0% found this document useful (0 votes)
13 views5 pages

Aisc Journal 2001 - V38 04 02

Uploaded by

hiram.roque
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
13 views5 pages

Aisc Journal 2001 - V38 04 02

Uploaded by

hiram.roque
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

2000-27.

qxd 11/8/2001 11:05 AM Page 176

Design Considerations for W-Shape Flexural Members


Built-Up from Plates
RICHARD M. DRAKE and MICHAEL E. BANKES

INTRODUCTION Compact Sections

I t is common practice in some areas of the world to build-


up or fabricate W-shapes from plate as alternates to rolled
W-shapes.
Compact sections are able to develop the full plastic buck-
ling capacity of the section. AISC specifications require that
flanges must be continuously connected to the web(s) for a
It is also possible that in order to meet project delivery section to be considered to be compact.
schedules, a fabricator may request approval from the engi- Compact sections must have web slenderness ratios, h/tw
neer-of-record to substitute a built-up W-shape for a tem- that are less than prescribed limits, λp. For rolled sections, h
porarily unavailable rolled W-shape. In either case, it would is defined as the distance between the web toe fillets. For
be practicable if the engineer-of-record can select his W- built-up sections, h is defined as the distance between
shape members using the AISC Manual (AISC, 1994) and flanges.
LRFD Specification (AISC, 1999) and companion design
aids. Flexural Shear
This paper addresses design and/or fabrication issues for
By elastic theory, horizontal flexural shearing forces (Fv)
the practice and attempts to arrive at “rules” to allow safe
are present across the cross-section, as determined by the
and timely application.
equation Fv = VQ/I. For rolled sections, these forces are
accommodated in the base metal. For built-up sections,
DESIGN ISSUES
AISC specifications require that the flange-to-web welds be
Residual Stresses proportioned to transfer these shear forces. AISC specifica-
tions require that all welds be in accordance with AWS D1.1
Residual stresses occur in rolled sections as a result of the
requirements (AWS, 2000). Neither AWS nor AISC specifi-
rolling and cooling process. Compressive and tensile
cations require a specific weld process or weld type for
stresses are “locked in” to the section as the rolled section
these welds.
cools. In built-up sections, residual compressive and tensile
Rolled W-shapes generally have excess flexural shear
stresses are “locked in” as the weld metal and heat-affected
capacity. The decision or need to substitute built-up sec-
zone cool. See Figure 1.
tions for rolled shapes offers the opportunity to reduce the
Compression residual stresses affect local stability and
thickness and weight of the W-shape web, yet still provide
are accounted for in AISC’s LRFD specification flexural
sufficient design shear strength. As presented, this study
design procedures. It is AISC’s position that compressive
residual stresses are higher in built-up sections than in
rolled sections. As a result, the LRFD specification pre-
scribes lower flexural strength for built-up shapes than
rolled shapes for beams where the inelastic buckling limit
state governs, typically medium span beams where Lb
exceeds the compact section limit Lp.
Note that residual stresses are not directly accounted for
in the AISC compression strength provisions because the
design equations are a reasonable approximation of several
different column design curves that consider buckling axis
and residual stresses. See the SSRC Guide for additional
discussion (SSRC, 1998).

Richard M. Drake, S.E. is principal structural engineer, Fluor


Daniel, Inc., Aliso Viejo, CA.
Michael E. Bankes, P.E. is design structural engineer, Fluor
Daniel, Inc., Aliso Viejo, CA. Fig. 1. Typical residual stress pattern in W-shape sections.

176 / ENGINEERING JOURNAL / FOURTH QUARTER / 2001

© 2003 by American Institute of Steel Construction, Inc. All rights reserved. This publication or any part thereof must not be reproduced in any form without the written permission of the publisher.
2000-27.qxd 11/8/2001 11:05 AM Page 177

does not consider a reduction in web thickness because of ry X 1


the resulting reduction in design moment strength. How- Lr = 1 + 1 + X 2 FL 2 (LRFD Eq. F1-6)
FL
ever, the practicing Structural Engineer is encouraged to
where:
consider the possibility of reducing web thickness if the
project schedule allows time for the investigation. π EGJA (LRFD Eq. F1-8)
X1 =
Sx 2
WELDING ISSUES 2
Cw  S x  (LRFD Eq. F1-9)
X2 = 4
Weld Processes I y  GJ 
Shielded Metal Arc Welding (SMAW) is the standard weld FL = smaller of (Fyf – Fr) or Fyw (LRFD Sec. F1.2)
process for structural welds, especially in the field. Sub-
Fr = compressive residual stress (LRFD Sec. F1.2)
merged Arc Welding (SAW) is a more common shop weld-
= 10 ksi for rolled shapes
ing process. SAW yields a larger heat-affected zone,
= 16.5 ksi for welded built-up shapes
resulting in a higher strength fillet weld than the same size
fillet weld made using SMAW. As a result, AWS and AISC
Flexural Yielding Limit State
prescribe larger effective throats (and design strengths) for
fillet welds using SAW than those using SMAW. There is The compression flange is sufficiently braced to allow the
no design difference for groove welds. full plastic moment, Mp, to be reached.
For compact sections with Lb < Lp:
AWS D1.1 Section 2—Design Provisions
Mn = Mp (LRFD Eq. F1-1)
Section 2.1.15 requires that “If two or more plates or rolled
where:
shapes are used to build up a member, sufficient welding (of
the fillet, plug, or slot type) shall be provided to make the Mp = FyZ ≤ 1.5My (LRFD Sec. F1.1)
parts act in unison, but not less than that would be required
to transfer calculated stress between the parts joined.” Inelastic Lateral-Torsional Buckling Limit State
Sections 2.1.16, 2.1.17, and 2.1.18 give maximum
The compression flange is not sufficiently braced to allow
spacing requirements for intermittent welds of built-up
the full plastic moment, Mp, to be reached. The nominal
members.
moment capacity is limited by inelastic lateral-torsional
buckling of the beam’s compression flange.
AWS D1.1 Section 5—Fabrication Provisions
For compact or noncompact sections with Lp < Lb ≤ Lr:
Section 5.21 requires control of distortion and shrinkage
  Lb − Lp  
when welding built-up members. This section contains
(
M n = Cb  M p − M p − Mr  )
  ≤ Mp
 Lr − Lp  
mostly “good practice” requirements to control and mini- 
mize distortion without locking in excessive residual
stresses. (LRFD Eq. F1-2)

AISC DESIGN REQUIREMENTS


Chapter F and Appendix F of the LRFD Specification pre-
scribes the flexural design requirements for W-shaped
members. The design requirements reflect distinct flexural
limit states, with each limit state bounded by a characteris-
tic range of compression flange unbraced length, Lb. See
Figure 2.
All notation is consistent with that defined in the LRFD
Specification.

Limit State Bounding Ranges


The flexural limit state bounding ranges are defined as:
E Fig. 2. Nominal flexural strength as a function of unbraced compression
L p = 1.76 ry (LRFD Eq. F1-4)
Fyf flange length and moment gradient (Cb = 1.00).

ENGINEERING JOURNAL / FOURTH QUARTER / 2001 / 177

© 2003 by American Institute of Steel Construction, Inc. All rights reserved. This publication or any part thereof must not be reproduced in any form without the written permission of the publisher.
2000-27.qxd 11/8/2001 11:05 AM Page 178

where: STUDY METHODOLOGY


Mr = FLSx (LRFD Eq. F1-7) In order to compare AISC rolled shapes to built-up shapes,
a spreadsheet was developed using MathCAD software
Elastic Lateral-Torsional Buckling Limit State (MathCAD, 1998) that calculated all of the required design
variables. The study addressed 50 ksi steel only, and the use
The compression flange is not sufficiently braced to allow
of metric plates to fabricate the built-up sections.
the full plastic moment, Mp, to be reached. The nominal
moment capacity is limited by elastic lateral-torsional
Spreadsheet Input
buckling of the beam’s compression flange.
For compact or noncompact sections with Lr < Lb: Study input variables were:
• Yield Strength, Fy = 50 ksi
Mn = Mcr ≤ Mp (LRFD Eq. F1-12)
• Modulus of Elasticity, E = 29000 ksi
where: • Shear Modulus, G = 11200 ksi
• Resistance factor, φb = 0.90
2
π  πE  • Bending coefficient, Cb = 1.00
M cr = Cb EI y GJ +  I y Cw (LRFD Eq. F1-13)
Lb  L 
b
• Available metric plate thicknesses

Cb S x X 1 2 X 12 X 2 It is recognized that plates are typically produced in dis-


= 1+ (LRFD Sec. F1.2)
Lb L 
2 crete plate thicknesses. The metric plate thicknesses con-
ry 2 b  sidered in this study were selected based on a study of
 ry 
metric plate availability for South American projects: 6, 8,
10, 12, 16, 19, 25, 31, 37, 43 and 50 mm.
Flange And Web Buckling Limit States
Input variables for the rolled shapes were:
Noncompact sections are capable of developing the mate-
• AISC rolled shape name, generally selected from the
rial yield stress across some, but not all, of the beam cross-
top and bottom of section lists in the AISC Manual
section before compression buckling. The nominal moment
• Flange width, from AISC Manual
capacity is limited by local buckling of either the beam
• Flange thickness, from AISC Manual
flange or web.
• Section depth, from AISC Manual
For noncompact sections:
• Web thickness, from AISC Manual
 λ−λp  • Compressive flange residual stress, Fr = 10 ksi
(
M n = M p − M p − Mr  ) 
 λr − λ p 
(LRFD Eq. A-F1-3)
Input variables for the built-up shapes were:
• Compressive flange residual stress, Fr = 16.5 ksi
where λ, λr, and λp are the limiting width-thickness ratios
for compression elements defined in LRFD Specification
Spreadsheet Calculations
Table B5.1.
Calculated variables for the built-up shapes were:
Effect Of Compressive Residual Stresses • Flange width, match rolled shape
• Flange thickness, from available metric plate thick-
Note that compressive residual stresses, Fr, are defined dif-
nesses
ferent for rolled shapes and welded built-up shapes. As a
• Section depth, match rolled shape
result all equations that include the effects of Fr will be dif-
• Web thickness, from available metric plate thick-
ferent for rolled shapes and welded built-up shapes. The
nesses
effect of the residual stresses is to reduce the design
• Compressive flange residual stress, Fr = 16.5 ksi
strengths of a welded built-up shape from the rolled shape
with the same flange and web plate dimensions. This
Calculated variables for both the rolled shapes and built-up
reduced strength can be compensated for by selecting plates
shapes were:
for the welded built-up shape that are larger than the com-
• Cross-section area, A
parable rolled shape.
• Weak-axis moment of inertia, Iy
• Weak-axis radius of gyration, ry
• Elastic section modulus, Sx
• Plastic section modulus, Zx

178 / ENGINEERING JOURNAL / FOURTH QUARTER / 2001

© 2003 by American Institute of Steel Construction, Inc. All rights reserved. This publication or any part thereof must not be reproduced in any form without the written permission of the publisher.
2000-27.qxd 11/8/2001 11:05 AM Page 179

Table 1. W8×40 Comparison Table 2. W12×65 Comparison

COMPARISON SHEET COMPARISON SHEET


Fy = 50 ksi Fy = 50 ksi
Size = "W8x40" Size = "W12x65"
Rolled_Section = "Compact" Rolled_Section = "Non-Compact"
Builtup_Section = "Compact" Builtup_Section = "Non-Compact"

200 400

150 300

( )
φMp_ROL L b φMp_ROL Lb ( )
kip⋅ ft kip ⋅ ft
100 200
( )
φMp_BUP L b φMp_BUP Lb ( )
kip⋅ ft kip ⋅ ft

50 100

0 0
0 5 10 15 20 25 30 35 40 0 10 20 30 40 50
Lb Lb
ft ft

Variable Rolled Section Built-Up Section Variable Rolled Section Built-Up Section
Flange Width bf = 8.07 in bf = 8.07 in Flange Width bf = 12 in bf = 12 in
Flange Thickness t f_ROL = 14.224mm t f_BUP = 16 mm Flange Thickness t f_ROL = 15.367mm t f_BUP = 16 mm
Section Depth d = 8.25 in d = 8.25 in Section Depth d = 12.12 in d = 12.12 in
Web Thickness t w_ROL = 9.144 mm t w_BUP = 10 mm Web Thickness t w_ROL = 9.906 mm t w_BUP = 10 mm
Plastic Length (Lp) Lp_ROL = 7.271 ft Lp_BUP = 7.309 ft Plastic Length (Lp) Lp_ROL = 10.772ft Lp_BUP = 10.815 ft
Moment Lb < Lp φMp_ROL ( 0.1 ⋅ ft) = 147.5 kip ⋅ ft φMp_BUP ( 0.1 ⋅ ft) = 163.3 kip ⋅ ft Moment Lb < Lp φMp_ROL ( 0.1 ⋅ ft) = 351.6 kip ⋅ ft φMp_BUP ( 0.1 ⋅ ft) = 365.9 kip ⋅ ft
Elastic Length (Lr) Lr_ROL = 26.252ft Lr_BUP = 34.391ft Elastic Length (Lr) Lr_ROL = 31.351ft Lr_BUP = 36.807ft
Moment @ Lr φbMr_ROL = 105.261kip ⋅ ft φbMr_BUP = 96.924kip ⋅ ft Moment @ Lr φbMr_ROL = 259.39kip ⋅ ft φbMr_BUP = 224.51kip ⋅ ft
Weight WeightROL = 39.49 plf WeightBUP = 43.96 plf Weight WeightROL = 63.887plf WeightBUP = 65.993 plf
Weight Increase ∆ = 11.32 % Weight Increase ∆ = 3.296 %

• Torsional moment of inertia, J φbMn vs. Lb plots for the rolled section and the built-up sec-
• Warping constant, Cw tion on the same graph at the same scale. The summary
• Torsional parameters, X1, X2 sheets also list selected variables of the built-up section
• Slenderness parameters, λ properties compared side-by-side to the rolled section prop-
• Compact element limiting parameter, λp erties, and a weight comparison.
• Noncompact element limiting parameter, λr Three representative comparisons were selected and are
• Plastic length, Lp presented as Tables 1, 2, and 3 to illustrate typical results
• Elastic Length, Lr and to support the authors’ conclusions. The three selected
• Yield moment, My comparisons are representative of the 86 sections analyzed.
• Design moment at Lp, φbMp
• Design moment at Lr, φbMr DATA INTERPRETATION
• Design moment at all Lb, φbMn
A total of 86 sections were analyzed and compared in the
• Section weight
study. The following general observations can be made:
The calculated section properties for rolled shapes do not
1. For compact sections, the built-up section yielding
agree exactly with those published by AISC because they
limit state strength is always greater than or equal to
were calculated based on rectangular dimensions with no
that of the rolled section.
consideration of web fillets. Any differences are small and
2. For noncompact sections, the built-up section flange
do not noticeably affect the study results.
and web buckling limit state strengths are always
greater than or equal to that of the rolled section.
Spreadsheet Output
3. For all sections, the limiting unbraced length Lr for the
After the necessary variables were calculated, summary built-up section is always greater than or equal to that
sheets were developed for each selected shape that included of the rolled section.

ENGINEERING JOURNAL / FOURTH QUARTER / 2001 / 179

© 2003 by American Institute of Steel Construction, Inc. All rights reserved. This publication or any part thereof must not be reproduced in any form without the written permission of the publisher.
2000-27.qxd 11/8/2001 11:05 AM Page 180

Table 3. W24×162 Comparison CONCLUSIONS


COMPARISON SHEET Although some of the data indicates some built-up shapes
Fy = 50 ksi
Size = "W24x162" falling below the design strength of the rolled shape, the
Rolled_Section = "Compact"
margins are very small, and are confined to a small range of
Builtup_Section = "Compact"
unbraced lengths near the transition between compression
2000
flange elastic and inelastic lateral-torsional buckling. The
margins are well within the accuracy of the loads normally
1500 applied to structures.
φMp_ROL Lb ( ) An examination of the data supports the conclusion that
kip ⋅ ft the built-up W-shapes can be safely substituted for rolled
1000
φMp_BUP Lb ( ) W-shapes provided that all of the following occur:
kip ⋅ ft
1. Flange and web plate thicknesses (tf and tw) are equal
500 to or larger than the rolled shape thicknesses;
2. Flange and web plate widths (bf and h) are equal to or
larger than the rolled shape thicknesses;
0
0 10 20 30 40 50
3. Plate material strengths (Fy and Fu) are equal to or
Lb
ft
larger than the rolled shape strengths; and
4. All welding is in accordance with AWS D1.1, and
Variable Rolled Section Built-Up Section
Flange Width
5. Fabrication shop inspection efforts are increased to
bf = 12.955in bf = 12.955in
Flange Thickness t f_ROL = 30.988mm t f_BUP = 31 mm
account for the greater quantity of shop welding.
Section Depth d = 25 in d = 25 in This study addressed only 50 ksi yield strength steel and
Web Thickness t w_ROL = 17.907 mm t w_BUP = 19 mm metric plate thickness increments. It is the author’s opinion
Plastic Length (Lp) Lp_ROL = 10.793ft Lp_BUP = 10.686 ft
3
that studies of steels with higher yield strengths and differ-
Moment Lb < Lp φMp_ROL ( 0.1 ⋅ ft) = 1745.8k ⋅ ft φMp_BUP ( 0.1 ⋅ ft) = 1.8 × 10 k ⋅ ft
Elastic Length (Lr) Lr_ROL = 32.395ft Lr_BUP = 37.054ft
ent plate thickness increments would yield the same results.
Moment @ Lr 3
φbMr_ROL = 1.235 × 10 kip ⋅ ft
3
φbMr_BUP = 1.043 × 10 kip ⋅ ft Studies of steels with lower yield strengths are less likely
Weight WeightROL = 161.683plf WeightBUP = 165.026plf to yield similar results because the residual stress values in
Weight Increase ∆ = 2.067 % the AISC Specification will be a larger percentage of the
yield stress, reducing the values of the built-up section
nominal moment strengths relative to those of the compara-
4. For all sections, the limiting buckling moment Mr for ble rolled section.
the built-up section is always less than or equal to that In addition, the practicing Structural Engineer is encour-
of the rolled section. aged to consider the possibility of reducing web thickness if
5. For all sections, the slope of the nominal strength the project schedule allows time for the investigation.
Equation F1-2 is steeper for the built-up section than
that of the rolled section. The start point for the equa- REFERENCES
tion (Lp, Mp) for the built-up section is always at an Mn American Institute of Steel Construction (AISC) (1994),
value greater than or equal to that of the rolled section. Manual of Steel Construction, Load and Resistance Fac-
Any understrength in the built-up section relative to tor Design, 2nd Edition.
the rolled section occurs near the other end of the
curves (Lr, Mr). American Institute of Steel Construction (AISC) (1999),
6. For all sections, any understrength in the built-up sec- Load and Resistance Factor Design Specification for
tion relative to the rolled section occurs at large values Structural Steel Buildings, December 27, 1999.
of Lb where the compressive flange strength will be American Welding Society (AWS) (2000), Structural Weld-
limited by either inelastic or elastic lateral-torsional ing Code – Steel, AWS D1.1-2000.
buckling. This implies that closer study may be war- MathCAD 8 Professional (1998), MathSoft, Inc.
ranted for industrial structures where flexural mem- Structural Stability Research Council (1998), Guide to Sta-
bers seldom have the continuous compression flange bility Design Criteria for Metal Structures, 5th Edition,
support of a concrete slab. edited by T. V. Galambos.

180 / ENGINEERING JOURNAL / FOURTH QUARTER / 2001

© 2003 by American Institute of Steel Construction, Inc. All rights reserved. This publication or any part thereof must not be reproduced in any form without the written permission of the publisher.

You might also like