10 Post-Tensioned Concrete - Principles and Practice

In the architectural design of the Watergate building, the floor plan was curved and columns were
located randomly in areas which substantially hid them, including walls, duct spaces, closets, etc. The
resulting column layout did not line up in either direction. No column was spaced any farther than about
22 feet from any other column; however, the concepts of gridlines, column strips, and middle strips were
meaningless. The structural designers of the slabs (a joint effort of T. Y. Lin & Associates and Atlas
Prestressing Corp.) were perplexed because, using conventional two-way slab techniques, there was no
obvious path for slab loads to columns. Someone in the team suggested connecting columns in one axis
of the building with imaginary straight lines between individual columns, and thinking of those lines as
a series of beams, or hard points. A “band” of tendons could be run along that line connecting columns
in one direction, then in the other direction tendons could be spaced uniformly over bands. With this
concept, the load path became obvious, and the forces and profiles for both the band tendons and the
uniform tendons could be easily calculated.

This tendon layout, with all of the post-tensioning tendons in one direction located in a narrow band
over columns, and tendons distributed uniformly with no regard for imaginary column strips and middle
strips, had never been done before. However, the Watergate design team saw no alternative and the
design and construction proceeded with the unique tendon layout. The performance of the slabs
appeared to be good, and the tendon installer reported a significant savings in tendon placing costs when
compared to the conventional “basket-weave” system. The primary labor savings resulted from the
elimination of tendon sequencing. In this new banded layout, all of the band tendons were placed first,
and then all of the uniform tendons. Ironworkers did not have to place individual series of tendons,
alternating in each direction, according to a complex sequence.

Typical Banded Tendon Layout

 K. Dirk Bondy & Bryan Allred 11

Since the Watergate Apartment building, built almost forty years ago, the banded tendon layout has
become the standard method for placing tendons in two-way post-tensioned slabs. The adequacy of the
banded tendon layout has been confirmed by the functional performance of hundreds of millions of
square feet in service, and numerous laboratory tests, starting with a landmark testing program at the
University of Texas at Austin in the early 1970s, supervised by the legendary professor and researcher
Dr. Ned H. Burns.

Four-Panel Test at University of Texas

Formation of the Post-Tensioning Institute

Engineers, contractors, and material fabricators in the post-tensioning industry recognized early that they
needed an industry organization to represent their interests and to disseminate design and construction
information relative to this specialized field. From the late 1960s through the mid 1970s the post-
tensioning industry was represented as a group within the Prestressed Concrete Institute (PCI), now the
Precast/Prestressed Concrete Institute. However, as the industry grew, it became apparent that a separate
organization, dedicated solely to post-tensioned concrete design and construction, was needed.

Pursuant to this, the Post-Tensioning Institute (PTI) was formed as an independent organization in 1976.
Now thirty years old, PTI provides all those with an interest in post-tensioned concrete a single unified
voice and source of design and construction information. Since its founding, PTI has been guided by
three extremely talented Executive Directors: Cliff Freyermuth, Gerry McGuire, and its current
Executive Director, Ted Neff.

PTI has matured and grown as the industry has grown. PTI now publishes a Journal with informative
articles about post-tensioning design and construction issues, and holds well-attended annual
engineering conferences. In many cases, PTI documents and committee reports establish the standard of
12 Post-Tensioned Concrete - Principles and Practice

care for design and construction of post-tensioned concrete structures. PTI is now recognized
internationally as the premiere source of information about post-tensioned structures worldwide.

Computers
No discussion of the history of post-tensioning in U.S. buildings would be complete without addressing
how it was influenced by computers.

When I graduated from UCLA in 1963, and started my first engineering job with T. Y. Lin and
Associates, my primary mathematical tool was a slide rule.

Slide Rule

My slide rule could multiply and divide, but it didn’t know where the decimal point was. I had to figure
that out myself. Now, 50 years later, I can do a dynamic analysis of a 40-story building, with all the
decimal points in the right place, on a flight between San Francisco and Los Angeles, on a laptop
computer about the size of a book…..while sipping a glass of Shiraz. Just kidding about the Shiraz; I
would never do that.

Just in one lifetime the changes in computing power have been astonishing. From slide rules to personal
computers the size of a notebook (with more power than the original mainframe computers that took up
an entire air-conditioned large room), the improvements in computing power are breathtaking!

Let’s take a look at my personal journey through this whirlwind of technology.

The Dawn of the Computer Age

My first contact with machine-assisted design came in 1966, when I was employed by Atlas Prestressing
Corp., and I was introduced to the Olivetti Programma 101.

The Olivetti was like a giant hand calculator. It was 19” wide, 24” deep, 7.5” tall, weighed 78 pounds,
and, for the first time, showed the potential for machine-assisted calculations. It sold in 1966 for about
$3,500. It could add, subtract, divide, multiply, and calculate a square root. It could record and execute a
limited number of program steps on plastic magnetic cards, and the output data was printed on a roll of
calculator paper, like an old “adding machine.” There was no programming language; communication
with the Olivetti was in machine language (enter a number into the X-register, arrow up into the Y-
register, divide Y by X, store the answer in register ZZ).
 K. Dirk Bondy & Bryan Allred 13

Olivetti Programma 101 (1966)

We would trim and tape the output to preprinted 8.5x11 calculation sheets with predetermined locations
for the input and output data. For a short time in the late 1960s scissors became a primary structural
design tool. When I was with Atlas I presented a series of one-day seminars on the design of post-
tensioned concrete to groups of practicing structural engineers throughout the country and in a few
foreign countries. At these seminars we provided the attendees with a design workbook whose cover is
shown below:

Seminar Workbook Cover

14 Post-Tensioned Concrete - Principles and Practice

One of the design examples I presented was for a post-tensioned beam, such as might be used in a cast-
in-place parking structure. First, I went through the hand calculations for the beam, part of which shows
the design for nominal strength:

Hand Calculations for Beam Strength Design

Next, I presented a sheet which showed the way we were actually doing this type of design at Atlas, a
preprinted sheet with input and output labels in prearranged locations. We had a series of recorded
“programs” on magnetic cards. For beam design we had four small programs: one which calculated
beam loads from input geometry and data, one which calculated section properties, one which
determined the prestressing force and flexural concrete stresses, and finally, one which calculated the
required amount of non-prestressed reinforcement.

We would take the output strips of calculator paper for each of the four parts of the design, trim them
with structural scissors, and tape them onto the preprinted calculation sheets in the proper locations. I
had the audacity to call this sheet a “Computer Example!” And I guess, in kind of a primitive way, it
was.
 K. Dirk Bondy & Bryan Allred 15

“Computer Example”

The Wang Era

The next landmark in the evolution of computer-aided design came with the introduction of the Wang
700 series of “super calculators.” The Wangs still used machine language, but they had much more
capacity than the Olivetti with many more registers and much more storage. Program steps were
recorded on audio cassette tapes, and output could be printed on large sheets with an IBM Selectric
typewriter. The Wang 700 machine is shown in the two figures below, the second showing it with
peripheral equipment (typewriter, auxiliary tape reader) attached:
16 Post-Tensioned Concrete - Principles and Practice

Wang 700 “Super Calculator”

Wang 700 With Peripheral Equipment Attached

With the Wang 700 series machines we could write a crude analysis program in a series of independent
routines, each small enough to fit into the machine’s storage capacity. When one routine was completed,
we would print the input and output, then delete as much of the data as we could to make room in
storage for the next routine. The answers, and some of the input from one routine, were used in the next.
It was tedious, but it was better than anything else that existed at the time, and it was certainly better
than doing the calculations by hand.

The Wang programs were based on analysis, not design. It simply did not have enough capacity to
perform a true design involving loops, decisions, and iteration to a final design. We would input a first
guess at geometry, force and profile, perform an analysis on the given configuration (calculate stresses
and reinforcement), and if we were not happy with the results we would change something and start
over, iterating manually to a final acceptable design. Moments and shears were calculated with 2-cycle
 K. Dirk Bondy & Bryan Allred 17

moment distribution, done either by hand or with a programmed routine on the Wang. The capacity of
the Wang did not permit the use of matrix techniques.

The First Personal Computer

Wang revolutionized the engineering world by introducing, in 1973, the first real personal computer, the
Wang 2200. It had its own programming language, a BASIC interpreter, and could perform loops and
mathematical decisions. It had a mighty 4 KB of random access memory (RAM); thus, it required some
tedious programming. As with its predecessor the 700, with the 2200 we would input some data, do
some calculations, print the answers with a Selectric typewriter, delete unnecessary data, input some
more data, do some more calcs…..and so on until we got an acceptable design.

Wang 2200

Program code on the 2200 was recorded (read only) on standard audio cassette tapes, and the output
printed on 8.5x11 sheets with the Selectric typewriter. For the first time, output could be printed in a
tabular format with rows and columns. The Wang 2200 created some great programmers – if you could
program this thing in BASIC with 4 KB of RAM you could program anything!

The IBM Personal Computer

The world changed forever in 1981 when IBM introduced the first IBM Personal Computer (the “PC”).
It had two floppy disk drives, each permitting both reading from and writing to the disk. Flexible
printing with virtually unlimited formatting could be done on a dot matrix printer. A large monitor made
input/output operation much easier. The first PC had 64 KB of RAM (luxurious, who could want more
than that?) and a dazzling processor speed of 4.77 MHz. There was no hard drive. It came with a
sophisticated “GW-BASIC” (nicknamed “gee-whiz”) language developed by Microsoft, which also
developed the disk operating system (DOS) which controlled access, storage, and all internal operations.
18 Post-Tensioned Concrete - Principles and Practice

The First IBM Personal Computer (1981)

In 1983, two years after the introduction of the first personal computer, IBM introduced the greatly
improved IBM XT. It was the first computer to have a built-in hard drive (a huge 10 MB). Along with
the hard drive, the XT had two floppy disk drives. RAM was doubled to 128 KB, and the processor
speed remained at 4.77 MHz.

IBM XT (1983)
 K. Dirk Bondy & Bryan Allred 19

The XT allowed vast improvements in programming. For the first time matrix methods became feasible,
and with the increased RAM we could do true design programming with decisions and iterations to a
final design. Much more sophisticated routines were possible, such as the one we developed for
punching shear in two-way slabs. Finally we didn’t have to spend most of our programming time
juggling and printing data.

The first version of our widely used post-tensioning design program PTData was developed on the IBM
PC and finalized for use on the XT. PTData was first made available to the engineering community in
1983 and eventually sold over 500 licenses. It was recently converted to Windows 64-bit and
[Link] by my son’s firm Seneca Software Solutions, Inc., which maintains and markets the
program now.

Windows
Perhaps the most significant landmark in computer history came in 1985 with the introduction of
Windows by Microsoft. Windows quickly replaced DOS as the prevalent PC operating system.
Microsoft also developed an improved BASIC programming language which it called Visual Basic.
third party programmers now had access to all the beautiful Windows interfaces and I/O routines.
Windows was originally a 16-bit system, but changed to 32-bit in 1994 requiring a major conversion for
16-bit programs. In 2009, Windows changed again to a 64-bit system and introduced [Link], an
advanced BASIC programming language.

With the advent of Windows came huge improvements in memory, processor speed, and disk storage.
RAM has increased from 64 KB on the first PC to a typical 4 GB now, an increase of more than 60,000
times. Processor speed has increased from 4.77 MHz to 1.4 GHz, an increase of about 3,000, and now
with 1+ terabyte hard drives easily available and inexpensive, disk storage has increased about 100,000
times since the first 10 MB hard drive on the XT.

Post-Tensioning Design Programs

In the early 1970s the market for post-tensioning in buildings was rapidly growing. One major reason
for this growth was the dissemination by Atlas of design information to practicing structural engineering
firms through seminars and design assistance to those firms. More engineers were learning how to
design post-tensioned buildings and, as a result, more post-tensioned buildings were being built. To
handle this volume of engineering work, Atlas grew and maintained a large staff of licensed engineers
(more than 100) in their various offices throughout the country.

With the availability of a true personal computer (the Wang 2200), and having for the first time the
capacity to run a real post-tensioned design program, Atlas decided that it was time to develop one. The
conversion of our Wang-based machine language routines to a comprehensive design program written in
BASIC looked like a major undertaking. Through PTI committee work I had become friends with
Merrill Walstad, the chief structural engineer for the VSL Corporation, a competitor active in post-
tensioned concrete building work. Atlas and VSL made the decision to develop the program jointly, with
Atlas providing most of the programming work (we had full-time programmers on staff) and VSL
providing funding and support. Surprisingly, this unlikely relationship worked, and in a period of about
six months the program was functional and both Atlas and VSL began to use it in-house for their design
services.

Both Atlas and VSL made the decision NOT to market the program commercially, but rather to limit its
use to in-house proprietary design services. However, Merrill left VSL and I left Atlas at about the same
20 Post-Tensioned Concrete - Principles and Practice

time in 1976 to start our own firms, Merrill a structural design firm (Walstad Engineering, Inc.) and me,
with my partner Chris Deetz, a construction firm and a structural design firm, both specializing in post-
tensioned concrete buildings (Seneca Construction Systems, Inc. and Seneca Structural Design, Inc.)
Merrill and I continued the relationship we had developed at Atlas and VSL and worked together on
several projects. We of course had the source code for the Wang 2200 program we developed at
Atlas/VSL, and we recreated the program and used it for our own design work for about five years. In
the early 1980s, when the first IBM PC was introduced, Merrill and I decided it was time to develop a
state-of-the-art post-tensioning design program which would run on the PC and which we would make
available to the general engineering public. To accomplish this, we formed a corporation called
Structural Data, Inc. (SDI) and made the decision to call the new program “PTData.”

Along with Merrill and me, an original SDI partner was Nick Watry, a close friend of both of ours, who
was running a growing structural design firm in the San Francisco Bay Area (Watry Design Group) and
who had an interest in the development of the program. As work on the program progressed, Nick’s
interest in SDI faded and we bought him out. I think Nick decided it was in his best interest to keep his
post-tensioning expertise in-house and not make it available, through the program, to potential
competitors. The buyout was amicable and Nick has remained a great friend and colleague throughout
the years.

PTData was completed and first offered for sale in 1983. The original version was written in GW-Basic
based on the DOS operating system. Merrill developed the original I/O routines and I did all the other
technical programming. The program was converted to 16-bit Windows in 1985 and then to the 32-bit
version in 1994. Merrill ran all other aspects of SDI business, sales, accounting, etc., and we jointly
handled the technical support. It was an extremely successful venture and we eventually sold about 500
licenses.

PTData was the first commercially available computer program sold for use in the licensee’s office.
However a program called POSTEN was previously available (since 1971) but it was “rented,” i.e. the
customer filled out data forms, submitted them to the POSTEN office, which ran the program on a
mainframe and returned the output to the customer. As powerful personal computers became available,
the concept of a “rented” program became less appealing to structural firms, and the use of POSTEN
decreased dramatically. In 1991 a licensed version of POSTEN was made available, but by that time it
was too late to compete with the other established programs which could be run in-house on the
licensee’s computers.

A few years after PTData began to be sold, another competitive program called “ADAPT” was
developed and marketed. From that time to the present day, a period of almost 30 years now, PTData
and ADAPT have dominated the market for post-tensioned concrete design software. The last several
years have also seen new software developers entering the market, mostly with very complex, three-
dimensional finite element analysis programs.

In his design classes at UCLA my son Dirk effectively integrates PTData computer calculations with
hand calculations, and the combination results in a very powerful learning tool for students, who today
are highly computer literate. Examples of the use of the computer as a learning tool can be found in the
design examples in this book, starting with Chapter 7.

In the late 2000s Windows converted to a 64-bit system and Microsoft introduced a new version of its
BASIC programming language, [Link]. Our 32-bit Windows version of PTData would not run
in the 64-bit environment, and faced with another major conversion and lured by retirement, Merrill and
I decided to end our long run and hand our interests in the program over to Dirk’s structural engineering
 K. Dirk Bondy & Bryan Allred 21

office. Dirk formed a new company, Seneca Software Solutions, to market the program. He successfully
converted the program to the Windows 64-bit environment and it is up and running on Windows 7!

Problems Faced by the Post-Tensioning Industry

Restraint to concrete volume change (shortening) was the first big pervasive problem faced by the
industry. The mechanics of volume change, and the restraint to that volume change, is different in post-
tensioned concrete members than in non-prestressed members. In non-prestressed concrete beams and
slabs restraint to shortening provided by connected elements (walls and columns) results in many closely
and uniformly spaced cracks throughout the length of the member. The ends of a non-prestressed
member tend to stay in their original positions, and the total shortening is simply the sum of the crack
widths along its length. Because of this, shears and moments induced into restraining connected
elements are relatively small.

In post-tensioned concrete members, however, the effect of the axial prestress force tends to close most
of the restraint-to-shortening cracks which would otherwise form between the ends of the member.
Unlike non-prestressed members, the total volume change along the length of the post-tensioned
concrete member is reflected in significant movement inwards at the ends. This induces large shears and
moments into the connected walls and columns. These shears and moments can result in large and
unsightly cracks in the post-tensioned member, and in the walls and columns themselves.

Engineers had to learn how to design post-tensioned concrete floor systems with levels of cracking
normally accepted in non-prestressed floor systems. This was accomplished over the years largely with the
use of joinery details (slip joints and pour strips) and the use of properly located and sized non-prestressed
reinforcement. Mitigation of restraint-to-shortening effects in the design of post-tensioned buildings has
become as large a part of the design process as the selection of the forces and profiles themselves.

But without doubt the biggest problem ever faced by the industry was tendon corrosion. The early
unbonded tendon sheathings and coatings (grease) were inadequate for aggressive environments, such as
those where de-icing salts are applied to exposed concrete surfaces. Serious corrosion problems began to
be apparent in such buildings within about 10 years of service. Most were repairable, and several
companies thrived by specializing in corrosion-related repairs.

Improved material specifications developed and enforced by PTI through certification and informational
programs have largely solved these corrosion problems. These include improvements in the quality and
performance of sheathing material, coatings, and the development of a complete tendon encapsulation
system for use in the most severe environments.

Summary
My career in post-tensioned concrete has been rewarding in every way. It was technically challenging,
always exciting, and it brought me into contact with some of the most interesting and talented people in
the construction industry. I am still dazzled by the power of post-tensioned concrete, and its inherent
economy. By “balancing” applied downward loads, post-tensioning eliminates a significant amount of
bending, and bending, because it does not use the full member cross section, is an extremely inefficient
way to carry a load from one place to another.

I never tired of working in the field of post-tensioned concrete. Every day was fun. Some days were
more frustrating than others, but it was always fun. Now, 50 years later, in these words, I have tried to
look at the history of this dynamic industry, as I saw it firsthand. For me, it was a great ride.
 2
Design Philosophy, Equivalent Loads & Concrete Stresses
on Determinate Systems

All references in this book to ACI 318 refer to the 2014 edition. All references to ASCE 7 are to the
2010 edition.

Working Stress versus Nominal Strength Design Methods

In the early days of structural engineering virtually all materials were designed using what was called
the “Working Stress” method. Highlights of the Working Stress Method are as follows:

Under “working” or “service” loads the material stresses are limited to a level that promotes
elastic behavior.
Factors of safety are inherent in the allowable stresses (for instance, 0.66F y for stable steel
beams) but the factors of safety are not explicitly stated and typically are not easy to
determine.
One benefit is that all calculations are performed using the same loading (stress checks,
deflections, etc.).
A significant drawback is that engineers cannot determine the true maximum (ultimate) strength.

Eventually, the American Concrete Institute (ACI) blazed a new trail, as they often have, and developed
the “Nominal Strength” method of design. Highlights of the Nominal Strength Method are as follows:

The actual maximum capacity of a member is calculated (the nominal strength).

A portion of the factor of safety is applied to the loading used in the elastic analysis (the load
factors, such as 1.2 x Dead Loads & 1.6 x Live Loads). This part of the factor of safety is
strictly a function of the loading and applies equally to all elements and modes of failure
(beam shear or column axial load, etc.).
The remainder of the factor of safety is applied through a “capacity reduction factor”, φ. The
capacity reduction factor varies according to the member and the mode of failure (i.e., φ = 0.9 for
beam flexure & 0.75 for beam shear) and is intended in part to promote more ductile (rather than
brittle) failure modes. It is also intended to account for variations in material properties.

Prestressed concrete, probably because it is so different than any other material or system in that it
actively supports load (as opposed to virtually all other materials which passively resist loading), has
always been (unfairly?) required to satisfy both working stress and ultimate strength design criteria in
the ACI codes. In this book we will first address the working (service) stress design requirements, and
then we will tackle the ultimate strength portion. As we work through this it should become clear that
this is also the proper design sequence used in the everyday design of prestressed concrete members.

23
24 Post-Tensioned Concrete - Principles and Practice

The working stress portion of the design will determine the prestressing force and profile, which will
become a “given” in the ultimate strength part of the design.

Pretensioned versus Post-Tensioned Prestressed Concrete

Prestressed concrete is the umbrella term that describes both pre-tensioned and post-tensioned concrete.
The “pre“ and “post” prefixes describe when the prestressing tendons are stressed (pulled or jacked)
relative to the concrete being placed and cured.

In pre-tensioned concrete the tendons are typically placed in steel bed forms and stressed prior to the
concrete being placed. The prestressing steel is usually straight, or “harped” in straight segments. These
tendons are bare and bond directly to the concrete. Once the concrete is cured the tendons are released
from the bed and the tension force is transferred through bond into the concrete, compressing the
concrete. Pretensioned concrete is almost always associated with precast concrete, or concrete cast in
beds within a precasting plant. The most typical pre-tensioned structural members are single tees, double
tees, beams, girders and planks.

In post-tensioned concrete the tendons are typically draped in a parabolic profile in the concrete forms,
unstressed. These tendons are coated in grease within plastic sheathing so that they won’t bond to the
concrete and can slide within the sheathing when they are stressed after the concrete had been placed
and hardened. Post-tensioned concrete is almost always associated with cast-in-place concrete, or
concrete cast at the actual site in its final location. The most typical post-tensioned structural members
are one-way slabs, two-way slabs, beams and girders.

One-way Slab Tendons Over a Post-tensioned Beam

This book will primarily focus on post-tensioned concrete, though for academic purposes pre-tensioned
concrete will be discussed occasionally.
 K. Dirk Bondy & Bryan Allred 25

Review of Section Properties

Before we jump in to equivalent loads let’s do a quick review of section properties. All of the stress
calculations that we do in this book will depend upon the elastic section properties of the members, so
it’s important to start off on the right foot.

Rectangular Beam Example:

c.g.c. =
I=
yt = y b =
& S=

∴ St = Sb = =
( )

In general: I = Io + Ad2 & S = I/y

Tee-Beam Example:

Area, A = (48 in)(6 in) + 30 in(14 in) = 708 in2

48 in(6 in)(6 in 2) + 30 in(14 in)(6 in + 30 in 2)

c. g. c. =
708 in

= 13.67 in from the top

∴ yt = 13.67 in ; yb = 36 in – 13.67 in = 22.33 in

I = 48 in(6 in)3/12 + 14 in(30 in)3/12 + 48 in(6 in)(13.67 in – 3 in)2 + 14 in(30 in)[13.67 in – (15 in+6
in)]2
= 87,719 in4
∴St = I/yt = 87,719 in4 / 13.67 in = 6417 in3
& Sb = I/yb = 87,719 in4 / 22.33 in = 3928 in3
26 Post-Tensioned Concrete - Principles and Practice

Equivalent Loads
Before we begin with equivalent loads we should first address the convention for drawing the moment
diagrams. Virtually all textbooks draw moment diagrams to the compression side (positive up and
negative down for horizontal members). However, most consulting structural engineering firms that
specialize in any type of concrete will reverse this convention and teach their young engineers to draw
moment diagrams to the tension side. One reason for this is that tensile reinforcing steel is obviously
placed where the concrete is in tension. It is much less likely that a young engineer will make the
mistake of drawing the reinforcement on the wrong side of the beam, slab or wall if the moment diagram
clearly indicates where the reinforcement should be placed. Another reason is that the tension side
moment diagram will approximate the deflected shape, which helps a young engineer develop a more
intuitive feel for bending behavior. In post-tensioned concrete it is even more important that the tension
side be clearly depicted in the moment diagrams because the controlling service level stresses are always
the tensile stresses in a proper design. This is one of a number of instances where the practicing world
and academic world have clear differences of opinion. In this book all moment diagrams will be drawn
to the tension side.

Definitions:

Drape, “a”: The maximum distance from the c.g.s. to a line connecting the support c.g.s. locations. This
is sometimes referred to as the “sag” of the tendons. Drape has nothing to do with the c.g.c. The drape
will be used to determine the equivalent loads in the span.

Eccentricity, “e”: The distance at any point in the span between the c.g.s. and the c.g.c. Whereas “a”
occurs at a single location in the span, the value of “e” can vary and can be determined at any location in
a span. The eccentricity is used to determine the primary prestressing moment at a discrete location,
which is defined as the force times the eccentricity.

Strand: Prestressing steel reinforcement consisting of 7 high-strength wires.

Tendon: (From ACI) In pretensioned applications, the tendon is the prestressing steel. In post-tensioned
applications the tendon is a complete assembly consisting of anchorages, prestressing steel, and
sheathing with coating for unbonded applications or ducts with grout for bonded applications.

c.g.s.: The center of gravity of the reinforcing steel or tendon group.

c.g.c.: The center of gravity of the concrete section.

F: The effective force in the tendon group, typically in kips.

M = Fe in both statically determinate and indeterminate systems. The primary prestress

moment will also be the equivalent moment in statically determinate systems.
 K. Dirk Bondy & Bryan Allred 27

Consider a tendon cast in a weightless but stiff material. Let’s assume also that the tendon has no
inherent stiffness (consider the tendon as a string or rubber band). The tendon is pulled (stressed) to a
tension force, F, and anchored at the end of the beam. The beam in this case is statically determinate. In
this case “a” also equals “e” at midspan.

Note that there is no external load on the system (beam). Now pull out the tendon from the system and
look at the forces acting on it. Notice that it wants to straighten, but the material resists that.

For this tendon with no stiffness to have this shape (2 straight portions) a load “P” must be acting as
shown. There is no other load or location that will result in this same shape. The load “P” is applied by
the surrounding material. For equilibrium, the ends of the tendon must each have a vertical component
of P/2.

Now look at the material only. Remember that there are no external loads on this weightless beam.

Here’s the point. When evaluated separately, the tendons and material have loads acting on them in
equilibrium. But when they are added together there is no net force.
28 Post-Tensioned Concrete - Principles and Practice

Let’s look at a free-body diagram of the tendon again and determine the load “P” acting on it.

Using small angle theory,

a
F =0 2F P=0
L
2
4Fa
P=
L

Another way to look at this (also utilizing small angle theory) is:

Summing moments about “O” yields:

L P L
Fa P + =0
2 2 2

PL 4Fa
Fa = or P =
4 L

Let’s look at the material again.

The equivalent moment being applied to the material is:

P L
M =
4
exactly offsetting the tendon moment. The “equiv” is short for “equivalent”, and we will begin referring
to these vertical loads in the span as “equivalent loads” or sometimes “balanced loads” because later we
 K. Dirk Bondy & Bryan Allred 29

will be “balancing” the applied loading. The material, since it has stiffness, resists this equivalent load
by creating internal stresses. Note that this is also under compression from “F”, therefore the total stress
in the material at midspan is (tension is positive, compression is negative):

M F
@ Top =
S A

M F
@ Bottom =
S A

Summary

Note: Mequiv equals F

multiplied by e, the
eccentricity of the c.g.s to
the c.g.c., at any point

P L
=
48EI
30 Post-Tensioned Concrete - Principles and Practice

Consider the same beam with a parabolic tendon profile.

Now pull out the tendon and analyze it alone. Once again, a tendon under tension wants to straighten,
but this is resisted by the stiffness of the material.

Think of this as a clothesline. An evenly distributed load (of clothes, say) creates a second order
parabolic shape, referred to as a catenary. Therefore, if a tendon is to retain its parabolic shape, it must
“feel” uniform loading acting upon it. No other load will create this same shape.

Now cut a free-body diagram at the centerline:

Sum moments about “O”:

L L
Fa w =0
2 4

w L
Fa =
8

8Fa
w =
L