OAW Oxyacetylene Welding sw Stud Welding
FB a} secs es esTorch Brazing
Welding @ ©ONW : Oxyh ydrogen Welding
FCAW ...
FOW Forge WeldinsArc
Flux Cored PAW Plasma-Arc Welding ™ Thermit Welding
FRW Friction Welding PEW : Percussion Weldin usw vo... Ultrasonic Welding
PGW Pressure Gas Welding uw +. ou. . Upset Welding

914
 BACK OR
SURFACING
BACKING

SINGLE-V-GROOVE WELDING SYMBOL INDICATING ROOT PENETRATION PROJECTION WELDING PROJECTION WELDING SYMBOL
REFERENCE MUST BE USED
eS PITCH (DISTANCE
o:: PER WELD) OK OF RPW firme CENTERS) OF
[DEPTH H OF CAME CHAMFERERINCS Ae
size PLUS = ogame 190:SS ROOT OPENING WELD MAY BE USED
INSTEAD FOR CIRCULAR
500 6 WEDS
JELROOT T PI GENETRATIO! yw GROOVE ANGLE PROJECTION WELDS. (4) ————_
NUMBER OF WELDS

DOUBLE-BEVEL-GROOVE WELDING SYMBOL SIZE (WIDTH OF WELD) SEAM WELDING SYMBOL LENGTH OF WELDS OR
STRENGTH IN LB. PER INCREMENTS.
‘X_ ARROW POINTS TOWARD LINEAR INCH MAY BE <30" 3=9: OMISSION INDICATES THAT
OMISSION OF SIZE DIMENSION MEMBER TO BE CHAM/ERED USED INSTEAD RSEW WELD EXTENDS BETWEEN
INDICATES A TOTAL DEPTH = |_*——" PROCESS REFERENCE ABRUPT CHANGES IN
OF CHAMFERING EQUAL TO MUST BE USED TO DIRECTION OR AS DIMENSIONED
THICKNESS OF MEMBERS
ROOT OPENING PITCH (DISTANCE BETWEEN
GROOVE ANGLE INDICATE PROCESS
DESIRED CENTERS) OF INCREMENTS
WELDING SYMBOLS FOR COMBINED WELDS
FLASH OR UPSET WELDING SYMBOL
T-3 PROCESS REFERENCE
MUST BE USED TO
t Fw INDICATE PROCESS
60° 4 DESIRED

PLUG WELDING SYMBOL [PITCH (DISTANCE SQUARE-GROOVE WELDING SYMBOL

BETWEEN CENTERS) OF
Le WELDS ,
DEPTH OF FILLING IN
SIZE (OIA. OF HOLE AT ROOT) —, 6 INCHES OMISSION OF
INCLUDED ANGLE 45° OMISSION INDICATES SIZE INDICATES
OF COUNTERSINK FILLING IS COMPLETE COMPLETE JOINT
PENETRATION
SLOT WELDING SYMBOL FLARE-V-AND FLARE-BEVEL-GROOVE WELDING SYMBOL
ORIENTATION, LOCATION
AND ALL DIMENSIONS
ae BEEING
OMISSION INDICATES
OTHER THAN DEPTH OF
FILLING ARE SHOWN
ON) THE DRAWING
SIZE
AS
IS CONSIDERED
EXTENDING ONLY
of 7
FILLING IS COMPLETE TO TANGENT POINTS ROOT OPENING

SPOT WELDING SYMBOL PITCH (DISTANCE EDGE- AND CORNER-FLANGE WELD SYMBOLS
NUMBER OF WELDS (5) __ fects CENTERS) OF
Os WELDS 143
SIZE (DIA. OF WELD). RSW [process REFERENCE t | 7 67 3
STRENGTH IN LB. PER MUST. 8E USED TO WS FIOEN rich eer pis
WPECDIM ATE BESUSED,
INSTEAD INDICATE PROCESS
DESIRED
sve oF weio —— "© OF Tancency

FLUSH-CONTOUR SYMBOL CONVEX-CONTOUR SYMBOL

FLUSH-CONTOUR SYMBOL oh
INDICATES FACE OF WELD FINISH SYMBOL (USER'S FINISH SYMBOL (USER'S
TO BE MADE FLUSH. WHEN STD.) CONVEX.CONTOUR SYMBOL STD.)
USED WITHOUT A FINISH INDICATES METHOD OF INDICATES FACE OF WELD INDICATES METHOD OF
SYMBOL, INDICATES WELD OBTAINING SPECIFIED TO BE FINISHED TO OBTAINING SPECIFIED
TO BE WELDED FLUSH ™ CONTOUR BUT NOT CONVEX CONTOUR CONTOUR BUT NOT
WITHOUT SUBSEQUENT DEGREE OF FINISH ra DEGREE OF FINISH
FINISHING Ll

Arrow-Side and Other-Side Member of Joint

ARROW SIDE
IDE | ARROW-SIDE OF JOINT =
OOF JOINT | MEMBER OF JOINT ZH
1
| |
S !
2 KE \
capes LSS"
ARROW SIDE | 4 OTHER-SIDE ARROW =
OF JOINT OF WELDING
ARROW OF | MEMBER OF JOINT
SYMBOL
WELDING SYMBOL | AR ROWE
WELDING OE
SYMBOL ~ OTHER SIDE
2
LAP JOINT EDGE JOINT Castal

DESIGNATION OF ZZ NONPREFERRED SYMBOLS;

S I ROCe coee.by, ns LZ USE PREFERRED SYMBOL WITH PROCESS REFERENCE IN THE TAIL
‘AAC Air Carbon-Arc Cutting
AC Arc Cutting
AOCE oe Oxygen-Arc Cutting
CAC Carbon-Arc Cutting
FOC Chemical Flux Cutting
MAC Metal-Arc Cutting
oc Oxygen Cutting
PAC Plasma-Arc Cutting
PoC asks Metal Powder Cutting

915
Table D-2 International System of Units (abbreviated SI) and
Symbols
Quantity Unit SI Symbol Formula

length meter m =
mass kilogram kg =
time second s =
thermodynamic temperature kelvin K =
plane angle radian rad =

acceleration meter per second squared — m/s?

angular acceleration radians per second squared _ rad/s?
angular velocity rddians per second — rad/s*
area square meter — m?
density kilogram per cubic meter — kg/m?
energy joule J N-m
force newton N kg-m/s?
frequency hertz Hz (cycle)/s
power watt WwW J/s
pressure newton per square meter — N/m?
quantity of heat joule J N-m
specific heat joule per kilogram-kelvin — J/kg°K
stress newton per square meter oa N/m?
thermal conductivity watt per meter-kelvin _— W/m-K
velocity meter per second _ m/s
viscosity, dynamic newton-second per square meter — N-s/m?
viscosity, kinematic meter squared per Second _ m7? /s
volume cubic meter — m?
work joule J N-m
Table D-3 Conversion Factors for SI Units (symbols of SI units
given in parenthesis)
To convert from To Multiply by

atmosphere (normal) newton/meter? (N/m?) 1.013250 X 105

British thermal unit (Btu) joule (J) 1.055056 X 10°
Btu/hour watt (W) 2.930711 X 107
Btu/ft?-minute watt/meters? (W/m?) 1.891489 X 10?
centipoise newton-second/meter? (N+s/m?) 1.000000 x 1073
centistoke meter? /second (m?/s) 1.000000 x 10°
cubic foot meters? (m*) 2.831685 X 10?
cubic inches meters? (m?) 1.638706 X 1075
cubic foot/minute meters? /second (m?/s) 4.719474 X 1074
cubic foot/second meters? /second (m?3/s) 2.831685 X 10?
cubic inches/minute meters? /second (m?/s) 2.731177 X 107
degree (angle) radian (rad) 1.745329 X 10?
degree Celsius (i.e. centigrade) kelvin (K) ty =te + 273.15
degree Farenheit degree Celsius (i.e. centigrade) to = (tp — 32)/1.8
degree Farenheit kelvin (K) ty = (tp + 459.67)/1.8
degree Rankine kelvin (K) ty =tp/1.8
fluid ounce (U.S.) meter? (m?) 2.957353 X 1075
foot meter (m) 3.048000 X 107!
foot/second meter/second (m/s) 3.048000 X 107!
foot/second? meter/second? (m/s?) 3.048000 X 107!
foot-pound-force joule (J) 1.355818
foot-poundal joule (J) 4.214011 X 10°?
foot-pound-force/hour watt (W) 3.766161 X 10
foot-pound-force/minute watt (W) 2.259697 X 107?
foot-pound-force/second watt (W) 1.355818
gallon (U.S. liquid) meter? (m*) 3.785412 X 107°
gallon (U.S. liquid)/minute meter? /second (m3/s) 6.309020 X 10°$
horsepower (550 ft-lb-f/s) watt (W) 7.456999 X 10?
hour (mean solar) second (s) 3.600000 x 10°
inch meter (m) 2.540000 X 10°?
inch of mercury (32° F) newton/meter? (N/m?) 3.386389 X 10°
inch of mercury (60° F) newton/meter? (N/m?) 3.376850 X 10°
inch/second meter/second (m/s) 2.540000 X 107?
inch/second? meter/second? (m/s?) 2.540000 X 10?
kelvin degree Celsius (i.e. centigrade) te =ty — 273.15
kip newton (N) 4.448222 x 10°
kip/inch? newton/meter? (N/m?) 6.894757 X 10°
kilowatt-hour joule (J) 3.600000 X 10°
mile (U.S. statute) meter (m) 1.609344 X 10°
mile/hour (U.S. statute) meters/second (m/s) 4.470400 X 107!
mile/hour (U.S. statute) kilometer/hour 1.609344
minute (angle) radian (rad) 2.908882 X 10~¢
moment ofinertia (Ibm-ft? ) kilogram-meter* (kg+m?) 4.214012 X 107?
moment of inertia (Ibm-in.? ) kilogram-meter? (kgm?) 2.926397 x 105
moment of section (second meter’ (m*) 4.162314 X 107
moment of inertia) (in.*)
poise (absolute viscosity) newton-second/meter? (N-s/m?) 1.000000 X 107!
pound-force (Ibf avoirdupois) kilogram-force 4.535924 X 107!
pound-mass (Ibm avoirdupois) kilogram (kg) 4.535924 x 107
pound-force/inch newton/meter (N/m) 1.751268 X 10~?
pound-force/foot newton/meter (N/m) 1.459390 X 10!
pound-force-inch newton-meter (N*m) 1.129848 X 107?
pound-force-foot newton-meter (N*m) 1.355818
pound-force-inch/inch newton-meter/meter (N*m/m) 4.448222
pound-force/foot? newton/meter? (N/m?) 4.788026 X 10!
pound-force/inch? (psi) newton/meter? (N/m?) 6.894757 X 10°
pound-force/inch? (psi) kilogram-force/mm? 7.030696 X 10~*
pound-mass/foot? kilogram/meter? (kg/m?) 4.882428
pound-mass/foot? kilogram/meter® (kg/m?) 1.601846 X 10!
pound-mass/inch? kilogram/meter? (kg/m?) 2.767990 X 104%
pound-mass/minute kilogram/second (kg/s) 7.559873 X 1073
pound-mass/second kilogram/second (kg/s) 4.535924 X 107
second (angle) radian (rad) 4.848137 X 10-4
section modulus (ft?) meter? (m?) 2.831685 X 107?
section modulus (in?) meter? (m?) 1.638706 X 10S
stoke (kinematic viscosity) meter? /second (m?/s) 1.000000 x 10~*
statute mile (U.S.) meter (m) 1.609344 X 10?
sqaure foot meter? (m?) 9.290304 X 107%
square inch meter? (m?) 6.451600 X 10~4
square foot/second meter? /second (m?/s) 9.290304 X 10°?
ton (short, 2000 Ibm) kilogram (kg.) 9.071847 X 10?
watt-hour joule (J) 3.600000 X 10?
watt-second joule (J) 1.000000
year (calendar) second (s) 3.153600 X 107
EE

917
Table D-4 Other Useful Conversions
To convert from To Multiply oy

British Thermal units (Btu) foot pounds (ft Ib) 7.780000 X 10?
cubic inch centimeter? (cm?) 1.638706 X 10!
dyne-centimeter newton-meter (N*m) 1.000000 X 1077
dyne/centimeter? newton/meter? (N/m?) 1.000000 X 107!
gram-force/centimeter? newton/meter? (N/m?) 9.806650 X 10!
gravity (standard) meter/second? (m/s?) 9.806650
kilogram-force/centimeter” newton/meter? /N/m? ) 9.806650 X 104
kilogram-force-meter newton-meter (N*m) 9.806650
kilogram-force-meter? newton/meter? (N/m? ) 9.806650
square inch centimeter? (cm? ) 6.451600

918
A strain hardening of, 69
Aluminum casting alloys, 71
Abrasion, 138 table of properties, 878
Abrasive wear in gears, 564 American Institute of Steel Construction
Absolute viscosity, 410 (AISC), 777
Active coils, spring, 722 American National Standards Institute,
Actuating force 363n
brakes, 695-711 American Society for Testing Materials
differential band, 699 (ASTM), 501
long shoe external block, 705 American Welding Society, 113, 171n
long shoe internal block, 706 Analytical model, 236
short shoe block, 702 Anderson, R. T., 357n
simple band, 698 Angle of contact, belt drive, 664, 666, 669
disc clutch, 685-86 Angular backlash, 524, 587
Addendum, 522-23 Anisotropy, 53
Addendum circle, 522-23 Annealing, 45
Adhesive black, 46
lap joints, 840-46 blue, 46
shear, 841-46 box, 46
stress distribution, 843-46 bright, 46
stress distribution factor, 845 flame, 46
Adhesive joints, 841-46 full, 44
Age hardening, 45 inverse, 44
Aging, 45 process, 44
artificial, 69 Annular Bearing Engineers’ Committee
AGMA (American Gear Manufacturers (ABEC), 478
Association) strength formula Anti-Friction Bearing Manufacturers As-
bevel gears, 646-48 sociation (AFBMA), 444, 472,
helical gears, 613 A72n.
spur gears, 545, 554-63 Arc welding, 173-75, 832--33
AGMA (American Gear Manufacturers Asbestos brake lining, 708
Association) wear formulas Assemblies of rolling bearings, 512-15
bevel gears, 649-51 Assumptions, 238
helical gears, 616 Austempering, 38, 46
spur gears, 568-80 Austenite, 21
Air springs, 745, 746 Avner, S. H., 37
AISI-SAE wrought steel designation, 65— Axial groove bearing, 437
67 Axial pitch
Alloy, 53 helical gears, 609
Alloying, 19 worm gears, 625, 633
Alloying elements, 53 Axially loaded rivets, 776-87
Alloying elements, steel, 54-55 Axisymmetric problems, 850-63
Almen, J.O., 108
Alpha iron, 20 B
Aluminum alloys, tables of properties of,
877-78 Back cone of bevel gears, 639
Aluminum alloys, wrought, 67—70 Backlash in gears, 524, 586-88
artificial aging, 69 Bainite, 36, 36n.
designation of, 68-70 Ball bearings
properties of, 877 angular contact, 451-58

919
Ball bearings (cont’d.) Block brake, 700-708
duplexing, 453-58 Bluing, 47
nomenclature for, 446-48 Bolts, 807-810
radial, 448-51 Boundary lubrication, 407
thrust, 459 Boyd, J., 425
Ball screws, 770-72 Brake actuation systems, 709-711
Band brake Brakes, 695-711
differential, 699 Brazing, 178-79
simple, 697 Brinell hardness, 44, 94-100
Base circle for gears, 522—23, 527 for gears, 568
Base pitch for gears, 529 Brittle lacquer, 242
Basic load rating of rolling bearings, 485 Buckingham equation, 565
Basic static load rating of rolling bearings, Buckling of springs, 732-35
483-84 Buckling stress in power screws, 769-70
Beam strength Burnishing of gears, 539
bevel gears, 645 Burnt steel, 47
helical gears, 613 Butt joint, riveted, 781-87
spur gears, 544, 547 Butt welds, 833, 834
worm. gears, 627
Bearing clearance, 425 C
Bearing materials
aluminum alloys, 439 Cantilever leaf spring, 737
babbitt alloys, 439 Capillary tube, 411
cast iron, 439 Carbon steel, 58
copper alloys, 439 Carbonitriding, 47, 49
plastic, 439 Carburizing, 47
porous metal, 439 effect of fatigue of, 129-30
rubber, 439 gears, 591
silver, 439 Case hardening
steel, 439 carbonitriding, 49
Bearing pressure of power screws, 766 carburizing, 47
Bearing stress cyaniding, 49
bolts, 809 flame hardening, 49
rivets, 780 gears, 572, 591
Bearings induction hardening, 50
hydrodynamic, 406 nitriding, 48
hydrostatic, 419 Cast iron, 22, 24-30, 307
partial, 408 alloy
plain or journal, 406-407 Meehanite, 29
thrust, 406 Ni-Resist, 29
Belt chilled, 29
center distance, 669, 670 gray, 27-28
contact angle, 664, 666, 669 malleable, 25—27
drive design, 665, 670 nodular (ductile), 28-29
length, 669, 670 ferritic, 28
Belt drive, 660-71 pearlitic, 28
adjustable speed, 661 white, 25
design, 665 Cast steel, 59
service factors, 667 Castigliano’s theorem, 278-81
transmitted power, 665 Casting, 148-56, 170-71
Bending of gear teeth centrifugal, 155
bevel gears, 642-45 ceramic mold, 152
helical gears, 613 continuous, 155-56
spur gears, 551 Che, Sx
worm gears, 627 full mold, 151
Bending moment, 244 investment, 152
Bending stress in power screws, 767 permanent mold, 151
Beta (8), 21n. plaster mold, 152
Bevel gears, 637-55 plastics, 170-71
Black, P. H., 108 sand, 148-50

920
 shell mold, 150-51 Compression member, 282
Cementite, 20 Compression springs, 721-35
Center distance for gears Compression springs, chart of endurance
crossed helical, 621 limit curves for, 890
helical, 611 Cone clutch, 688—90
spur, 521—23 Conformability of bearing materials, 438
worm, 625 Coniflex bevel gears, 640-41
Center of gravity, 245 Conjugate action, 525
of weld, 836-39 Constant pitch screw thread, 797
Centipoise, 411 Contact length for gears, 529
Centrifugal casting, 155 Contact ratio for gears, 529
Centrifugal clutches, 691, 692 Contact stress number for gears, 568, 575
Ceramic mold casting, 152 Continuous casting, 155-56
Chain Controlled cooling, 50
chordal action, 673 Copper alloys, tables of properties, 879-
failure, 675 80
inverted tooth, 672 Copper and copper alloys, 71-72
pitch, 672-73 brass, 72
roller, 673 bronze, 72
silent, 672 table of properties, 879-80
Chain drive, 660, 671-75 Cores
variable speed, 662 heat treatment, 50
Chemical milling, 185-86 Corrosion
Chordal action, 673 effect on fatigue of, 132
Church, A. H., 346n. fretting, 120-21, 132, 138
Circular pitch, gears galvanic, 139
helical, 609 Corrosive wear in gears, 564
spur, 520-23 Cost, 13-14
worm, 626 Couplings
Circumferential groove bearing, 435 flexible, 382-89
Clamps, 805-806 rigid, 378-82
Clash allowance, 723 Crandall, S. H., 346n.
Classes of fits, 193-96 Creep, 134-35
tables of, 904-911 Critical load, 282
Clearance in bearings, 425, 440 Critical load, springs, 732-35
Clearance of gear teeth, 522-23 Critical speed of shafts, 357-58
Clearance ratio of bearings, 440 Crossed helical gears, 603, 620-23
Clips, 806 Crown gears, 640-41
Closed die forging, 158 Cummings, H. N., 108
Clutches, types of Cumulative damage, 325
cone, 688-90 Curvature, 250
disc or plate, 682-88 Curved beams, 862, 863
electromagnetic, 679 Curved members, 862, 863
hydraulic, 680 Cyaniding, 49
positive contact, 678, 680-82 gears, 459
square-jaw, 681 Cycloidal gears, 525
Coarse screw thread, 796-97 Cylinder, thick wall, 850-54
Coefficient of absolute viscosity, 410
Coefficient of friction D
bearings, 428
brakes and clutches, 685 Dahl, N. C., 346n.
worm gears, 631 Damping capacity, 54
Cold shortness, 53 Davis, G. H. B., 415
Cold working, effect on fatigue of, 125-29 Dean, E. W., 415
Columns Decarburization, 50, 111n.
eccentrically loaded, 286 Dedendum, gears, 522-23
short, 284 Deflection due to bending, 250-60
slender, 281-84, 288-90 Delta iron, 21
Combined stress in power screws, 769 Density of oils, 413
Compatibility in bearings, 438 Design flow, diagram, 2—7

921
Design hints, 212-30 Elastic limit, 84
Design of springs, 725 Elastic stability, 281-90
Diametral pitch Elastic stability, springs, 732-35
helical gears, 610 Elasticity, 81
spur gears, 520, 523 modulus of, 87-88
standard values of Elastomers, 76-77
spur gears, 582 ASTM designation and general prop-
worm gears, 625 erties, 76
Die casting, 152 Electrodes, welding, 832
gears, 536 Electrolytic grinding, 186
materials for, properties of, 882 Electromagnetic brake, 709
Die casting materials, table of, 882 Electromagnetic clutch, 679, 691
Differential band brake, 699 Elliptical shaft, 270
Digital computer, 290-94 Embeddability of bearing materials, 438
Dimensioning, 189-90, 203-204 Embrittlement, 54
Dimensioning terms, definitions of, 190- End leakage for bearings, 430
92 Endurance (fatigue) strength, 100-102
Disc brake, 700 factors influencing, 108
Disc clutch, 682-88 Endurance (fatigue) limit, 102-109, 315,
Disks, rotating, 856-59 325
Displacement, 851, 852 gear teeth, 567
Distortion energy theory, 313 working equation for the, 119-20
fatigue loading, 323 Energy
Double enveloping worm gear, 624 complementary, 278
Double shear, 778 strain, 278
Drawing, 33, 52. See also Tempering Energy methods, 278-81
of gears, 537 Energy and power considerations for
Drop feed oiler, 437 brakes, 695-97
Ductility, 90 Energy storage in flywheels, 858, 859
Dummy load, 279 Epicyclic gear train, 520
Dynamic factor Equilibrium diagram, iron-carbon, 19-22
bevel gears, 646, 649 Equivalent load for rolling bearing, 485—
helical gears, 613 93
spur gears, 554-56, 568, 570 Equivalent number of teeth
Dynamic load bevel gears, 642
bevel gears, 645 helical gears, 612
helical gears, 612 Error in action for spur gears, 558
spur gears, 582 Errosion, cavitation, 139
worm gears, 627 Euler load, 283
Eutectic point, 21
E Eutectoid, 21
Exact theory, 236
Eagan, T. E., 116 Extension springs, 735, 736
Eccentrically loaded rivets, 776, 788-93 External bevel gear, 640
Eccentrically loaded welds, 836-39 Extra fine screw thread, 797-99
Eccentricity of journal bearings, 427, Extruding of gears, 537
429, 441 Extrusion forming, 169-70
Eccentricity ratio of journal bearings, 429,
44]
F
Effective rivet diameter, 779
Efficiency Factor of safety, 8-11, 306, 319
power screws, 762-65 bevel gears, 648, 651
riveted connections, 783 helical gears, 613
spur gears, 594 spur gears, 560, 562, 575, 580
worm gears, 628-31 ultimate strength, 10
Elastic analysis on bolts, 817-22 yield strength, 10
Elastic coefficient Failure pressure, 861
bevel gears, 649 Failure of rolling bearings, 483
helical gears, 616 Failure theories, 301-326
spur gears, 568-69 Faires, V. M., 346n.

922
Fatigue, effect of manufacture on, 111-13 Fracture, brittle and ductile, 17-18
cold working, 125-29 Fracture strength, 84
grinding, 125 Free height of spring, 723, 724
machining, 111 Frettage, 138
polishing, 125 Fretting corrosion, 120-21, 132, 138
welding, 112 Fretting fatigue, 120, 138
Fatigue, effect of stress concentration on, Friction
113-19 brakes and clutches, 685
Fatigue, effects of material variations on journal bearings, 428
(reliability factor), 109 worm gears, 631
Fatigue, influence of size on, 110-11 Friction, coefficient of, 663, 668
Fatigue factors Friction coefficients for power screws,
definition of, 114 763
nonquantitative, 120-34 Full journal bearing, 407
reliability, 109 Full mold casting, 151
size, 110 Fundamental law of toothed gearing, 525
stress concentration, 113-19
surface finish, 110
G
Fatigue failure, 314-26
Fatigue loading on bolts, 817-22 Gadd, E. R., 130
Fatigue of springs, 726-28 Gage factor, 242
Ferrous metals, tables of properties, 868— Galling, 138
76 Galvanic corrosion, 139
Fillet welds, 835-41 Gas welding, 171-72
Fillets on gear teeth, 522, 523 Gaskets in bolted connections, 822
Fine screw thread, 796, 798 Gauge decimals, table of standard, 888
Finishing methods for gears Gear finishing methods
burnishing, 539 burnishing, 539
grinding, 539 grinding, 539
honing, 539 honing, 539
lapping, 539 lapping, 539
shaving, 539 shaving, 539
Fits and tolerance, 854 Gear lubrication, 588-89
tables of, 904-911 Gear manufacturing methods
Fitted bearing, 408 die casting, 536
Flaking, 54 drawing, 537
Flame hardening, 49 extruding, 537
of gears, 591 generating hob, 536
Flat belt drives, 663 generating rack-cutter, 534
Flexible shafting, 395 generating shaper cutter, 535
Flowchart, computer, 292 injection molding, 539
Fluctuating loads on bolts, 817—22 milling cutter, 534
Fluctuating stress, 314 sintered powder, 538
Fluid film lubrication, 406 stamping, 538
Fluid flow in bearings, 428 Gear materials
Flywheels, 857-59 cast irons, 590
Force fit, 854 nonferrous metals, 591
Forced-feed bearings, 435 steels, 590-91
Forging, 157-60 Gear milling, 534
closed die, 158 Gear ratios, 525
open die, 158 Gear tooth systems, 528
upset, 158-60 Gear train, 521, 544, 594
Form factors for Gears, 637-55
bevel gears, 645 bevel, 637-55
helical gears, 613 helical, 602-623
spur gears, 547-50 spur, 519-99
worm gears, 627 worm, 623-37
Formative number of teeth Generated gear teeth, 534-36
bevel gears, 642 Geometric dimensioning and tolerancing,
helical gears, 612 207

923
Geometry factor for gears Homogeneity, 54, 81
bevel gears, 646, 648-50 Homogenizing, 52
helical gears, 613-615, 618 Honing of gears, 539
spur gears, 554, 557, 558, 569, 572, 573, |Hooke’s law, 88
576 Horsepower load for gears, 544
Goodier, J. N., 765 Hot compression molding, 167
Goodman criterion, 324 Hydraulic clutch, 691-92
Grain size, 18 Hydrodynamic lubrication, 406, 421
Graphitizing (temper carbon), 52 Hydrogen embrittlement, 54
Grease lubricated bearing, 438 Hydrostatic lifts, 419, 441
Grinding of gears, 539 Hydrostatic lubrication, 419-20
Grooves in bearings, 435 Hypereutectoid steel, 22
Hypoeutectoid steel, 22
H Hypoid bevel gears, 638, 655

Hagan-Poiseuille law, 411 I

Hand of helical gears, 603 :
Hand oiling of bearings, 437 I section, 247
Hardenability, 39-45 Identification symbols for screws, 815
Jominy test, 41 Included angle, belt, 664
Hardening, strain, 18-19 Induction hardening, 591
Hardness, 94-100 Inertia force, 664
Hardness ratio factors Inertia loading, 856, 857
bevel, 651 Initial tension in bolts, 8i5—17
helical, 613 Initial tension of spring, 735
spur, 575, 578=79 Initially curved members, 862, 863
Heat balance of bearings, 430 Injection molding, 167—68
Heat dissipation gears, 539
bearings, 432 Inlet hole bearing, 437
brakes, 696—97 Instability, 282
gear cases, 632 Interference, 854-56
Heat transfer coefficient gear teeth, 531
bearings, 433 Interference fits, 854-56
worm gear casings, 632 Intermetallic compound, 19
Heat treatment of steels, 19, 31-33 Internal gears
critical cooling rate, 31 bevel, 640
definitions, 45—53 spur, 520, 523
effect on fatigue, 121-25 Internal pressure, 853
martensite, 31-33, 33n. International Nickel Company, 29
quenching, 32, 52-53 Interstitial solution, 18
spheroidite, 33 Investment casting, 152
spheroidizing, 25n., 53 Involute gears, 525
tempering, 33, 53 Iron-carbon diagram, 19-22
Heating Iron-carbon phases
of bearings, 430 alpha iron, 20
of brakes, 695-97 austenite, 21
of worm gear sets, 631-33 cast iron, 22
Helical gears, 601-623 cementite, 20
Helical springs, 721-36 delta iron, 21
Helical torsion spring, 740-42 eutectic point, 21
Helicoil screw inserts, 804, 806 eutectoid, 21
Helix angle, (lead angle), 757 ferrite, 21
helical gears, 607 hypereutectoid, 22
worm gears, 625 hypoeutectoid, 21
Herringbone gears, 606-607 ledeburite, 21
Hertz equation, 565 pearlite, 21
Hertz, H., 446 steelsa 22
High strength, low alloy steel, 58 Isothermal transformation diagrams (TTT
High temperature service steel, 64-65 or S diagrams), 34-38
Hobbing of gears, 536 austempering, 38, 46

924
 martempering, 38 Life of rolling bearings, 483-84
Isotropic, 54 Limit load for wear
Isotropy, 81-82 for bevel gears, 645-51
for helical gears, 616-20
J for spur gears, 564-81
for worm gears, 628
Johnson formula, 285 Limiting error in action for spur gears,
Jominy test, 41 558
Journal bearings, 406, 407 Line of action of gears, 522—23, 527
Juvinall, R. C., 117 Line of centers for gears, 522
Linear backlash, 524, 587
K Lining pressure for brakes and clutches,
685
Keys eIpsonaG@nw lily)
feather, 372 Listing, computer program, 293
Nordberg, 372 Load distribution factor for gears
standard flat, 368-70 bevel, 646-49
standard square, 364-68 helical, 613, 617
Keyways, stress concentration factors Spun 54-554 69=73
for, 892 Loads
Killed steel, 54 bearings, 425
Kinematic viscosity, 414 gear teeth
Kinematics of gear teeth, 519 bevel, 651-55
hypoid, 655
L spiral, 654
straight, 651-53
Laminar flow, 409, 421 helical, 604-605, 612
Laminating, 170 spur, 541-45, 582
Lap joints worm, 629
adhesive, 841-46 Loads, journal bearings, 425
riveted connections, 781-85 Locknuts, 801
Lapping of gear teeth, 539 Long shoe block brake
Ready ai795 external, 702-706
screws, 795 internal, 706-708
worm gears, 625 Long and short addendum gears, 533
Lead angle of worm gears, 625, 633 Low alloy structural steel, 58-59
springs, 737—40 Low temperature service steel, 65
Ledeburite, 21 Lubricants, 440
Length of contact in gears, 529 gas, 406
Length/diameter ratio of bearings, 425, grease, 438
440 oil, 406
Lewis equation Lubrication
bevel gears, 645 spur gears, 589-90
helical gears, 613 thick film bearing, 406
spur gears, 544, 547 Lubrication of rolling bearings, 500-507
worm gears, 627 Lundberg, G., 483
Lewis form factor
bevel gears, 547, 550 M
helical gears, 613
spur gears, 547-50 Machining
worm gears, 627 electrochemical, 185
Life adjustment factors for rolling bear- electrodischarge, 186
ings, 495-97 external cylindrical surfaces, 184
application conditions, 496 flat surfaces, 182
materials, 495—96 internal cylindrical surfaces, 183
reliability, 493-95 ultrasonic, 187-89
Life factor for gears Magnesium alloys, 72-73
bevel, 648-51 table of properties, 881
helical, 613 Malleability, 91
spur, 560-61, 575-77 Malleabilizing, 52

925
Manufacturing processes, classification of, of weld, 836
148 Mooney, R. N., 13
Marin, J., 108 Moore, R. R., 103
Martempering, 38, 52 Multiple disc clutch, 687
Martensite, 33 Multiple V belt drive, 668
Materials
bearings, 438 N
brakes and clutches, 685, 708
gears, 589-92 National Lubrication Grease Institute
screws and bolts, 810-15 (NGLI), 501
Materials, properties of, Natural frequency of springs, 729
carbon and alloy steels, 870-71 Needle roller bearings, 462-63
carburizing and hardening grade steels, Neutral surface, 244
872-75 Newton-Raphson method, 290-92
cast copper base alloys, 880 Newton’s law of viscous flow, 409
cast ferrous metals, 868-69 Newton’s method, 290
die casting metals, 882 Ni-Resist, 29
nickel alloys and nickel base alloys, Nickel alloys, tables of properties, 883-84
883-84 Nickel and nickel alloys, 73-74
plastics, 885-86 table of prope:ties, 883-84
wrought aluminum alloys, 877-78 Nitralloy Corporation, 131n.
wrought and cast magnesium alloys, Nitriding, 48, 52
881 effect on fatigue, 130-32
wrought copper alloys, 879 gears, 591
wrought stainless steels, 876 Nodular iron, 590
Materials for rolling bearings, 479-82 Nominal size of weld, 835
Maximum normal stress theory, 303 Noncircular sections in torsion, 267
Maximum shear, 853-63 Nonmetallic gears, 591-92
Maximum shear theory, 309ff. Normal force, belt, 663
fatigue loading, 321 Normal load on gears
Mechanical mixture, 19 bevel, 651-55
Mechanical properties of materials, 81— helical, 604-605
109 spur, 540, 545
table of, 887 worm, 628-29
Meehanite, 29 Normal pitch
Metal-inert gas arc welding, 833 helical gears, 610
Metallurgical terms, supplementary, 53- worm gears, 627
54 Normal pressure angle
Miles, L. D., 13 helical gears, 610
Milling of gears, 534 worm gears, 627, 633
Miner’s rule, 325 Normalizing, 37, 52
Minimum edge distance, 781 Non-ferrous metals, tables of properties,
Minimum film thickness, 407, 427 877-81
Minimum number of teeth Notch sensitivity, 114
bevel gears, 640 Notch-sensitivity curves, chart of, 892
spur gears, 582 Number of teeth in contact, 529
Miscellaneous primary processes Numerical methods, 290-94
for metals, 161-67 Nuts, 800
for plastics, 171 locknut, 801
Miter gear, 640 lockwasher, 801
Modulus speednut, 801
of elasticity, 87-88
of resilience, 92 O
of toughness, 92-94
Mohr circle, 303 Oil bath bearings, 433
Molding Oil grooves in bearings, 435
hot compression, 167 Oil ring bearing, 433
injection, 167-68 Open die forging, 158
transfer, 168-69 Overhauling of power screws, 761-62
Moment of inertia, 246 Overheated, metal, 52

926
Overload correction factor Plastics, forming of, 167—71
bevel gears, 646, 649 Plastics, table of properties, 885-86
helical gears, 613 Plastics welding, 831
spur gears, 554, 555, 568 Plate clutch, 682-88
Overrunning clutches, 693, 694 Poise, 411
Poiseuille, 411
P Poisson’s ratio, gears, 565
Polar coordinates, 851
Palmgren, A., 483 Polar moment of inertia, 266
Partial bearing, 408 Porous metal bearing, 439
Pearlite, 21 Positive contact clutches, 678, 680-82
Pedestal bearing, 432 Powder metallurgy, 156~—57
Perfect lubrication, 406 Power, transmitted, for belt drive, 665
Permanent mold casting, 151 Power screws, 749-72
Peterson, R. E., 116 definitions for, 757-58
Petroff’s equation, 418 efficiency of, 762-65
Photoelasticity, 242 overhauling of, 761-62
Photoforming, 187 stress considerations for, 765—70
Physical properties of materials, table of, thread forms of, 750—57
887 acme, 750-53
Pillow block bearing, 432 buttress, 754
Pilot bearing, 406 square, 753-54
Pins, as fasteners, 806—807 torque equations for, 758-61
Pitch, 757 Precipitation hardening, 52
rivets, 782 Preheating, 52
screws, 794-96 Preload on bolts, 815-16, 819
Pitch, chair, 672, 673 Pressure angle in gears
Pitch angle, bevel gears, 639-40 bevel, 640
Pitch circle of gears, 520, 522-23 helical, 610
Pitch cone of bevel gears, 639 spur, 522-523, 527
Pitch diameter Pressure distribution in journal bearings,
bevel gears, 639 427
helical gears, 610 Pressure line, 527
spur gears, 522-23 Pressure lubricated bearings, 435, 438
threads, 796-97 Pressure vessel, 320, 850-54
worm gears, 625 Primary processes
Pitch of gear teeth for metals, 148-67
axial, 610 for plastics, 167-71
helical, 610 Principal directions, 303
worm, 625 Principal planes, 303
base, 529 Proof stress, 84
circular, 520, 522-23 Proportional limit, 84
diametral, 520, 523
normal, 610
transverse, 610
Q
Pitch line velocity, 542 Quenching, 52
Pitch point, 522-23
Pitting, 138 R
gears, 564
Plain bearing, 406 Rack-cutter for gears, 534
Planetary gear train, 597, 598 Radial clearance, 408
Plaster mold casting, 152 Radial interference, 854-57
Plastic analysis, 859-62 Radial load on gears
Plastic bearing, 439 bevel, 653
Plastic design, 861, 862 helical, 604-605
Plasticity, 82 spur, 541, 545
Plastics, 74-75 worm, 628-29
properties of, 885 Radial stress, 851-63
thermoplastic, 75 Radiation effects, 139-40
thermosetting, 75 Radius of gyration, 284

927
Raimondi, A. A., 425 Rubber bearings, 439
Ratchet wrench, 248 Rubber compression mounts, 743, 744
Rating life of rolling bearings, 483-84 Rubber mounts, 743-45
Red shortness, 54 Rubber shear mounts, 744, 745
Redundant support, 258
Reliability, 11-13 S
factor in fatigue, 109
Reliability factor for gears Safe stress line, 318
bevel, 648 Safety, 14-15
helical, 613 Safety factor, 306, 319
spur, 560, 562 St. Venant, 271
Resilience, modulus of, 92 St. Venant’s principle, 248
Resistance welding, 175-78, 830-31 Sand casting, 148-50
Restrained member, 275 Saybolt universal seconds, 414
Retaining rings, 806-807 Saybolt viscosity, 414
Reversed stress, 325 Schram, J. 353n.
Reyn, 411 Schulte, W. C., 108
Reynolds, O., 411, 421 Scissor gears, 588
Reynolds’ equation Scoring of gears, 564
one-dimensional flow, 421 Screws, 794-801
two-dimensional flow, 425 Seals for rolling bearings, 509-512
Rimmed steel, 54 Secant formula, 286
Ring-oiled bearings, 437 Secant method, 292
Riveted connection, 778-93 Secondary production processes, 181-89
butt joint, 781-87 chemical milling, 185-86
lap joint, 781-87 electrochemical machining, 185
Rockwell hardness, 41, 44, 94-100 electrodischarge machining, 186
Roller bearings electrolytic grinding, 186
cylindrical, 460-61 external cylindrical surface machining,
needle, 462-63 184
spherical, 468 flat surface machining, 182
tapered, 463-67 internal cylindrical surface machining,
Rolling bearings 183
assemblies of, 512-15 photoforming, 187
ball bearings, 446-60 ultrasonic machining, 187-89
basic load rating of, 485 Self-contained bearings, 432
basic static load rating of, 483-84 Self-drilling screw, 804—805
basis for failure of, 483 Self-energizing brakes
equivalent load for, 485—93 differential band, 699
friction of, 482 long shoe external block, 705
life of, 483-84 long shoe internal block, 707-708
life adjustment for, 495-96 Self-locking
lubrication of, 500—507 differential band brake, 699
materials for, 479-82 long shoe external block brake, 705
nomenclature and types of ball bear- long shoe internal block brake, 707-708
ings, 446-60 worm gears, 631
rating life of, 483-84 Sellers thread, 795
roller bearings, 460 Semikilled steel, 54
seals for, 509-12 Sems, 802
speed limitations of, 507—509 Set screws, 802-803
standard dimensions for, 472-78 Shaft angle
survival probabilities of, 493-95 bevel gears, 640-41
thrust bearings, 459, 460 crossed helical gears, 621
tolerances for, 478-79 Shaft deflection
variable loads in, 497-500 computer method for, 353-56
Rolling bearings subjected to variable numerical method for, 347-52
loads, 497-500 Shaft design by computer, 353-56
Root diameter for gears, 522-23 Shaft design when subject to fluctuating
Root of equation, 290 loads
Rotating disks, 856—59 distortion energy theory, 340

928
 maximum shear theory, 338-39 Sommerfeld number, 425
shock load considerations, 341-42 Spalling, 123n.
Shafting, 322 Specific gravity, lubricating oils, 414
Shafts Specifications, 2-4, 11, 15
critical speed of, 357-58 Speed limitations of rolling bearings, 507—
flexible, 395 509
materials for, 333 Speed nut, 801
torsional stiffness of, 358-60 Speed ratio, chain drive, 671
types of, 332-33 Speed ratio of gears, 525
Shaping of gears, 535 Spherical bearing, 406
Shaving of gears, 539 Spherical roller bearings, 468
Shaw process, 152 Spheroidizing, 53
Shear Spiral bevel gears, 638, 654
center, 263 Spiral torsion spring, 742, 743
deflection, 262 Splash system of lubrication, 437
stress, 262-66, 268-73 Splines
torsional, 265 involute, 374-77
transverse, 261 involute serration, 377
Shear in adhesives, 841-46 straight, 373-74
Shear force, 251 Spot welding, 830, 831
Shear modulus, 266 Spring index, 724
Shear strength Spring loaded scissor gears, 588
direct, 88-89 Spring mares 22s aod,
yield point, 89-90 Spring wires, chart of minimum, tensile
Shear stress, 778, 780, 791 strengths, 889
bolts, 809 Springs
power screws, 767-68 air, 745, 746
rivets, 778, 780, 791 bending, 732-35
Shearing stress in power screws, 767-68 buckling, 732-35
Shell mold casting, 150-51 cantilever, 737
Shock absorber bolt, 800 clash allowance, 723
Shock factor, 322 compression, 721-35
Shoe-type thrust bearing, 406 critical load, 732-35
Short shoe block brake, 700-702 design, 725
Shot peening, effect on fatigue of, 126-27 endurance limit, 726
Shrink fit, 854 endurance limit curves for compres-
SI units, tables of, 916-18 sion, 890
Silver soldering, 178n. extension, 735, 736
Simple support, 255, 282 fatigue, 726-28
Sines, G., 108 free height, 723-24
Single enveloping worm gear, 624 helical, 721-36
Single shear, 778 helical torsion, 740-42
Sintered metal gears, 538 initial tension, 735
Size correction factor leaf, 737-40
bevel gears, 646, 647, 649 minimum tensile strength of wire for,
helical gears, 613 889
spur gears, 569-70 natural frequency, 729
Size of gear teeth, 523 rubber, 743-45
Sleeve bearing, 406 Soderberg criterion, 726—29
Slenderness ratio, 284 solid deflection, 723
Sliding velocity of gear teeth solid height, 723
crossed helical, 621 spiral torsion, 742-43
worm, 631 squared ends, 722
Society of Automotive Engineers (SAE), stress, 724
425 surging, 729-32
Soderberg criterion, 318-21 torsion, 740-43
Soderberg criterion for springs, 726—29 torsion bars, 719-21
Solid deflection of spring, 723 working deflection, 723
Solid height of spring, 723 Spur gears, 519, 599
Solid solution, 19 Squeeze film, 420, 442

929
Stability, 282 thermal, 274
Stability of springs, 732-35 three-dimensional case, 305
Stable lubrication, 406 two-dimensional field, 302
Stainless steel, 58-63 Stress concentration, 316, 318, 322
austenitic, 60-61 Stress concentration factors, charts of,
classification of, 60-62 892-901
ferritic, 60-62 Stress concentration in gear teeth, 553
martensitic, 61 Stress concentration mitigation, 117-19
precipitation hardening, 62 Stress considerations for power screws,
properties of, 876 765-70
Stamping of gears, 538 Stress distribution in adhesive joints, 843-
Standard dimensions for rolling bearings, 46
472-78 Stress distribution factor in adhesive, 845
Statically indeterminate problems, 258 Stress relieving, 53
Steclsw22 Stress-strain laws, in polar coordinates,
alloying elements in, 54—55 851
burnt, 47 Stribeck,; H., 446
carbon in, 55—58 Structural steel, 58-59
cast, 59 Stud bolt, 800
classification of, 55—65 Stulen, F. B., 108
H band, 42 Substitutional solution, 19
high strength, low alloy, 58 Superposition, 252
high temperature, 64-65 Surface condition factor
killed, 54 bevel gears, 649
low temperature, 64-65 helical gears, 569, 574
rimmed, 54 spur gears, 649
semikilled, 54 Surface durability
stainless, 59-63 helical gears, 616
structural, low alloy, 58 spur gears, 564-68
tables of properties, 870-76 Surface quality, 208-212
tool, 63-64 Surface rolling, effect on fatigue of, 128-
Stiffness constant for bolts, 817 29
Stokes, G. G., 414 Surging of springs, 729-32
Straight tooth bevel gears, 637, 639-42 Survival probabilities of rolling bearings,
Strain, 83, 238 493-95
free, 274
true, 274
i
Strain-displacement relationships in polar
coordinates, 851 T section, 246
Strain energy, 278, 312 Tangential load on gears
Strain gages, 239 bevel, 653
Strain hardening, 18-19 helical, 604-605
aluminum, 69 spur, 541, 545
Strength of materials, 236 worm, 628-29
Stress, 83 Tangential stress, 851-63
bending, 242 Tapered roller bearings, 463-67
combined, 302 Temperature
concentration, 237 effect on fatigue of, 133-34
equivalent, 319 effect of high, 137-38
maximum, 237 effect of low, 136—37
maximum shear, 305 Temperature in brakes and clutches,
mean, 318 maximum, 685
normal, 302 Temperature effect on viscosity, 415
at a point, 302 Temperature factor
power screws, 768-69 bevel gears, 648, 651
principal, 303, 314 helical gears, 613, 616
range, 318 spur gears, 560-61, 575, 578
residual, 274 Temperature sensitive properties of ma-
rivets, 779 terials, 135-38
spring, 724 Tempering, 53

930
Tensile strength, 82—83 U
Tensilock screw, 803
Tension, belt, 664 Ultrahigh strength steel, 65
Tension test, 309 Undercutting of gear teeth, 533
Thermal capacity of worm gears, 631-33 Unified thread, 795-99
Thermal expansion coefficient, 274 Uniform pressure
Thermal stress, 274-78 disc clutch, 686, 688
Thick wall cylinder, 850-54 short shoe block brake, 701
Thickness of gear teeth, 522, 523 Uniform wear for disc clutches, 683-85
Thin film lubrication, 406, 407 Universal joints, 390-95
Thompson, W. T., 357n. Upset forging, 158
Thread forms of power screws, 750-54 U. S. Steel Corporation, 34
Threads, stress concentration factors for,
892 Vv
Throat dimension of weld, 835
Thrust bearing, 419-20 V belt, 660, 664-71
Thrust load V screw thread, 794
bevel gears, 653 Velocity gradient, 410
crossed helical gears, 622 Velocity ratio for gears, 525
helical gears, 607 Vidosic, J. P., 10
worm gears, 628-30 Virtual number of teeth
Tightening torque on bolts, 815-17 bevel gears, 645
Tolerances helical gears, 612
accumulation and nonaccumulation of, Viscosity, 409-410, 414, 501
204-205 Viscosity index, 415
casting, 148-56
definitions of, 190-207 Ww
selection of, 196-206
stackup of, 205 Waisman, J. L., 108
statistical determination of, 205—207 Warping, 269
tables of, 904-911 Waste-packed bearing, 433
Tolerances for rolling bearings, 478-79 Wear, 138-39
Tool steel, 63-64 Wear of clutch lining, 683-84
Torque, V belt, 665 Wear in gears
Torque capacity abrasive, 564
brakes, 695-711 corrosive, 564
long shoe external block, 705 pitting, 564
long shoe internal block, 706 scoring, 564
short shoe block, 701 Wear load
simple band, 698 bevel gears, 645
disc clutch, 685, 686 helical gears, 616
Torque equations for power screws, 758- spur gears, 567-68
61 worm gears, 628
Torsion, 265 Wear load factor
slotted tube, 273 bevel gears, 645
thin rectangular section, 272 helical, 616
Torsion bars, 719-21 spur gears, 567
Torsion springs, 740—43 worm gears, 628
Torsional deflection, 266 Weld
Torsional rigidity, 273 arc, 832-33
Torsional spring constant, 267 butt, 833, 834
Torsional stiffness of shafts, 358-60 center of gravity, 836-39
Toughness, modulus of, 92—94 eccentrically loaded, 836-39
Tower, B., 421 fillet, 835-41
Transfer molding, 168-69 moment of inertia, 836
Transformation range, 53 nominal size, 835
Transmitted power, belt drive, 665 resistance, 830, 831
Transverse pitch helical gears, 609 shear loading, 835-41
Tungsten-inert gas welding, 833 spot, 830, 831
Turbulent flow, 409 tension loading, 833-34

931
Weld (cont’d.) seam, 176
throat, 835 spot, 175-76
Welded joints, fundamental types of, 180 upset, 177
Welding, 171-81 shielded metal, 173
arc, 173-75 submerged, 173-74
brazing, 178-79 symbols for, 914-15
dip, 179 thermit, 179
furnace, 178 ultrasonic, 179
induction, 178 Welding symbols, standard, 914-15
torch, 178 Welds, 830-41
cold, 179 subject to non-planer loads, 839-41
diffusion, 179 Whiskers, metal, 18
electron beam, 181 Whitworth thread, 795
electroslag, 181 Whole depth of gear teeth, 522-23
explosion, 179 Wick-feed oiler, 437
forge, 179 Working depth for gears, 522—23
friction, 179 Worm gears, 623-37
gas, 171-72 standards and proportions, 633-34
oxyacetylene, 172 Wrought iron, 22-24
oxyhydrogen, 172 Wrought steel, designations for, 65-67
pressure, 172
induction, 179 Y
laser beam, 179
plasma, 174-75 Yield point (strength), 85
resistance, 175—78
flash, 177 Z
percussion, 177—78
projection, 176-77 ZEROL bevel gears, 638, 652-53

932
 id’

od
5)
y
2ee
ie
wa):
gS eeeexre
Seanen
Pebretee

s}
i 134

Pim eas =

eas =

sfarht
Spataetee
S