CHAPTER 26-PRESTRESSED CONCRETE

With the exception of shear, bond, and the addition of c/J to the ultimate ..
strength equations, this chapter is similar to the report of ACI-ASCE Commit-
tee 423 (formerly 323), Prestressed Concrete.*

2602-Scope
The entire Code applies to prestressed concrete except where excluded or in
direct conflict with Chapter 26.
The following is a discussion of why some sections of the main code are
excluded from use in the design of prestressed concrete:
Section 906-Requirement for T-beams- These empirical provisions devel-
oped for conventional reinforced concrete would exclude many standard pre-
stressed products in satisfactory use today.
By excluding Section 906, no provisions for prestressed concrete T -beams
appear in the Code. This does not mean that prestressed concrete T-beams can-
not be designed. Instead it leaves the problem of defining the width of flange to
the judgment of the engineer. Where possible, the flange widths in Section 906
should be followed unless experience has proven otherwise. It is not necessarily
conservative in elastic design consideration to use the maximum flange width as
recommended in Section 906.
Section 911-Minimum reinforcement of flexural members- The provisions
for minimum steel in prestressed concrete are in Section 2609(c).
Section 913-Limitsfor reinforcement of columns- The empirical limitations
such as minimum bar size obviously do not apply to prestressed concrete.
Chapters 13 and 18- Bond and anchorage- The subject of bond of the pre-
stressing steel is included in Section 2611. However, the provisions of Chapters
13 and 18 are applicable for bond and anchorage of any auxiliary reinforcement
that may be used with the prestressing steel.
Chapter 14-Reinforced concrete columns- Working stress design- The ma-
terial is not applicable to prestressed column design where both working stress
..
and ultimate strength investigation must be made. No provisions are given for
design of prestressed columns, but this does not exclude their design by some
rational method.
Section 1508-Control of cracking- This section is obviously applicable to
conventional reinforced concrete only.
*ACI-ASCE Committee 323, "Tentative Recommendations for Prestressed Concrete," ACI Journal, Pro-
ceedings V. 54, No.7, Jan. 1958, pp. 545--578.
74
 PRESTRESSED CONCRETE 75

Section 2001(a)-Concrete joist floor construction- The empirical limits of

this section are justified for conventional reinforced concrete but not for pre-
stressed concrete.
Chapter 21-Fiat slabs with square or rectangular panels- The design of pre-
stressed concrete slabs requires recognition of secondary moments induced by
the undulating profile of the prestressing steel. Also volume changes due to the
prestressing force can create additional loads on the structure that are not ade-
qtiately covered in this chapter. Because of these unique properties associated
with the act of prestressing, many of the design procedures of Chapter 21 are
not appropriate for prestressed concrete structures.
'- Section 2504(b)-Fiexural design-Ultimate strength design- These restric-
tions do not apply to prestressed concrete where the "excessive tensile cracking"
can be controlled by the amount and location of the prestressing force.

2603-General consideration
As has been past practice in the design of prestressed concrete, the design
investigation should include all load stages that may be significant. The three
major stages are: (I) initial stage-when the tensile force in the prestressed
steel is transferred to the young concrete; when stress levels are high relative
to cylinder strength, (2) working load stage-after long-time volume changes
have occurred, and (3) the ultimate capacity load stage. There may be other
load stages that require investigation. For example, if the cracking load is signif-
icant, this load stage may require study.
From the standpoint of satisfactory behavior, the two stages of most impor-
tance are ultimate capacity and working stress (or service load). It is necessary
to investigate both of these stages because of the nature of prestressed concrete.
For example, a beam could be prestressed along its longitudinal axis in such a
manner that it will support the service loads without objectionable deflection
but the ultimate capacity could be well below that required for an adequate
safety factor. Similarly, a design based on ultimate capacity alone may provide
unsatisfactory behavior, at service loads, i.e. excessive camber or deflection.

2605-AIIowable stresses in concrete

Working stress design is based on a system of allowable stresses similar to
conventional reinforced concrete design.
Allowable stresses for concrete are given for the load stage that occurs at
time of transfer of the prestress force to the concrete and for the load stage
existing after the reduction in the prestress force from the various losses listed
t
in Section 2607.
Section 2605(a)- The temporary allowable concrete stresses are applicable
immediately after transfer of the prestressing force but prior to the occurrence
of losses due to creep and shrinkage. In other words, the concrete stresses at
this stage should not exceed the recommended values caused by the force in the
steel at transfer reduced by the losses due to some relaxation of the steel, elas-
tic shortening of the concrete, plus the stresses due to self weight of the member.
76 COMMENTARY-BUILDING CODE

Shrinkage loss is not included at this stage because if the member has been
properly cured in a moist atmosphere, or if moisture loss has been stopped by a
vapor barrier curing blanket, little shrinkage has taken place. Elastic shortening
is included because as soon as the force in the tensioned steel is transferred to
the concrete, clastic shortening does take place in a pretensioned member.
The maximum allowable compressive force at the initial stage is the same as
that reported by ACI-ASCE Committee 423, except that no differentiation is
made between pretcnsioned and post-tensioned clements.
The allowable concrete tensile stress, anywhere in the member, at the initial
stage is 3VJ:{ unless auxiliary reinforcement is used in the tensile stress zones.
The auxiliary steel is proportioned on the assumption of an uncracked section.
In other words, the stress diagram due to initial prestress and the weight of
•
the member is drawn, and enough steel area is provided to carry all the tension.
The allowable stress in this auxiliary steel is not specified but it would be con-
servative to usc those given in Section 1003 since the effects of creep and shrink-
age begin to reduce the tensile stress almost immediately.
Section 2605(b)- The limits for allowable compressive and tensile stresses
after losses have occurred arc similar to those reported by ACI-ASCE Commit-
tee 423. The wording concerning tensile stresses has been changed slightly, to
state more clearly that tensile stresses arc allowed unless the conditions of ex-
posure arc such that the cracking would expose the prestressing steel to a
corrosive atmosphere.
The statement that the upper limit of 6-vf[ may be exceeded when not detri-
mental to proper behavior is a necessary one. There are many designs where this
tensile stress is exceeded and where deflection is within the desirable limits for
the specific structure. In fact, it may be necessary to exceed this limit for satis-
factory behavior. For example, in the case of a multiple bay roof member, de-
sign for full snow .load may result in camber that creates drainage problems
because the full design load is not always present. If a greater tension will allow
control of this camber, even though the member may flexurally crack under the
full snow load, it should be considered. Each case should be considered on the
basis of its particular loads and framing plan. The tension should, of course,
not take the member into an inelastic range, nor should it allow a situation
where water can pool on the deflected roof in ever increasing depths as deflec-
tion occurs. Thus the matter is one that properly should be decided by the engi-
neer on the basis of the details of the actual structure. ...
The entire matter of allowable stresses is being seriously studied by several
committees at the present time. The values given in this Code are excellent
guides, and arc satisfactory for most members. However, there is no reason why
they cannot be exceeded in specific cases provided that a thorough evaluation of
the resulting service load behavior is made by the engineer. Section I 04 provides
the procedure by which these stress values can be exceeded.

2606-AIIowable stresses in steel

The allowable steel stresses are the same as those in current use.
 PRESTRESSED CONCRETE 77

2607-Loss of prestress
The causes for loss of prestress arc listed. For an explanation of how to
compute these losses, see the report of ACI-ASCE Committee 423.
The lump sum losses of 35,000 psi for pretensioning and 25,000 psi for post-
tensioning that appeared in the report, are not given in the Code, but can be
used. Actual losses greater, or smaller, than the lump sum values have little
effect on the ultimate strength of the member, but affect service load behavior
such as deflection, camber, or cracking load. Overestimation of losses of pre-
stress can be almost as bad as underestimation, since the former would result in
excess camber and horizontal movement.

2608-Uitimate flexural strength design

As mentioned previously, the nature of prestressed concrete requires an
investigation at ultimate capacity to determine the safety factor, or ratio of
ultimate load to service load.
The equations are the same as those in the report of ACI-ASCE Committee
423, except for the addition of the capacity reduction factor c/J. Develop-
ment of the equations and a full explanation of these equations is contained in
a paper by Warwaruk, Sozen, and Siess.* The methods and procedures shown
in this paper are usable for the present Code when the ultimate moment is re-
duced by the capacity reduction factor c/J. The introduction of the c/J-factor
creates no difficulties in the process of determining the ultimate capacity of a
member that has been proportioned on the basis of allowable stresses. The re-
quired ultimate moment can be calculated by multiplying the estimated dead
load moment and live load moments by the load factors. If this required mo-
ment is then increased by dividing by l/J the calculation of the required area of
steel and its location remains straightforward.
Eq. (26-6) is an approximate expression for steel stress. It may be advanta-
geous to calculate the steel stress by a more accurate method such as that pre-
sented in the paper by Warwaruk, Sozen, and Siess.*
For members where the prestressing steel is unbonded, Eq. (26-7) severely
restricts the stress in the steel at ultimate strength. This restriction results from
the difficulties of determining the actual steel stress in an unbonded member. lt
should be noted that this question pertains to the cases where information for
the determination of ]su is not available. When rigorous methods are feasible,
or when test data is available the more precise values for ]su may be used.

2609-Limitations on steel percentage

The limitations on the reinforcing steel index, p ]su/f/ = 0.3, was originally
set by ACI-ASCE Committee 423, Prestressed Concrete, as the dividing line be-
tween under-reinforced and over-reinforced members. It is based on the fact
that when the reinforcing index exceeds 0.3, the ultimate capacity as predicted
by Eq. (26-4) (without the c/J) does not correlate well with test results.
•warwaruk, Joseph; Sozen, Mete A.; and Siess, Chester P., "Investigation ol Prestressed Reinforced Con-
crete lor Highway Bridges. Part 3-Strength and Behavior in Flexure ol Prestressed Concrete Beams," Bulletin
No. 464, Engineering Experiment Station, University ol Illinois, Urbana, 1962, 105 pp.
78 COMMENTARY-BUILDING CODE

No provision for minimum amount of tensile steel appeared in previous de-

sign recommendations in this country. The provision is a precaution against
abrupt flexural failure resulting from rupture of the prestressing steel when
ultimate capacity is reached immediately after cracking. The usual member re-
quires considerable additional load beyond cracking to reach ultimate capacity.
Thus, considerable deflection warns that the ultimate capacity is being ap-
proached. However, if ultimate capacity occurs shortly after cracking the
warning deflection may not occur. .~

261 O-Shea r*
These provisions arc based upon a critical appraisal of 244 bonded pre-
stressed beams which failed in shear, including both monolithic and composite
sections up to 39 in. in width and 25-1/2 in. in depth. The equations in Section
2610 have also predicted satisfactorily the ultimate strength in shear of thirty
five 9 x 18 in. prestressed [-beams with web reinforcement.
A very limited number of prestressed concrete slabs have been tested in
shear. The provisions of Section 2610 are not intended for the design of slabs.
The shear provisions arc adequate to include draped or depressed prestress-
ing steel and partial prestressing.
Two equations arc presented for proportioning web steel. Eq. (26-11) defines
the minimum area of web steel unless members meet the conditions of Section
2610(c). Eq. (26-10) defines the amount of steel required by shear, and provides
that stirrups, stressed to the yield point, must carry the shear above V...
The equation for minimum shear reinforcement was revised because in the
equation of ACI-ASCE Committee 423 ( 1958) more steel was required as the
web width b' increased. In members with multiple webs and subject to uniform
loads, the term h' shall be taken as the sum of the widths of all webs of the
member. Eq. (26-11) can be simplified by multiplying top and bottom of the
right hand side by bd. If j/ = 250,000 and],. = 40,000, the expression for Av
becomes:
A,. = 0.08pbs ff,"
Eq. (26-10) may be derived as follows: The ultimate shear that can be su-
stained by a beam at any cross-section is the sum of the amount of shear su-
stained by the concrete plus the amount sustained by the web reinforcement,
i.e. V., = ,,, ( V,. + V' ). where V., is the shear force due to specified ultimate
load, V,. is the shear carried by concrete, V' is the shear carried by web reinforce-
ment, and 1/> is the strength reduction coefficient = 0.85. V' is given empirically
by the expression: V' = A,,f,. djs which represents the vertical force carried by
the stirrups, A,djs, within a length d, at the yield stress fv· Fig. 16 illustrates
the agreement between this expression and experimental data. Then

V tl = ,,, (v c AvJ,.d)
-1---.\'- or A,,=
( V" - ,,, V,.)s
1/>f,.d .

which is Eq. (26-10).

*See also: Mast, Paul, "Short Cuts for the Shear Analysis of Standard Prestressed Concrete Members,"
Journal. Prestressed Concrete Institute, V. 9, No. S, Oct. 1964, pp. IS· 47.
 PRESTRESSED CONCRETE 79

IOOOr----.---.------.--.----.,-----,

900

600

700

600
~
':-. 500
>

... .. ..
400

300

200

100
•"'
......•
~,.
.
0 100 200 300 400 500 600
Fig. 16-lncrease in shear strength of a Ay
prestressed member due to web reinforcement ;bl fy

To solve Eq. (26-10), a value for Vc must be determined. This requires an

investigation into the two types of shear cracking that can occur: Type I is re-
lated to flexural cracking, Type 2 appears in portions of the member not flcx-
urally cracked. Fig. 17 illustrates these two types of cracking and the general
areas where they can occur.
Eq. (26-12) determines V<'i for Type I cracking. Eq. (26-13) determines Vcw
for Type 2 cracking. The smaller governs the design and is used for Vc in Eq.
(26-10).
To reduce the capacity of a beam, a diagonal crack must have a projection
on the longitudinal axis of the beam equal to the depth of the beam. Consider
a given Cross Section B-B (Fig. 18). A flexural crack distance d away (in the
direction of decreasing moment) may lead to a diagonal crack which could be
critical for Section B-B. The principal tensile stresses along the path of the
incipient diagonal crack will be increased by flexural cracking within distance
d. The principal tensile stress triggering the diagonal crack occurs near the
centroid of the beam. A flexure crack occurring at a distance d/2 from Section
B-B is a sign indicating the imminence of a Type I crack.
The terms·
Mer V
M d + d
v 2
in Eq. (26-12) are the shear corresponding to the formation of this crack .

Flexure
) cracks
Support' Support
Fig. 17-Types of diagonal cracks occurring in a prestressed concrete member
80 COMMENTARY-BUILDING CODE

Diagonal shear crack A s'

Flexural cracks

Fig. 18-Theory for Eq. ( 26-12) for V.;

To derive Eq. (26-12) consider Section B-B (Fig. 18), where, due to ex-
ternally applied loads, the moment* is M and the shear is V. The moment at
Section A-A is Men with the corresponding shear Ver· Since the change in
moment from one cross section to another is equal to the area of the shear
diagram between the sections
M -M rr -_ V+Vcr!{
2 2
In a typical prestressed concrete member the difference between V and Ver
over the distance d/2 will be small. Hence:

M- Mer= V ~
M
or
v
and,
Mer
v M d
v 2
Mer being the moment due to applied loads when flexural cracking occurs,
the middle term of the equation for V,.; is the shear due to applied loads when
flexural cracking occurs, even though expressed in terms of moment.

Mer = ~ (6V]7 + fpr- ]d)

where 6v']7 = modulus of rupture of the concrete, ] 1,. = compressive stress in
concrete due to prestress, and]d = stress due to dead load.
The total shear Ver. due to both applied loads and dead loads when the
critical flexural crack occurs is:

MJV- d/2 + Vd
Dead load shear, Vd, is considered separately for two reasons: 11

I. Dead load is usually uniformly distributed, whereas live loads can have
any distribution. ·
2. The dead load effect is always computed for the prestressed section alone.
The live load effect is computed for the composite section in composite
construction.
• A footnote in th~ Code refers to M and V us caused by "external ultimate loads." Their rutio, M IV, how-
ever, is not affected by loud factors.
 PRESTRESSED CONCRETE 81

The shear Vc; which causes the Type I diagonal crack has been found to be
equal to the shear corresponding to the formation of the flexure crack at a
distance d/2 from the section under consideration, plus a shear which appears
to be a function of the dimensions of the cross section and the strength of the
concrete. This increment of shear is accounted for by the term 0.6b' dv]7.
The total shear at the formation of a Type I diagonal crack is therefore

V,; = 0.6b' dv]7 + M M,~ + Vd

v 2
which is Eq. (26-12)
Fig. 19 shows the close agreement between Eq. (26-12) and available experi-
mental data.
A lower limit of 1.7b' dv]7 is proposed for Vci because the only test beams
in which V,; fell below this lirnit were those having an abnormally low degree of
prestress.
Type 2 diagonal cracking occurs when the maximum principal tensile stress
becomes equal to the tensile strength of the concrete. The principal stresses can
be calculated using the classical elastic theory equations. The expression for
Vc ... presented in the Code by Eq. (26-13) is a simplification of the classical
principal stress formula.
V,w is the shear in the nonflexurally cracked member at the time that diagonal
cracking occurs in the beam web. The difference between Vcw and V" is the shear
9~---r----.----.----.---~.---·
•
8
7 ••••••
• • ••
• •••.........
• •
•
• •• • • • •
• •

0 4 5 6
Mer/(~-~) b'd~
Fig. 19-Diagonal cracking in those regions of beams previously cracked in flexure
82 COMMENTARY-BUILDING CODE

to be carried by stirrups. Tests have shown that the maximum principal tensile
stress will usually occur near the centroid of the cross section.
The capacity of the member is reached if:

ft = ~ Vcr/ + (12Y - ~c
where f, = tensile strength of concrete, ,,,..,. = shear stress, and j,,.. com pres-
sive stress due to prestress. This relation yields:

or,
(!, + f2J v,.,..
2

(r
+ 2
fpc

Vcw = f, ~I +fpc
f,
A value for f, of 4V'j'/ appears to be substantiated by tests, but since v""'
is the nominal shear stress and not the actual maximal one, f, = 3.5-v:j/ is used
in the expression for Vcw:

Vcw = 3.5yj/ (~I+ 3 .5 0-J:)

The curve representing this equation is plotted on Fig. 20, as a solid line.
The equation may be simplified to the form:
Vcw = 3.5yl'jj + 0.3fpc
which is shown by the dashed lines.
10~--~--~--~--~--~--~----~--~~

9
8 Yew= 3.5./f~ [vf+ fpc ]
3.5~

Yew
7
6
5
-- --
~4
3
2 ...

0 2 4 6 8 fpc 10 12 14 16 18
~
Fig. 20-Relationship between nominal shear stress at Type 2 cracking and compressive
stress at centroid
 PRESTRESSED CONCRETE 83

The nominal shear stress,''""·· equals the shear force due to loads, v..,.,
minus
the counteracting component of the prestressing force, V", divided by b'd:
V,...,- Vp
b'd
Substituting this expression for I'""'

Vcwb'd
- Vp =
3•5 V- 'J!'
}c + 0 •3 jpc
{

and
" Vcw = b'd (3.5V}7 + 0.3jpc) + Vp,
which is Eq. (26-13).
In Eq. (26-13), dis taken as 80 percent of the total beam depth or as the
value used in Eq. (26-12) whichever is greater. This is valid because the section is
uncracked. Thus the web area available to compute nominal shear stress is not
directly a function of the location of the centroid of the prestressing steel. The
value of 80 percent is substantiated by test data which showed that 80-85 per-
cent of the total depth yielded reasonable results for this type of analysis.
Lightweight aggregate concrete- The equations for shear in lightweight
aggregate prcstrc!!sed concrete beams were derived from the equations for
normal weight concrete, Eq. (26-12) and (26-13), by analogy, using the pro-
visions of Section 1708 as a guide. Separate equations for lightweight aggregate
concrete are necessary because such concrete can have a lower tensile strength
than normal weight concrete of the same compressive strength.
Section 1708(a) I requires that the limiting value for shear stress on the
unreinforced web of a lightweight aggregate concrete member may not exceed
0.3 ~ Fspv'J7 .
Section 1701(c) requires that for normal weight concrete this limiting value
be2~V]7.
For these two limiting values to be equivalent, the value of F,,, for normal
weight concrete must be 6.67, which is in reasonable agreement with the test
value found for sand and gravel concrete. From this relationship, Eq. (26-12A)
and (26-13A) were derived by analogy.
Guide for application- The web reinforcement in excess of the specified
minimum can be governed by either V..;, Eq. (26-12), or V""'' Eq. (26-13). Both
values must be computed unless the procedure can be abbreviated, for uniform
loads, by realizing that V,.,. usually governs ncar the support, and Vc; near the
quarter point. The center portion of the span is usually governed by the mini-
mum value of V,.;.
It may be advantageous to determine the web steel requirements by plotting
the various allowable and prevailing shears along the member. Fig. 21 is such a
typical and schematic diagram with the computed values of V,..,., V,.;, and Vc; """
at various points along the span. Plotting enough points to define the curves
illustrates where each type of shear governs and its relationship to the actual
84 COMMENTARY-BUILDING CODE

4 of span

Legend

11!1111 Cross·hatched area indicates

1-------1 web reinforcement in ucess
of A.,.mln.
IZJ Dolled areas indicate the web
reinforcement required.

This strip represents the

shear carried by the minimum
web reinforcement, Avmln.

Fig. 21-Schematic diagram of shear computations for Vcu· 1 Vc;, and Vc~. m '".

ultimate shear force, V11. This visual interpretation of Eq. (26-10) indicates
where VII exceeds V, .... or V,.; and where extra stirrups must be provided.
General comments- These provisions for design in shear were developed
from tests on beams having either rectangular or !-shaped cross sections, with
webs of constant width. If the web width of a member varies over its height, the
use of the minimum width, as required by the code, will lead to conservative
designs.
In lieu of the term d, which is specified differently for most of the equations
of Section 2610, an effective depth d, could be defined as either the distance
from the extreme compression fiber to the centroid of the prestressing tendons,
or 80 percent of the over-all depth of the member at the cross section under
consideration, whichever is greater.
This value d, could replace d in all equations except Eq. (26-12) and
Eq. (26-12A).
Attention is called to the provision at the end of Section 2610(c), that not
even the minimum web reinforcement need be provided if it is shown by tests
that the required ultimate flexural and shear capacity can be developed without
web reinforcement. The problem is the determination of a parameter that will
accurately predict when no web reinforcement is required. In the absence of
such a parameter, minimum shear steel is required unless load tests have shown
otherwise.
When such tests are made, care should be taken to reproduce as closely as
possible the conditions of loading, and to simulate the support conditions of
the actual structure.

2611-Bond
The provisions of section 2611 apply to pretensioned, prestressed concrete
members using three or seven-wire steel prestressing strands. They are intended
to ensure that failure of a pretensioned prestressed member shall not occur by
the strand pulling through the concrete as a result of failure in flexural bond.
 PRESTRESSED CONCRETE 85

The provisions were established on the basis of tests performed on normal

weight concrete specimens with a minimum cover of 2 in. In general, this section
will control only in the design of cantilevers and short span members.
The expression for development length I may be re-written as:

I = -4- D + (f,, - f,.) D

where I and Dare in inches, andj,, andf,, are in kips per square inch.
The first term represents the transfer length of the strand 1,, i.e., the distance
over which the strand must be bonded to the concrete to develop the prestress
~ f,, in the strand (Fig. 22). The second term represents the distance in addition to
the transfer length over which the strand must be bonded to the concrete, in
order that a stress j,, may develop in the strand at ultimate strength of the
member.
The variation of strand stress along the development length of the strand is
shown in Fig. 22. The build-up of stress in the transfer length, 1,, is much more
rapid than in the remainder of the development length. The expressions for
transfer length, and for the additional bonded length necessary to develop an
increase in stress of (f.,, - f,.) are based on tests* of members prestressed with
clean, 1/4, 3/8, and 1/2-in. diameter strands for which the maximum value
off,, was 275 kips per sq in.
The transfer length of strand is a function of the perimeter configuration,
area and surface condition of the steel, the stress in the steel, and the method
used to transfer the steel force to the concrete. Strand with a slightly ru~ted
surface can have an appreciably shorter transfer length than clean strand.
Gentle release of the strand will permit a shorter transfer length than abruptly
cutting the strands.
The value of f,.Df3 for transfer length is an average value based on data
reported by Kaar et al., * who also demonstrated that transfer length is not a
function of concrete strength for strands of up to I /2 in. diameter within the
normal range of strength used for prestressed concrete.
Bond failure of strand takes place in two stages: (I) adhesion between strand
and concrete is destroyed by high local bond stresses occurring after cracking,
a slight slip occurs along the entire length of the strand, which in most cases is
arrested by mechanical interlock between strand and surrounding concrete, and
(2) with a further increase in strand stress, the mechanical interlock breaks down
and collapse of the member occurs as the strand pulls through the concrete.
Test data by Hanson and Kaar show that the increase in bonded length
(I - 1,), beyond the transfer length, needed to permit development of strand
stress f,. is proportional to the increase in strand stress from the effective pre-
stress],. to the strand stressf,,. Fig. 23 shows beams prestressed with clean I /4,

•Hanson, N. W., and Kaar, P. H. "Flexural Bond Tests of Pretensioned Beams," ACI Journal, Proceedings
V. 55, No. 7, Jan. 1959, pp. 783-802. Also IJel'elopment Departmeltl Bulletin D28, Portland Cement Associa-
tion, Skokie, Ill.
Kaar, P. H.; LaFraugh, R. W.; and Mass, M.A., "Influence of Concrete Strength on Strand Transfer
Length," Jotmw/, Prestressed Concrete Institute, V. 8, No.5, Oct. 1963, pp. 47-67. Also De••e/opmeltl Departmeltl
Bttlletin D11, Portland Cement Association, Skoie, Ill.
86 COMMENTARY-BUILDING CODE

3/8, or 1/2-in. strand, where failure occurred by the strand pulling through the
concrete. Open points indicate increase in stress to where slip-of-the-strand was
first observed at ends of beams. Solid points indicate increase in stress to com-
plete bond failure when beams collapsed after strand pulled through the con-
crete. The data arc represented reasonably conservatively by the straight line:

I. C.

I - /, = (f.<., - f".)D
which is the second part of the equation. ·~
The expression for development length in Section 2611 leads to a required
bonded length of 150 strand diameters for f,., = 250 kips per sq. in.; i.e. 56 in.
for 3/8-in. strand and 75 in. for I /2-in. strand. When fm is not the specified
ultimate strength of strand, the required development length will be correspond-
ingly shorter than if the ultimate strength were to be developed flexurally.
Section 2611 docs not apply to plain wires nor to end anchored tendons.
The length for smooth wire could be expected to be considerably greater due to
the absence of mechanical interlock. Flexural bond failure would occur with
plain wire when first slip occurred.

Prestress only
Ill
Ill
...
-
Q)

( /)

-
Q)
Q) -
ii!

- .
( /)
Q)
1/1

~·I··----- usu-tse>D ---~

Distance From Free End of Strand

Fig, 22-Variation of steel stress with distance from free end of strand
 PRESTRESSED CONCRETE 87

2612-Repetitive loads
Repetitive loading is the type of load that occurs on a bridge. In a building
it can occur as load imposed on a floor by heavy reciprocating machinery.
Ample laboratory tests and field testing (AASJ-10 Road Test bridges) have
shown clearly that the fatigue resistance of prestressed concrete is excellent.
Section 2612(a) warns the engineer of the possibility of bond failure when
flexural cracking is expected. The use of the equation for flexural bond in
i Section 2611 should eliminate such difficulties since it is based on general bond
slip. Additional bond capacity still remains from the mechanical interlock
between the helically wound strand and the concrete.
"' Section 2612(c) also warns against shear cracking at repetitive loads lower
in magnitude than the static loads used for design. Thus it would appear prudent
to use at least the minimum shear steel expressed by Eq. (26-11 ), even though
static tests show that no shear steel is needed and Section 261 O(c) would permit
its elimination.

v
140

~
120 T 4

~
0

.5
c:r
@
A .
r
en
100
...
Q)
.....~
0.
en
.9- 80
,o/. ~

~
~ @c
A 0
en » 0
en
@

-
...
Q)

en
a;
-
Q)

en
60

I 7
@v
c
Q)
en
cQ)
...
0
c
40

20
/
I/
v )
efs~rfse;increase in steel stress
of bond failure
ofsb- fse;increase in steel stress

/
when slip of strand
1 first occurred at beam
ends 1 1
0
0 20 40 60 80 100 120 140
( 1-lt )/D
Fig. 23-Development length of prestressing strand
88 COMMENTARY-BUILDING CODE

2613-Composite construction
Composite members are designed in accordance with Chapter 25, with the
exception that Section 2504(b) does not apply to prestressed concrete.
Section 2505(c) defines the limits on the spacing of vertical ties used to hold
together the cast-in-place and precast elements. The spacing limitations of four
times the slab thickness cannot always be met in both directions in the case of a
thin composite slab bonded to a precast tee, double tee, or channel. -However,
it can be met at the stems in the area used in design and this would meet the
intent of the section.
2614-End regions
•
The conditions in the end region arc so varied that specific requirements
arc not given.

2618-Concrete
For the first time since the Bureau of Public Roads "Criteria for Prestressed
Concrete," a minimum concrete strength is suggested for time of transfer of
the prestressing force. It may be advantageous, both from an economic and
from a camber control standpoint, to use higher concrete strengths at release.
There arc products such as precast piling where camber is not a considera-
tion and where loss of prestress could be accounted for in the design. In such
instances it may be possible to transfer the prestress force at a lower concrete
strength.*
2621-Application and measurement of prestressing force
Section 2621 (a) is a requirement to insure that the amount of tension
assumed for the steel in design is actually placed in the steel. A similar provision
appeared in the report of ACI-ASCE Committee 426.
Section 2621 (c) calls attention to changes in prestress force that can occur
when large differences exist between the temperature of the steel and concrete.
It is apparent that a long length of prestressing steel that is tensioned at very
low temperatures and then embedded in warm concrete could expand enough to
lose a measurable amount of its tension.

2623-Formwork
For extended information on formwork, sec the reports of ACI Committee "'
347 (formerly 622).t

APPENDIX
Since the appendix applies to the methods for two-way slabs, commentary
on the appendix will be found under the commentary on Chapter 20, Joists and
Two-Way Slabs.
tACI Committec622, "Formwork lor Concrete," ACI Journal, l'roceedings V. 57, No.9, Mar. 1961, pp.
993-1040.
ACI Committee 347, "Recommended Practice lor Concrete Formwork (ACI 347-63)," ACI Journal,
l'roceedings V. 60, No.2, Feb. 1963, p. 169.