NON-TENSIONED STEEL

IN PRESTRESSED
CONCRETE BEAMS

A. F. Shaikh
University of Wisconsin-Milwaukee
Milwaukee, Wisconsin

Non-tensioned steel can be used

in prestressed concrete beams to
serve various purposes. However,
due to a lack of sufficient understanding of the behavior of prestressed concrete beams containing
non-tensioned steel, only a very limited application is to be found in
practice.
The use of non-tensioned steel in
prestressed concrete has been referred to as partial prestressing,
which, in general, is taken to mean
either or both of the following conditions:
1. Tensile stresses are permitted
under working loads
2. Non-tensioned steel is used in
addition to tensioned prestressing steel.
Some work on partial prestressing
has been done in the last fifteen
years (1,2,3,4,5,6,7), but most of it deals
with the form of partial prestressing defined under item 1 above.
Abeles ( 1,3 ) has made brief references
to the use of non-tensioned steel and
questioned the generally accepted
notion of its ineffectiveness in uncracked prestressed concrete beams.
Abeles( 3 ) mentions that "with more
non-tensioned steel of lower
strength, as compared to less non14

D. E. Branson
University of Iowa
Iowa City, Iowa

tensioned steel of higher strength,

the loss of prestress will be directly
more. However, as better control on
camber and cracking is likely, a
vital need for research in this direction exists."
The only other report on the use
of non-tensioned steel appears to be
that of Hutton and Loov( 7 ), which
was published in December 1966.
This paper contains observed camber and deflection curves of a limited number of beams containing
non-tensioned steel.
Practically no analytical work or
conclusive experimental work has
been reported on the use of nontensioned steel in prestressed concrete beams with reference to camber, loss of prestress, and deflections
of cracked sections.
OBJECTIVES

This paper details the findings of

an analytical and experimental study
on the effects of both the quality
(type) and quantity of non-tensioned
steel on the following behavior characteristics of prestressed concrete
beams:
1. Camber (short-time and timedependent)
2. Loss of prestress
PCI Journal

A study of the effects of non-tensioned steel on the behavior of

prestressed concrete beams is presented. Effects on camber, loss of
prestress force, cracking, and deflections are included. Analytical
results are compared with the observed behavior of twelve, simply
supported, prestressed concrete beams, ten of which contained
non-tensioned steel.

3. Crack formation
4. Deflections under working
loads and overloads.
DESCRIPTION OF
EXPERIMENTAL INVESTIGATION

The experimental program consisted of the testing of four series

of pretensioned, prestressed concrete beams. Each series included
three simply supported, 6 x 8-in, by
15-ft. beams, for a total of 12 beams.
Table 1 shows the details of the test
beams.
The three beams in each of the
four series were designed for the
same prestress force. Series I was
designed to study the effect of quantity of non-tensioned steel on the
behavior of prestressed concrete
beams. Series II and IV were designed to study the effects of quality
as well as quantity of non-tensioned
steel. The distinction between the
two series was the total amount of
steel. Series III was designed to
study primarily the effect of steel
prestress level.

Measurements, methods and

instrumentation.
1. All test beams, shrinkage specimens and control cylinders were
February 1970

moist cured for 7 days by keeping

them covered with wet burlap.
The temperature in the laboratory
ranged from 70F to 80F, with
an average value of 72F. All test
beams were prestressed at 7 days.
2. Steel collars with electrical
SR-4 strain gages served as load
cells for measuring the individual
strand prestressing force. Fully
temperature-compensated, f o u r arm bridge circuitry was employed.
3. Initial and long-time midspan camber values were obtained
using two dial gages, one on each
side of the beam. The discrepancy
between the readings of the two
dial gages was found to be insignificant.
4. Initial and long-time concrete strains were obtained using
a Whittemore mechanical strain
gage with a 10-in. gage Iength.
Each beam had three gages distributed from top to bottom on
both sides of the beam.
5. Records of temperature and
relative humidity were kept
throughout the time-dependent
study. A sling psychrometer was
used to obtain the relative humidity data.
15

Table 1. Details of test beams()

Series No.
f'^ (psi)

3
.

5400
(28 day)

4120
(7 day)

Beam No.

Non-tensioned
steel

#4

000

#4

000

Prestressing
strand dia. (in.)

5/16

5/16

Design Fa (kips)

30

Actual F o (kips)

5890
(28 day)

4380
(7 day)
1

IV

III

II

6570
(28 day)

4830
(7 day)
1

5880
(28 day)

4300
(7 day)
1

Of0

OHO

#4

OO

3/8

3/8

3/8

3/8

30

30

26

26

26

30.5

29.8

29.8

25.2

25.8

24.4

0.116

0.173

0.240

0.240

0.160

0.160

0.160

0.200

0.400

0.310

0.080

0.310

0.600

0.30

0.30

0.30

0.45

0.60

0.60

0.40

0.40

0.40

1.50

0.15

0.50

1.00

0.80

0.20

0.80

1.50

2.30

3.46

0.50

1.73

3.46

1.30

0.50

1.94

3.76

0.23

0.27

0.31

0.20

0.20

0.21

0.20

0.25

0.31

0.28

0.28

0.29

21.3

23.8

26.2

19.0

19.8

19.8

19.5

25.6

29.6

24.8

27.7

26.1

#4

000

000

000

#5

O0

5/16

5/16

3/8

20

20

30

20.2

20.0

19.7

0.173

0.116

0.116

0.400

0.600

0.058

0.45

0.45

0.45

p' =A; / bd (%)

0.50

1.00

A$ /A,, = p'/ p

1.15

f'
, fi
p f, + p f,
Design M (kip-ft.)

5/16

#4

OYSO

OO

5/16

5/16

5/16

30

30

20

29.8

29.0

30.1

AS (sq. in.)

0.173

0.173

As (sq. in.)

0.200

p = A$/ bd (to)

All beams 6 x 8-in.; d = 6.5 in.; span = 15 ft. simply supported.

OPrestressing steel
high strength steel

#4

O.O

3/8

Non-tensioned 33 ksi minimum yield steel

o Non-tensioned 60 ksi minimum yield steel

#5

1.30

At
A
g 1.20

f,
f' = 5000 psi
f )
f su = fs (1-0.5p s
c
fy = 33 ksi
fs = 250 ksi
n = 7
Q
4

1.10

le
^aSi^
^1
aati ogle
tir o ^^, 4

4 o
4

1.00

0.1

,1e

,^aq1

6^ ^ o g^S 4

oar

0'

0
4
p + = 0 p = variable
f
f U.'
0.3
'su+
P ' f,
y
c
c

0.2

Fig. 1A. Area ratio vs. steel percentage parameter

was found to be negligible. The
number and height of visible
cracks and the cracked length of
each beam were recorded at various stages of loading.

6. At the end of the time-dependent study period, the beams

were loaded to failure under a
two-point loading applied with a
Universal testing machine using a
spreader beam. Mid-span deflections were recorded at one-half
kip increments with two longreach dial gages. As in the case
of camber, the discrepancy between the two dial gage readings

CAMBER

The effect -of non-tensioned steel

on the factors that influence initial
and time-dependent behavior of
prestressed concrete beams was

1.40

ye
^aS le
I 1.30
t
I
9

o+

4
o

1.20

Stiff

^a
%+ 4

4^

e
ayy
_yam
mat

4^.

1.10
P
1.00

^.e

ail

0.1

0.2

0, P = variable
0.3

0.4

f su fy
_

+ p' f
Pi
f
c
c

Fig. 1B. Moment of inertia ratio vs. steel percentage parameter

February 1970

17

studied. A method is presented

which, in conjunction with any of
the available methods for predicting
camber of prestressed concrete
beams without non-tensioned
steel( 8'9 >, will enable the prediction
of camber for prestressed concrete
beams with non-tensioned steel.
Short-time (initial) and long-time
(initial plus time-dependent) camber
are considered separately.

Short-time camber. For a prestressed

concrete beam without non-tensioned steel, it is usually satisfactory
to use gross section for computing
section properties. However, when
non-tensioned steel is used, depending upon the amount and location,
the effect on the transformed area
and/or the moment of inertia of the
transformed section may become
quite significant. This is demonstrated in Figs. 1A and 1B, which
plot the area ratio (area of transformed section to area of gross section, At/A9) and the moment of inertia ratio (moment of inertia of the
uncracked transformed section to
moment of inertia of the gross section, It/I9) against the steel percentage parameter, p+
f^

p' f^' . The steel

f^

percentage parameter is indicative

of the ductility of the beam crosssection. The plots in Figs. 1A and
1B were obtained for a rectangular
cross-section with both the prestressed and non-tensioned steel located at an eccentricity of 0.4 h.
For

f =
p' y

0.3 (which de-

termines the maximum amount of

steel permitted for ultimate strength
computations per ACI, AASHO and
PCI codes), At/A9 = 1.25 and It/Ig
= 1.38 for a beam which contains
no prestressing steel (an ordinary
18

reinforced concrete beam). These

ratios are equal to 1.04 and 1.07
respectively for a fully prestressed
beam (without non-tensioned steel).
The effect of transformed section in
the latter case may be ignored, but
should be considered in the former
case for accurate results.
Between these two extremes, there
is a family of curves which pertain
to prestressed concrete beams with
some non-tensioned steel. Depending upon the amounts of non-tensioned and prestressing steel, and
also upon the accuracy desired, a
decision regarding the use of the
gross section or the transformed section may then be necessary.
For the test beams the maximum
area ratio, At/A9 was 1.08 (Beam I
,
B3) and the maximum moment of
inertia ratio, It/I , was 1.05 (Beam
IVB2). Thus, for the test beams,
the use of the gross section was
considered satisfactory. Initial camber values for the test beams were
predicted using the gross section and
compared with the measured values.
These are shown in Fig. 4 and Table
3 to be in good agreement.

Long-time camber. Long-time camber in a prestressed concrete beam

consists of initial camber and timedependent camber. Initial camber
occurs at release of prestress; timedependent camber is caused by
strain changes due, primarily, to
creep and shrinkage of the concrete.
The strain changes due to shrinkage and creep of concrete bring
about a loss of prestress which has
a two-fold effect: first, a reduction
in initial curvature due to reduction
in prestress; and second, a change
in creep rate (decrease) due to a
reduction in concrete stresses. In
other words, the changes in initial
deformations are caused by an inPCI Journal

teraction between creep and shrinkage of the concrete and loss of prestress. Other factors which also
influence these changes are steel relaxation and the increase in modulus
of elasticity of concrete with time.
All of these factors are both timedependent and inter-dependent.
There are basically two methods
for computing deflections of prestressed concrete beams: a detailed
method( " , 5 > that considers the effects
of shrinkage and creep separately;
and a simplified method that lumps
together the effects of shrinkage,
creep and the loss of prestress into
a combined time-dependent coefficient. This is rather an over-simplified approach, as it does not take
into consideration the stress level
and distribution, prestress loss, quality of concrete, increase in concrete
modulus with age and the presence
of non-tensioned steel. The ACIASCE Joint Committee( 1 method
is an example of the simplified method where the ultimate creep coeffi-

e'

Asl

cient, Cu, of 1.0 to 3.0 is actually

the combined time-dependent coefficient. This combined time-dependent coefficient will be referred to as
a camber coefficient( $ >, Bt, as it is
directly used with camber.
Some recent studies (11,8,9) have
been made to determine the effects
of non-uniform stress distribution,
shape of specimen, loss of prestress,
increase in concrete modulus of elasticity with age and effect of variable
stress levels.
In order to use both the detailed
method and the simplified method
for computing deflections of prestressed concrete beams containing
non-tensioned steel, three modifica.
tion factors are derived. Let ash, a,
and a be three factors such that,
when respectively applied to the
shrinkage strain, creep coefficient
and camber coefficient values of a
prestressed concrete beam without
non-tensioned steel, they will give
the corresponding quantities to be
used in the case of an identical

_c^c._

E - . .

oA 0

L
A1 11. Bt1 L

Ecidx

^'

Fo

AsEsBtl L

o Ecidx

Beam 1 (Amount of non-tensioned steel A'

sl

A'
s2
0A o

e'

c.g.c._
As2EsBt2
--

F ^
o

A5ESBt2

L
o (

Ecidx

L I o L Ecidx

' )
Fran 2 (Amount of non-tensioned steel A'2 >As1
s

Fig. 2. Comparison of beams with different amounts of non-tensioned

steel
February 1970

19

beam with non-tensioned steel.

The basis for evaluating these
modification factors follows.
During time-dependent deformations, work done by forces in the
steel less work done against beam
dead load is the change in internal
strain energy of the beam, or
Ws Wd = E b
(1)
The modification factor, a, for the
camber coefficient, B t, is derived as
follows:
Consider two beams which are
identical in every respect except the
amounts of non-tensioned steel (Fig.
2). Let Bn be the camber coefficient
for Beam 1. Let a be a factor (may
be time-dependent) such that aBtj
= Bt2, the camber coefficient for
Beam 2. Also let As1 < As2 where
As = area of non-tensioned steel.
Referring to Fig. 2, during an interval of time, dt, Eq. (1) in differential form for Beam 1 is:

dW31 dWdl = dEbt

(2)

r,

dWsi = F,, dtt

AS1 E;Bt1

w Br1 J yvdx

Eb = FoBt2 e,.^dx
0

2A,
r,

2 AE S B
wB t4
0

dBti

yticlx

Now assuming that changes in the

strain energies of the two beams
during time-dependent deformations
are proportional to their respective
camber coefficients (intuitively this
seems a reasonable assumption as
the stress in concrete for the two
beams would be approximately the
same and the time-dependent strains
would be B tiev and Btzeti):

1+

Ewdx

0
L

s1

EsAs

e dx
L

(8)

eetidx
s

s2 0
L

e^adx

+
0

Ebi =1'oBti f ejdx

0

20

E , A,

a=

Substituting Eqs. (3) and (4) in Eq.

(2) and integrating from t = 0 to

2A,E8 Bt1

(7)

Using Eqs. (5), (6) and (7):

(4)

where io = the dead load of the

beam
y j = initial camber

t = t:

(6)

yjdx

Bc1 _ 1
Eb2 B 2 a

As 1 E' Bti d
dt i ^ Lef`dx
(3)
0

dWd1 = w d
t

dx

(L

(5)

Similarly for Beam 2:

A,E8Bci dBt1^1,ecidx

IE 2 dx

L
0

E,v dx

where CZ and e are initial strains at

the levels of prestressing steel and
non-tensioned steel, respectively, and
are computed as:
PCI Journal

__ _[F,
E`"

EG _

Fo

Fee
I

Mxe^

(^12

F,ee^ M.,ex^ (g)

EsAs^
1+

J(1 )

1 + E ;A s2

For constant eccentricities (i.e. e, _

e and e X =e'):

E,A ,

Fo + F"e
^^=C A
I )
a=

1 r( F
A
P

S u E

( )

11

Fee')

wL2e'
120
I )]

(12)

rL
E'

"dx

edx
0

the following approximation may be

made:
rL

J Ec^ dx
dx

Further, in the case of bonded

cables, the effective modulus of elasticity of the prestressing steel is approximately the same as that of nontensioned steel bars, and Eq. (15)
reduces to:

A,

a=

(16)

A,

To establish a comparison of camber coefficients of two beams, one

without non-tensioned steel and the
other with non-tensioned steel, let
A;1 = 0 and A'2 = A' . Eq. (16) is
written as:

,.

Eri
=

that Beam 1 contains no non-tensioned steel, the following expressions are obtained:

(13)

() Li4
L/4

where the subscript L/4 indicates

that the ratio is computed at the
quarter-point of the span using Eq.
(13). Eq. (8) may be rewritten as:

(17)

Modification factors for creep and

shrinkage may be derived on a similar basis (12,13'. For the condition

(x

February 1970

(15)

1
__
a 1+A' /AS

r,

EL

L/4

1+ ' 81
E.SAS

1+-

It is noted here that for the case

of constant eccentricities, the x-dependent part in Eqs. (11) and (12)
is that due to M, and is relatively
small. As the quantity of interest is
the ratio

l
Eci

1} s A s

_ wL2e Fo F,ee'
61(A- + I
+

(14)

Ei

EA;^

6I ^A + I
1wL-'e

L/4

E'A'

wL 22 eF, Fe-

+ 120 ( I )

Eel

However if e = e' for all values of

Ay, then Eq. (8) reduces to:

_ 1
So ^`idx

E,A,

a=

C t1

1
=

CO

E 'AS ^'E 12dx

1+

L
E,A, I e?zdx
0

(18)
21

as^^
(Esh)1
(EgT^)z

ESAs
1+E8A.

(19)

Under the same assumptions as in

the case of the modification factor,
a, the alternate expressions for a,;
are:
a =

E'A'
E8A.

1+
a0 =

'2
--

Era L/4

1+

(20)

( 21)
S S

E,A,

and for the condition E',= E:

(22)

a, 1 + A; /A,t
a4''

=1

(23)
'

Table 2. Comparison of modification

factor with ACI reduction factor for
compression steel

A
A,

Eq. (17)

Reduction factor for

compression steel

1.0

0.50

0.40

0.5

0.67

0.65

1.00

1.00

Some comments on the previously

derived equations are in order:
1. Even though the modification
factors a, a, and cesh, were assumed to be functions of time, the
resulting final expressions are independent of time. This was verified by the observed camber behavior of the test beams.
2. It was observed that the reduction of time-dependent camber is
not directly proportional to the
area of non-tensioned steel. As a
22

matter of fact a law of diminishing returns seems to apply, as

shown in Fig. 4, which plots Eqs.
(17), (22) and (23).
3. For very large amounts of nontensioned steel, the modification
factors previously derived would
no longer be accurate, because of
the transformed area effects. However, from design considerations,
large percentages of p' are rarely
used because of the desirability
of achieving ductile (under-reinforced) beams and because of
economic considerations.
4. The effect of non-tensioned
steel used to control time-dependent camber is similar to the effect
of compression steel on deflections
of ordinary reinforced concrete
beams due to creep and shrinkage. Table 2 compares Eq. (17) to
the reduction factor for compression steel in reinforced concrete
beams proposed by ACI Committee 435( 14 ), and as found in the
ACI Building Code (318-63).
Comparison of theoretical and test

results. The time-dependent camber

curves for the test beams are shown
in Fig. 5. The top curve in each figure is labeled 100 percent, and refers to the beam without non-tensioned steel or with the minimum
amount of non-tensioned steel in
each of the four series. The other
percentages are the average observed time-dependent cambers of
the other two beams in each series
compared to the beam with minimum non-tensioned steel.
A comparison between test results
and theoretical results is made in
terms of the modification factor, a,
for the camber coefficient, B. Table
3 shows the observed and computed
values of a along with the observed
range of the values of a. The obPCI Journal

T
CD

Ic
0

CD

Table 3. Experimental and theoretical values of the modification factor and of initial and time-dependent camber
Series No.

IV

II

III
3

Beam No.

As/ As

1.15

2.30

3.46

0.50

1.73

3 .46

1.30

0.50

1.94

3.76

Experimentally
observed range
of a, %

100
to
100

70
to
77

60
to
65

100
to
100

62
to
66

32
to
40

100
to
100

70
to
80

43
to
50

100
to
100

78
to
85

39
to
45

Average of the
observed values
of a, %

100

74

63

100

65

37

100

77

48

100

83

42

Theoretical value
of a, Eq. (8), %

100

74

56

100

56

42

100

73

48

100

54

44

Theoretical value
of a, Eq. (13), %

100

72

55

100

55

40

100

71

45

100

51

42

0.142

0.140

0.225

0.225

0.219

0.204

0.205

0.184

Observed initial
camber, in.

0.251

Computed initial
camber, in.

0.254

0.254

0.254

0.138

0.138

0.138

0.218

0.218

0.218

0.202

0.202

0.202

Observed time-dependent camber, in.

0.254

0.191

0.157

0.123

0.075

0.042

0.249

0.213

0.106

0.130

0.106

0.041

172

172

172

140

140

140

124

124

124

123

123

123

Time-dependent
study period, days

0.252

* Readings could not be obtained as beam shifted significantly upon release of prestress force.

Bt, (for beam 1)

'H

Bt2 (for beam 2)

Bt2 = a8t1

dt

or: a = Bt2/Btl
Time,

Fig. 3. Camber coefficient vs. time

served range is seen to be fairly

narrow, thus verifying the theoretical conclusion that a is independent
of time. Table 3 also gives the computed and observed values of initial
camber and observed time-dependent camber. A comparison of modification factors is seen to be good,
except in the case of Beam IVB2.
The discrepancy in this case is attributed to experimental errors.
EFFECT OF NON-TENSIONED STEEL ON
LOSS OF PRESTRESS

The initial prestress force applied

to a prestressed concrete beam decreases at a decreasing rate with
time. The major contribution to the
loss of prestress (usually 70 to 80
percent of the total loss) is due to
shrinkage and creep of the concrete.
In a prestressed concrete beam
without non-tensioned steel any loss
of prestress force results in an equiv-

alent reduction of force on the concrete section. However, when nontensioned steel is included in a
prestressed concrete beam, this reduction of force on the concrete is
equal to the loss of prestress force
plus the forcetransferred to the
non-tensioned steel. Thus, when nontensioned steel is used, a distinction
between the loss of prestress and the
reduction of force on the concrete
must be made. A determination of
the reduction of force on the concrete permits an evaluation of the
change in stress level in concrete
from which the net stress in concrete
can be computed. This net effective
stress in concrete is of primary importance from the point of view of
creep rate and behavior (deflections
and extent of cracking) under service loads.
It has been shown (Fig. 4) that
non-tensioned steel reduces creep
and shrinkage strains. This reduction in strains results in a reduction
in the loss of prestress. To arrive at
a relationship between the loss of
prestress and the reduction of concrete force, consider two beams:
Beam 1 is without non-tensioned
steel and Beam 2 contains some nontensioned steel. Define two parameters /3 and y as:
_ loss of prestress for Beam 2
loss of prestress for Beam 1 (24)
reduction of force in
y =

concrete for Beam 2

(25)

reduction of force in
concrete for Beam 1

This loss of prestress ratio due to

creep effects, /3^, may be expressed
as:
0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

A0E8Ct2

4.0

A' IA
lC

Fig. 4. Relation between modification

factor and steel ratios

24

A,S E OC t1

L
L

,dx

0
L

e01dx
PCI Journal

0.30
0.25
ma
C ^a
e

0.20'-

ev

F Z 0.10

0.15
q
uy A 0.10
a^
nq1 ,c 0.05
E9

0.25

Ca

on

50

g0F
Q

0.10

aI a 0.05
0

0'

u5^'

0.^ 4^

s p.

Series II
di = 0.141 in.
avg
avg. meas.

= 0, Fo

040 , pH

4 60
=
p = p,40%, p33

y o
n
0.05
P.
05

2 (77%)

= 29. 8

3 (48%)

100

0,20%, Fo -

100

Series III
pi = 0.221 in.
avg. meas.
= 0.218 in.
comput
computed

124

25.2k

1 (100%)
2 (83%)

0.80%, F o = 25.8
1.50%, Fo = 24.4k

50

_ ^omputed

140
1 (100%)

50

0.10

ev

172

150

100
k
= 3p .5

6
= 0.80%, F o
%,
= 0 .60 P33

.,, 0.15
C 1

Series I

p = 0.307, p3 3 = 1.00%, Fo = 19 ,7k 3 (37%)

0.15

3 (63%)

= 20.2k
1 ((100%)
p' = 0.15%, F o
30%
k
F = 20.0 2 (65%)
0.50%,
= 0,30%, p6

0.20

100

50

2 (74%)

0.252 in. avg. meas.

0.254 in. computed

^i

0.10
0

1 (100)

= 29 "K
0.5%, F o
0, 45%,Q33 = 1.00%, Fo = 29.Ok
=
k
0 .45% , p1
= 1,57 F 0 = 30.1
0.45%, 4g3
p S

Series IV
= 0.198 in.
avg. meas.
= 0.202 in.
computed

3 (42%)
123

Fig. 5. Time dependent midspan camber (total minus initial)

plotted against time in days
r

Ct2

R` =

crl - a,

(26)

and the ratio of concrete force reduction due to creep, y c, as:

AE, Edx

yC = aC I +

(27)

A,E, 0E,1dx
If e, = e, then:

1L Ec.Idx
ASEsCt" L
o
+ ASE'C t 1 ^ ' E',dx
L
yc=
1 ('L
A,E S C tl L J Edx
u
February 1970

L
E1,v dx =

Edx
0

^0

ae =

A,E,
1+
A,E0
25

Table 4. Computed loss of prestress and reduction of concrete force for the test beams
N

Series No.
Beam No.

II

IV

III

1(1)

1(')

1(')

1(1)

1.15

2.30

3.46

0.50

1.73

3.46

1.30

0.50

1.94

3.76
0.44
1.13

13, Eq. (31)( 2 )

1.00

0.74

0.56

1.00

0.56

0.42

1.00

0.73

0.48

1.00

0.54

y, Eq. (32) ( 2 )

1.00

1.05

1.08

1.00

1.00

1.09

1.00

1.00

1.12

1.00

1.00

(2)

The beam in each of the four series without non-tensioned steel or with the minimum amount.

Values of /3 and y used to compute the loss of prestress and reduction in the concrete force for the test beams. For example, if the loss of prestress and the reduction of the concrete force in Beam lB1 are k s ,F. and k0 ,Fo, respectively, then the
loss of prestress and the reduction of the concrete force of Beam lB2 would be
pk s ,F. (where p = 0.74) and yk,,Fo (where y = 1.05), respectively.

Table 5. Number, height and distribution of visible-cracks in test beams

Series No.
Load, P
Beam No.
No. of cracks(')

C)

C)

II

II

III

IV

6 kips

5 kips

6 kips

6 kips

6 kips

IV
8 kips
2

12

10

11

12

13

12

Max. height of
a crack( 2 ) (i n.)

4.00

4.55

3.20

5.20

4.55

4.55

5.68

5.20

5.12

4.8

4.64

4.00

4.55

4.00

3.44

5.44

4.48

4.08

Ave. height of
cracks (in.)

3.60

3.20

2.04

4.64

4.00

3.84

5.20

4.80

4.48

4.40

3.20

3.20

4.16

3.52

2.56

4.80

4.16

3.76

Length of beam
cracked( 3 ) (ft.)

3.2

4.4

5.0

3.4

3.7

5.1

3.7

4.8

5.2

3.8

3.5

4.6

3.7

4.8

5.3

4.4

5.4

5.8

Refers to cracks in one half of each test beam.

Observations of the other half indicated approximately the same results.

(2'

The beams were all 8 in. deep.

3'

Refers to the distance over which the cracks in

one half of each beam were distributed. The
beams were all 15 ft. long.

y0

(28)

=1

Similarly during shrinkage:

Rsh = ash

(29)

ys^^ =

(30)

For the camber coefficient:

(31)

/3=a
y=a1+

AS E,

A,E,

E,.tdx

(32)
E

^dx

If e x = ey
y=1

(33)

Thus, if a beam without non-tensioned steel has an initial prestress

force, F0, and loss of prestress at a
particular time, t, of

(AFtsh + AFt ),

then the reduction of concrete force

will be equal to the loss of prestress,
AF t"h + AF, C . If non-tensioned steel
is provided in this beam such that
modification factors to shrinkage and
creep are, aa,, and a, respectively,
then the loss of prestress would be
a sh /F t + a4F t and the reduction
of concrete force would be yshAF'tsh
+ y AF t. , where y$ n and yo are given
by Eqs. (30) and (27) respectively.
If the combined coefficient, Bt, is
used and if the loss of prestress in
the beam without non-tensioned
steel is AF t , then the reduction of
concrete force is AF t . However, in
the beam with non-tensioned steel
the loss of prestress would be ahFt,
and the reduction of concrete force
would be yAF t , where y is given
by Eq. (32).
The computed loss of prestress
and the computed reduction of concrete force are compared in Table 4.
For this comparison define a prestress loss coefficient, k st, such that
the loss of prestress will be equal to
February 1970

k,,F 0 . Also let k, t be a reduction of

concrete force coefficient such that
the reduction of concrete force will
be equal to k, t F 0 . The loss of prestress and the reduction of concrete
force for the other two beams in
each of the four series are expressed
relative to the beam with the minimum amount of non-tensioned steel
in each series. It is noted that, in
the case of a beam without non-tensioned steel, kst = k0t.
Table 4 shows that the total concrete force is relatively insensitive to
the provision of non-tensioned steel
(it is invariant for e = e' since y, =1,
ysh = 1 and y = 1). In other words,
any reduction in the loss of prestress
appears as the force in the non-tensioned steel. Even for e'/e = 0.5
(Beam IVB3 which contains the
maximum amount of non-tensioned
steel), where the loss of prestress is
56 percent less than that in Beam
IVB1, the reduction of concrete
force is only 13 percent more than
that in Beam IVB1. In practice the
ratio e'/e is usually close to 1 and
thus, for all practical purposes, it
may be assumed that the provision of
non-tensioned steel does not influence the effective force in concrete.
CRACK FORMATION, DEFLECTION AND
ULTIMATE STRENGTH BEHAVIOR

The existing philosophy for the

design of prestressed concrete members is to allow either no tensile
stresses under working loads (fully
prestressed concrete) or no cracking
under working loads, even though
some tensile stresses may exist (a
limited form of partially prestressed
concrete). Nevertheless, the behavior
of cracked prestressed concrete
members is of importance from the
point of view of overloads. Knowledge regarding ultimate strength is
of interest in providing criteria for
27

10.0
8.0
6.0
4.0
ro
0
o
2.0

2.0

1.0

4.0

3.0

Midspan Deflection, in.

B -- Series II

A -- Series I

10.0
8.0

6.0
x
a 4.0
v
0
0
2.0
0
0

1.0

2.0

C -- Series III

3.0

4.0
D -- Series IV

Fig. 6. Observed midspan deflections vs. load

design. Increasing interest is also being shown in the design of prestressed concrete members that
would crack under working loads.
Since substantial cracking occurs
under working loads in ordinary reinforced concrete members, cracking
in prestressed concrete members
should be acceptable provided that
all safety and serviceability requirements are met. The presence of prestress force might then be considered an advantage as compared to
the corresponding reinforced concrete member.
The behavior of prestressed concrete members and ordinary rein28

forced concrete members is similar

under cracked conditions. Consequently, the extensive work that has
been done on ordinary reinforced
concrete members should provide a
strong basis for predicting the behavior of cracked prestressed concrete members.
It is reasonable to expect that nontensioned steel does not influence
the cracking moment of prestressed
concrete beams. This appears to be
verified by the observed load-deflection response of the test beams
(Fig. 6). The curves seem to deviate
from an initial linear relationship at
about the same load for all beams
PCI Journal

Beam IV Bi
p = 0.40%

ph = 0.20%
F = 25.2 kips
0

f
f
pfsu + p'fY
= 0.28
C

ult. - 8.95 kips

Beam IV B2
p = 0.40%
p

0 - 0.80%

Fo = 25.8 kips
f
f
pfsu + p'-= 0.28
c
c
P

ult.

= 10.11 kips

Beam IV B3
p = 0.40%
p 33 = 1.50%
Fo = 24.4 kips
f
f
p fsu + p'- = 0.29
c
c
P ult. = 9.23 kips

Fig. 7. Condition of Beams IVB1, IVB2 and IVB3 under a load of 6 kips
in each of the four series. For example, it is seen that in the case of
Series I this load is 3.3 kips. The
three beams differ only in the
amount of non-tensioned steel.
Whereas non-tensioned steel does
not influence first cracking, it has
February 1970

quite a significant cumulative effect

on the number, height and distribution of cracks. Studies of crack formation were made on an area of
one-half of one side of each of the
test beams. Other areas of the test
beams exhibited similar crack for29

Table 6. Comparisons of computed and measured values for the test beams
Series No.

II

Beam No.

Ill

IV

Observed ultimate load,

Pu
(kips)

8.7

9.4

9.9

6.8

7.2

7.3

7.0

9.3

9.7

8.9

10.1

9.2

Working load, P 24,('> (kips)

3.3

3.3

3.3

2.3

2.3

2.3

4.0

4.0

4.0

3.0

3.0

3.0

Load factor, P/ P

2.62

2.84

3.00

2.96

3.14

3.16

1.75

2.32

2.42

2.96

3.36

3.06

8.3

9.3 -

9.5

5.5

6.5

6.5

6.5

7.5

9.5

8.0

8.5

8.5

95

99

96

81

90

89

93

81

98

90

84

92

7.0

8.0

8.0

5.5

6.0

6.0

6.0

7.0

8.0

8.0

8.0

8.0

2.12

2.42

2.42

2.38

2.60

2.60

1.50

1.75

2.00

2.67

2.67

2.67

Pmai 1 2 1
P1

(kips)

P() X 100

P o` 3 '

(kips)

Applied overload ratio,

Po / Pw C4'

Worst discrepancy in
(percent)
deflection( 5 )

+15

+13

+17

-14

+6

-3

-19

-3

+12

+3

+9

+15

(k-ft.)

23.8

25.9

27.4

18.6

19.9

20.0

19.2

25.6

26.8

24.6

27.8

25.4

Computed M,1(k-ft.)

21.3

23.8

26.2

19.0

19.8

19.8

19.5

25.6

29.6

24.6

27.7

26.1

Observed M ((

Ratio of observed to
computed M,,

-o

C,

1.12

1.09

1.05

0.98

(1) For the test beams, the working load was assumed to represent
the condition that cracking would occur as soon as this load is
exceeded. P,, were computed values. Note that the load factors,
even for this assumption, tend to be on the high side for the
test beams.
(2) Represents the maximum load for which deflections were recorded.

1.00

1.01

0.98

1.00

0.91

0.99

1.00

0.97

(3> Represents the load at which the discrepancy between the ob-

served and computed values of deflection is the greatest.

> Gives an indication of the range of overload in which the discrepancy in deflections is the greatest.
s> Plus or minus indicate that computed deflection is greater than
or smaller than the observed deflection.

mation behavior. Table 5 lists the

number, maximum and average
heights of cracks, and the length of
the cracked portion of the beam for
each half of the test beams.
As an example, in the case of
Beam IVB1 under a load of 6 kips
(Table 5 and Fig. 7) there are 7
visible cracks in a length of 3.7 feet
with a maximum height of crack of
4.55 in. and an average height of
cracks of 4.16 in. The corresponding
quantities for Beams IVB2 and IVB3
are: 9 cracks over 4.8 ft., 4.00 in.
max. and 3.52 in. ave.; 9 cracks over
5.3 ft., 3.44 in. max. and 2.56 in. ave.
The three beams are identical except
that Beam IVB2 contains 4 times as
much non-tensioned steel and Beam
IVB3 contains 8 times as much nontensioned steel as Beam IVBI. All
have roughly the same ultimate load
capacity as shown in Fig. 7.

Deflections. The similarity of the

behavior of prestressed concrete
members and ordinary reinforced
concrete members under cracked
conditions led to the investigation of
the available methods given in the
literature (15,16,17,18,19) for computing
deflections of reinforced concrete
members. Since ordinary reinforced
concrete is normally cracked under
working loads, most methods for
computing these deflections do take
into account the effect of flexural
cracking.
Branson's method( 14,18 ) was used
to compute the deflections of test
beams. Based on a sizable number
of tests on rectangular beams (simple span and continuous) and Tbeams, Branson( 18 ) presented an empirical expression for the effective
moment of inertia at a given section,
Leff. The expression is given in a
form that includes the effect of
cracking as:
February 1970

1111",

IC/s_

+ [l_

(Mcr)

ill

4^ IC,

(34)

where M = moment at a particular

section
I = moment of inertia of the
cracked transformed
section.
An expression for an average effective moment of inertia for the entire length of a simply supported
beam under uniformly distributed
load is also given:
M

I`ff(i1i^^.^
+

XI,)

(-Y J

Ic,.

(35)

where Mma, = maximum moment in

the span. Eqs. (34) and (35) apply
only when M or Mmar is greater than
Mer; otherwise Ieff = I, (or It).
For continuous beams, the average
of the values for positive and negative moment regions is recommended m14 ' 19 . Although Eq. (35)
was originally established for simply
supported beams under uniformly
distributed loads, its use is considered quite adequate for two-point
test loading as well as for other
loads that are approximately symmetrical about the center line of
the beam.
The effect of non-tensioned steel
on deflections under cracked conditions is evident from Fig. 6. The
deflections of the beams with more
non-tensioned steel are considerably
less than the deflections at corresponding loads of identical beams
containing smaller amounts of nontensioned steel. For example, Beam
IB1 under a load of 8 kips shows a
deflection of 2.88 in., whereas Beam
31

IB3, which contains three times as

much non-tensioned steel, shows a
deflection of only 1.95 in.
Midspan deflections of the test
beams were measured at loads ranging from 81 to 99 percent of the
ultimate loads. Eq. (35), along with
the following expressions for M0,
and Ii,,., was used to compute these
deflections.

,I

yt + f` t 9
Me ,. = F tex +A
t
li

I,,..

^'

( 3d) s

(36)

+ n A s (d kd)2

+ n' A, (d' kd) 2

where k =

(np + n'p') 2 + (2np + 2n'p'

(np + n'p')

(37)

d
d'

(38)

The modulus of rupture, f' b, was

obtained by bending tests on plain
concrete specimens of the test
beams.
Maximum discrepancies in observed and computed values of deflection are indicated in Table 6.
Table 6 also gives the maximum
loads for which the deflections were
recorded.
The midspan deflections shown
in Fig. 6 are relative to the positions
of the beams just before the application of the transverse load. If the
deflections from the positions of the
beams before prestressing are desired, the total camber (initial plus
time-dependent, Table 3) just prior
to the application of the two-point
loading must be subtracted from the
deflections in Fig. 6.
Note that, with the use of nontensioned steel, greater not smaller
net deflections (as referred to the
position of the beam before prestressing) occur under working load.
32

This is because non-tensioned steel

reduces time-dependent camber and
thus, there is less total camber to be
cancelled before the beam deflects
downward from the position of the
beam before prestressing.
For example, in the case of Beams
IB1 and IB3 (having non-tensioned
steel percentages of 0.5 and 1.5 percent respectively) the total camber
values are 0.251 + 0.254 = 0.505 in.
and 0.252 + 0.157 = 0.409 in. respectively (Table 3). Under a transverse
load of 4 kips the observed deflections (Fig. 6) for the two beams are
0.534 in. and 0.514 in. respectively.
Thus the deflections relative to the
positions before prestressing are
0.534 0.505 = 0.029 in. and 0.514
0.409=0.105 in. respectively.
Whereas the deflection of Beam IB3
relative to its position just before
application of the transverse load is
smaller than the corresponding deflection of Beam IB1, its deflection
relative to the position before prestressing is significantly greater.
After first cracking, the increase in
deflection of a beam with a smaller
amount of non-tensioned steel will
be greater than the increase in deflection of an identical beam containing a larger amount of non-tensioned steel. This is due to a better
distribution of cracks and a reduction in the extent of crack development with a greater amount of nontensioned steel. The net deflection
(relative to the position of the beam
before prestressing) of the beam
with a larger amount of non-tensioned steel may be greater, comparable or considerably smaller depending on whether the applied
transverse load is approximately
equal. to, somewhat greater than, or
considerably greater than the cracking load.
In the case of most prestressed
PCI Journal

concrete beams with non-tensioned

steel, under increasing load the nontensioned steel would yield before
the ultimate load of the beam is
reached. This will certainly be the
case if the non-tensioned steel is of
lower strength than the prestressing
steel, and the beam is under-reinforced. However, for the usual
percentages of steel, the reserve
strength after yielding of non-tensioned steel is only a small percentage of the ultimate strength of the
beam due to the precompression in
the non-tensioned steel.
This observation is corroborated
by the load-deflection response of
the test beams. The only beam in
which non-tensioned steel yielded
(that is, up to the maximum load
for which the deflection was recorded) is Beam IBl. The contribution of its non-tensioned steel to the
quantity, p f;u -I- p' f; , is the least
(about 14 percent) f of all the test
beams. The yielding of non-tensioned steel seems to have originated at a load of about 8 kips. The
observed deflection is smaller than
the computed deflection at 8 kips,
but grows rapidly thereafter. Between 8 kips and 8.34 kips, the increase in observed deflection is
about three times the increase in
computed deflection. Even in the
case of Beam IBl, the load of 8 kips
amounts to about 92 percent of the
ultimate load.
CONCLUSIONS

1. The use of non-tensioned steel

in prestressed concrete beams may
necessitate the use of uncracked
transformed section properties as opposed to gross section properties for
reasonable accuracy (see Fig. 1).
2. The effect of non-tensioned steel
February 1970

on time-dependent camber is primarily due to restraints imposed on

creep and shrinkage of the concrete
as embodied in the modification
factors a, and a,,,, (see Eqs. (18)
through (23)). The factor a that
combines these two effects is given
by Eqs. (8) and (14) through (17).
The simplified Eqs. (15), (22) and
(23), in which e,, = e',, could be used
in practice to estimate the gross
effect of non-tensioned steel in reducing prestress loss and camber
when the eccentricities of the tensioned and non-tensioned steels are
approximately equal and on the
same side of the centroidal axis.
3. The effect of non-tensioned steel
in reducing time-dependent camber
of prestressed concrete beams is similar to the effect of compressive reinforcement in reducing long-time deflections of ordinary reinforced
concrete members (see Table 2) .
4. A distinction must be made between the loss of prestress force and
the reduction of the concrete force
in beams containing non-tensioned
steel. The loss of prestress is greatly
reduced due to the restraining action of the non-tensioned steel on
the creep and shrinkage of the concrete. However, the total effective
concrete force is quite insensitive to
the provision of the non-tensioned
steel. (See the discussion of /3 and y
defined by Eqs. (24) and (25), and
the results presented in Table 4.)
5. From a practical point of view,
the non-tensioned steel does not influence first cracking (i.e. cracking
moment) of prestressed concrete
beams.
6. If the net deflection under working loads is downward relative to
the position of the beam before prestressing, then, by using non-tensioned steel, this deflection would
be larger (assuming no cracking has
33

occured) because of the substantially reduced time-dependent camber.

7. Whereas non-tensioned steel
does not have a substantial effect on
first cracking, its effect on subsequent crack formation is quite pronounced. The additional bonded,
non-tensioned steel tends to distribute cracks and restrict their progression. Increased flexural rigidity and
reduced deflections under cracked
conditions are thus realized (see Fig.
6).
8. The total deflection of a prestressed concrete beam containing
non-tensioned steel, when compared
to the deflection of an identical beam
without non-tensioned steel, may be
greater if the applied load is equal
to the cracking load; comparable if
the applied load is slightly larger
than the cracking load; or considerably smaller if the applied load is
considerably larger than the cracking load.
9. Due to the similarity between
the behavior of ordinary reinforced
concrete and prestressed concrete
under cracked conditions, the methods uscd for computing deflections
of ordinary reinforced concrete
members may be applied to the deflections of cracked prestressed concrete members. This is accomplished
by properly defining the cracking
moment and -the effective moment
of inertia (see Eqs. (35) through
(38) and Table 6). This method predicted deflections up to 80 percent
of the ultimate load within 19 percent of the measured values in all
cases.
10. For all normal provisions of
non-tensioned steel, yielding (even
for a 33 ksi yield steel) occurs close
to the ultimate load and deflections
at loads of about 80 percent of the
ultimate load may be computed by
34

assuming that the non-tensioned

steel has not yielded.
11. Regarding the contribution of
non-tensioned steel to the ultimate
strength of an under-reinforced prestressed concrete beam, the usual
practice of considering that the nontensioned steel provides a tension
force equal to its area times its stress
at ultimate is satisfactory.
12. The selection of type and
quantity of non-tensioned steel
should be based on the behavior desired under various service conditions: desired reduction in time-dependent camber, acceptable
deflections under working loads,
desirability of limiting deflections
underoverloads, and the required
factor of safety against failure.
13. The only unfavorable effects
appear to be the possibility of greater deflections under working loads
(see conclusions 6, 7, 8). In general,
non-tensioned steel affords a powerful means which, with proper judgment, can be used to meet even the
severest serviceability and safety requirements of prestressed concrete
beams.
NOTATION

A9 = area of gross concrete seetion

At = area of uncracked transformed concrete section
A$ = area of prestressing steel
A8 = area of non-tensioned
steel
B t = camber coefficient for
prestressed concrete beam
defined as the ratio of
time-dependent camber to
initial camber
Ct = creep coefficient defined
as the ratio of creep strain
to initial strain
EP = modulus of elasticity of
concrete at the time of
prestressing
PCI Journal

Eb

= internal strain energy of

a beam
e
= eccentricity of steel
Fo
= prestress force at release
dF t = loss of prestress force at
time, t
Ft
=force in non-tensioned
steel at time, t
f^
= 28-day concrete strength
modulus of rupture of
concrete
= nominal ultimate strength
fs
of prestressing steel
= calculated stress in pref8
stressing steel at ultimate
load
f9
=nominal yield strength of
steel
h
= total depth of a beam
I
= moment of inertia
I
= moment of inertia of gross
concrete section
It
= moment of inertia of uncracked transformed concrete section
= moment of inertia of
Ic,.
cracked transformed concrete section
Ieff = effective moment of inertia of concrete section
k
= coefficient determining
the depth of neutral axis
under cracked conditions
= prestress loss coefficient
k 0 ,, = reduction of concrete
force coefficient
L
= beam span (center to center of supports)
M
= bending moment
= cracking moment
M max = maximum moment in a
beam
n
= modular ratio: n = E 3 /E c ;

n' = E8 /E,

= ratio of area of steel to

area of concrete: p = A S /
bd; p' = A /bd (p33 for 33
ksi yield steel, p60 for 60
ksi yield steel and p' for

February 1970

Wd
W8

w
y
Yt
a

a,
8h

f3

fish

ye

high strength steel)

= work done against dead
load of a beam
= work done by force in
steel
= uniform distributed beam
dead load
= camber of a beam
= distance of extreme tension fiber from centroid
of concrete section
= modification factor for
combined time-dependent
camber coefficient
= modification factor for
creep coefficient
=modification factor for
shrinkage strain
= ratio of loss of prestress
in a beam with non-tensioned steel to the loss of
prestress in a beam without non-tensioned steel
= ratio of loss of prestress
due to creep in a beam
with non-tensioned steel
to the loss of prestress in
a beam without non-tensioned steel
= ratio of loss of prestress
due to shrinkage in a
beam with non-tensioned
steel to the loss of prestress in a beam without
non-tensioned steel
= ratio of reduction of the
concrete force in a beam
with non-tensioned steel
to reduction of the conCrete force in a beam
without non-tensioned
steel
= ratio of reduction of the
concrete force due to
creep in a beam with nontensioned steel to reduction of the concrete force
in a beam without nontensioned steel
35

ya, = ratio of the reduction of

the concrete force due to
shrinkage in a beam with
non-tensioned steel to reduction of concrete force
in a beam without nontensioned steel
ec
t
= initial concrete strain at
the level of steel
= curvature or angle change
(
per unit length of the
beam
ACKNOWLEDGMENTS
This investigation was conducted
in the Materials Testing Laboratory,
Department of Civil Engineering,
University of Iowa. The project was
sponsored by the Iowa State Highway Commission and the Bureau of
Public Roads. Steel used in the
preparation of the test beams was
donated by CF & I Corporation and
Armco Steel Corporation.
REFERENCES
1. Abeles, P. W., "Static and Fatigue
Tests on Partially Prestressed Concrete
Constructi on," ACI Journal, Proceedings Vol. 50, No. 7, Dec. 1954, pp.
361-376.
2. Abeles, P. W., "Partial Prestressing
and Possibilities for Its Practical Application," PCI Journal, Vol. 4, No. 1,
June 1959, pp. 35-51.
3. Abeles, P. W., "Partial Prestressing in
England," PCI Journal, Vol. 8, No. 1,
Feb. 1963, pp. 51-72.
4. Abeles, P. W., "Studies of Crack
Widths and Deformation Under Sustained and Fatigue Loading," PCI
Journal, Vol. 10, No. 6, Dec. 1965.
5. Magura, D. and Hognestad, E., "Tests
of Partially Prestressed Concrete Girders," Journal ASCE Structural Division, Proceedings Vol. 92, No. ST1,
Feb. 1966, pp. 327-343.
6. Burns, N. H., "Moment Curvature Relationships for Partially Prestressed
Concrete Beams," PCI Journal, Vol. 9,
No. 1, 1964, pp. 52-63.
7. Hutton, S. C. and Loov, R. E., "Flexural Behavior of Prestressed, Partially
Prestressed and Reinforced Concrete
Beams," ACI Journal, Proceedings Vol.
63, No. 12, Dec. 1966, pp. 1401-1408.
36

8. Branson, D. E. and Ozell, A. M.,

"Camber in Prestressed Concrete
Beams," ACI Journal, Proceedings Vol.
57, No. 12, June 1961, pp. 1549-1574.
9. "Deflections of Prestressed Concrete
Members," ACI Committee 335, Subcommittee 5 Report, ACI Journal,
Proceedings Vol. 60, No. 12, Dec.
1963, pp. 1697-1728.
10. ACI-ASCE Joint Committee 323,
"Tentative Recommendations for Prestressed Concrete," ACI Journal, Proceedings Vol. 54, No. 7, Jan. 1958, pp.
545-578.
11. Zia, P. and Stevenson, J. F., "Creep
of Concrete Under Non-Uniform Stress
Distribution and Its Effect on Camber
of Prestressed Concrete Beams," North
Carolina State Highway Commission
and Bureau of Public Roads, Project
ERD-110R.
12. Shaikh, A. F., "Use of Non-Tensioned
Steel in Prestressed Concrete Beams,"
Ph.D. Thesis, University of Iowa, August 1967,
13. Branson, D. E. and Shaikh, A. F.,
"Favorable and Unfavorable Effects of
Non-Tensioned Steel in Prestressed
Concrete Beams", Iowa State Highway Commission Research Project No.
HR-123 and Bureau of Public Roads
No. HPR-1 (3) (Iowa), June 1967.
14. "Deflections of Reinforced Concrete
Members," ACI Committee 435 Report, ACI Journal, Proceedings Vol. 63,
No. 6, June 1966.
15. "Deflection of Reinforced Concrete
Members," Progress Report of ACI
Committee 307, ACI Journal, Proceedings Vol. 27, 1931, p. 351.
16. "Deflection of Reinforced Concrete
Members," Bulletin ST-70, Portland
Cement Association, 1947.
17. Yu, Wei-Wen and Winter, George,
"Instantaneous and Long-Time Deflections of Reinforced Concrete Beams
Under Working Loads," ACI Journal,
Proceedings Vol. 57, No. 1, July 1960,
pp. 29-50.
18. Branson, Dan E., "Instantaneous and
Time-Dependent Deflections of Simple and Continuous Reinforced Concrete Beams," Report No. 7, Alabama
Highway Research Report, Bureau of
Public Roads, Aug. 1963, (1965).
19. Bewtra, S. K., "A Study of Different
Methods for Predicting Short-Time
and Long-Time Deflections of Reinforced Concrete Beams," MS Thesis,
University of Iowa, Aug. 1964.
PCI Journal