Sie sind auf Seite 1von 0

86 ACI Structural Journal/January-February 2000

ACI Structural Journal, V. 97, No. 1, January-February 2000.


Received September 4, 1998, and reviewed under Institute publication policies.
Copyright 2000, American Concrete Institute. All rights reserved, including the
making of copies unless permission is obtained from the copyright proprietors. Perti-
nent discussion will be published in the November-December 2000 ACI Structural
Journal if received by July 1, 2000.

ACI STRUCTURAL JOURNAL TECHNICAL PAPER
A simplenoniterativemethod for predicting theultimatestrength
and themodeof failureof reinforced concretemembraneelements
subjected to pureshear is presented. This method is a simplification
of the results of the modified compression field theory (MCFT),
which is thebasis of an alternativedesign method in thelatest edi-
tions of AASHTO LRFD Specifications and theCanadian Build-
ing Code(CSA-A23.2). Theshear strength is related to theamounts
of transverse and longitudinal reinforcement and to the concrete
strength in the element. To check the accuracy of the method, the
observed shear capacity and modeof failureof 46 reinforced concrete
panels subjected to pureshear arecompared with theresults of the
method, and good agreement is obtained. Based on thenew method,
a limit on themaximumamount of reinforcement required to insure
yielding is proposed. Part II of this series of two papers extends the
applicability of theproposed method to beams subjected to shear com-
bined with bending moment and axial load.
Keywords: failure; membranes; reinforced concretes; shear; strength;
trusses.
BACKGROUND
Research on the shear behavior of reinforced concrete beams
started near the end of the 19th century
1
and led to the devel-
opment of the famous 45 degree truss model. This model as-
sumes that, after cracking of the concrete, the behavior of a
reinforced concrete beam becomes analogous to that of a truss
with a top longitudinal compression chord, a bottom longitu-
dinal tension chord, vertical steel ties, and diagonal concrete
struts inclined at 45 degrees (Fig. 1). The model assumes that
the diagonally cracked concrete cannot resist tension, and this
assumption eliminates the need for diagonal tension members
perpendicular to the concrete struts. I t is also assumed that the
shear force is resisted by the transverse steel. The steel stress

s
is given by
(1)
where

t
= ratio of transverse reinforcing steel; and
f
yt
= yield strength of transverse steel.
Comparing the results of the 45 degree truss model with ex-
perimental results showed that this model underestimates the
shear strength, especially in lightly reinforced members. To im-
prove the correlation between the model and experimental re-
sults, an empirical concrete contribution
c
was added to the steel
contribution
s
(given in Eq. (1)) to form what became known as
the modified 45 degree truss model. This model is the basis for
the shear design equations in many design codes such as the ACI
Code.
2
The popularity of the ACI shear design method comes
mainly from its simplicity. Because this method is semi-empirical,
however, eight different equations for
c
and numerous limits are
required to account for the effects of the amount of longitudinal

s

t
f
yt
=
reinforcement, the bending moment, axial tension or compres-
sion, and prestressing. In addition, the
c
equations poorly ac-
count for the effects of longitudinal reinforcement and bending
moment. I n fact, the ACI -ASCE Committee on Shear and Diag-
onal Tension
3
recommended that ACI
2
Eq. (11-5), which ac-
counts for these two parameters, no longer be used.
For members not subjected to axial loads, the ACI
2
expres-
sion for
c
is given by Eq. (11-5). For pure shear, this expres-
sion is reduced to
(2)
where f
c
is the specified compressive strength of the concrete.
On the other hand, more rational models for shear behavior
have been under development, especially in the past 30 years.
Most notably, the modified compression field theory
(MCFT)
4-7
is capable of predicting the full response of sections
subjected to a combination of shear, bending, axial load, and
torsion. A simplified method based on the MCFT, namely, the
general method, has been adopted by the 1994 Canadian Build-

c
0.292 f
c
MP a =
Title no. 97-S10
Shear Strength of Reinforced Concrete: Part IMembrane
Elements Subjected to Pure Shear
by Khaldoun N. Rahal
Fig. 1Truss model for reinforced concretebeams in shear.
87 ACI Structural Journal/January-February 2000
ing Code
8
and the 1994 AASHTO LRFD Specifications
9
as an
alternative to the traditional modified 45 degree truss model.
This method unifies the design for shear, combining the ef-
fects of prestressing, bending and torsional moments, and axial
load all in one set of equations. A recent study
10
showed that
this general and unified method is as conservative as the ACI
modified 45 degree truss method, yet it gives a narrower scat-
ter relative to experimental results, and therefore provides a
more uniform factor of safety. The simplification of the equa-
tions of the MCFT into the general method involved assump-
tions and considerations, which were detailed by Collins et al.
10
and Collins and Rahal.
11
Five years after adopting the general method, the traditional
modified 45 degree truss model remains the more popular design
method for two reasons. First, the traditional method is easier to
apply in the calculation of the shear capacity mainly in nonpre-
stressed members subjected to shear, bending moment, and axial
tension. I n fact, the general method was shaped to suit design
cases instead of capacity calculations. Second, design using the
general method is perceived to be a time-consuming iterative
process, even though designers who gain minimal experience
with the method are able to make a sufficiently accurate first
guess and do not need to iterate for the correct solution.
This paper proposes an alternative simplification of the
MCFT to produce a noniterative method for calculating the ul-
timate strength and the mode of failure of reinforced concrete
membrane elements subjected to pure shear. The proposed
method blends the simplicity of ACI s modified 45 degree truss
equations with the accuracy and rationality of the MCFT. I n
calculating the strength, this method accounts for the effects of
the amounts of longitudinal and transverse reinforcement and
the concrete strength.
This paper is the first of a series of two papers dealing with
the subject of shear in reinforced concrete. Part I I
12
of this se-
ries extends the proposed approach to beams subjected to
shear, axial load, and bending moment.
RESEARCH SIGNIFICANCE
This paper proposes a new simplified approach for calculat-
ing the shear strength and mode of failure of reinforced con-
crete membrane elements. The proposed method combines the
simplicity of the ACI Code equations and the accuracy of the
modified compression field theory. Similar to the general meth-
od, the proposed method has the potential of accounting for the
effects of axial loads, bending and torsional moments, and pre-
stressing in reinforced concrete beams.
MODIFIED COMPRESSION FIELD THEORY FOR
SHEAR
The MCFT is a behavioral model that uses the equations of
equilibrium of forces, compatibility of average deformations,
and realistic stress-strain relationships of uncracked and
cracked reinforced concrete to calculate the full response of
sections subjected to combined shear, moment, torsion, and ax-
ial load (tension or compression). The calculated response in-
cludes precracking, postcracking, and postultimate behavior.
Detailed information on the stresses and strains of the concrete
and steel is also obtained. This method is similar in its gener-
ality to the well-known flexure theory. Details on the MCFT
are not provided in this paper. Readers are referred to Refer-
ences 4 through 7 for full details on this theory.
SIMPLIFIED STRENGTH CURVES
Figure 2 shows a concrete element reinforced in the longitu-
dinal and transverse directions and subjected to shear. Experi-
mental evidence showed that the shear strength of members
properly reinforced in both directions depends on three main
parameters: the amount of transverse reinforcement, the
amount of longitudinal reinforcement, and the concrete
strength. The MCFT quantifies the effects of each of these pa-
rameters and identifies other less significant parameters such
as the aggregate size and the crack control characteristics of
the reinforcement. These two parameters become more signif-
icant in members reinforced only in one direction.
5,10
The major parameters affecting the shear strength are com-
bined in the following two nondimensional indexes
(3)
(4)
where
= reinforcement ratio; and
f
y
= yield strength of reinforcing bars.
The subscripts t and L refer to transverse and longitudinal, re-
spectively. The index , commonly referred to as the mechan-
ical reinforcement ratio, may be thought of as the maximum
force that the steel can resist relative to the maximum force
that concrete can resist.
The equations of the MCFT were implemented in a computer
program, which can be used to calculate the ultimate shear
strength v
u
of members with different
L
and
t
. Results show
that the normalized ultimate stress
u
/ f
c
is independent of the
value of f
c
used in the analysis, and hence the effects of the
three major parameters
L
,
t
, and f
c
can be combined in a sin-
gle graph. I t is to be noted that a parabolic stressstrain rela-
tionship for concrete in compression was assumed in the
formulation of the computer program.
Figure 3 shows a plot of
u
/ f
c
versus
t
curves for different
values of
L
. Each curve represents a relationship between
t
and
u
/ f
c
at a specific
L
. The choice of the maximum aggregate

L
f
y L
f
c
------------ =

t
f
yt
f
c
---------- =
ACI member Khaldoun N. Rahal is an assistant professor in theDepartment of Civil
Engineering, Facultyof Engineering and Petroleum, Kuwait University. Heholdsdegrees
fromtheAmerican Universityof Beirut, theUniversityof Michigan, Ann Arbor, Mich., and
theUniversityof Toronto. Heisa director of theACI-Kuwait Chapter.
Fig. 2Reinforced concretemembraneelement subjected to shear.
88
size and the spacing of the cracks slightly affect the results of
the figure at low reinforcing levels. Similar to the general
method,
5,8-10
a 300 mm crack spacing is used because it is be-
lieved that this value is appropriate for the full range of beams
containing stirrups.
10
Unlike the general method, which uses
a 19 mm maximum aggregate size, a conservative value of 10
mm is used in the proposed method.
Figure 3 shows that for a given level of
L
, the strength in-
creases with an increase in the index of transverse reinforce-
ment
t
. The rate of increase, however, drops beyond a certain
level of
t
, where crushing in the concrete occurs before yield-
ing of the steel. Similar behavior is observed in flexural mem-
bers when the amount of longitudinal reinforcement exceeds
the maximum value. I ncreasing the amount of longitudinal
reinforcement has a similar effect on the shear strength due to
the symmetry of the problem of pure shear. I n fact, the input
values of
L
and
t
can be exchanged without affecting the val-
ue of
u
/ f
c
obtained from Fig. 3.
At low levels of reinforcement (low
L
and
t
), the strength
is governed by the cracking strength in the concrete, taken as
0.33f
c
MPa for members with f
c
50 MPa. This lower limit
plots as a horizontal line for every curve in Fig. 3, but different
concrete strengths have different lines because the coordinate
axis in the figure is a function of f
c
instead of f
c
. To maintain
clarity, this limit is not shown in Fig. 3; however, values of
u
obtained from the figure should be compared to the cracking
shear stress, and the larger value should be retained.
MODE OF FAILURE
When calculating the shear strength, it is advantageous to
also calculate the mode of failure (MOF) of the element. The
mode of failure depends on whether the steel yielded at ultimate
load or not. I f the steel in both directions yields, the element is
considered under-reinforced. I f the steel in both directions does
not yield, the element is considered over-reinforced. I f steel in
one direction only yields, the element is considered partially un-
der-reinforced (or partially over-reinforced).
Two yield curves are plotted in Fig. 3, corresponding to the
amounts of reinforcement where
t
=
yield
and
L
=
yield
.
These curves divide Fig. 3 into four regions where:
1.
t
>
yield
and
L
>
yield
. I n this case, both the longitudinal
as well as the transverse reinforcement yield when the ultimate
strength is reached (under-reinforced);
2.
t
>
yield
and
L
<
yield
. I n this case, only the transverse
reinforcement yields when the ultimate strength is reached
(partially under-reinforced);
3.
t
<
yield
and
L
>
yield
. I n this case, only the longitudinal
reinforcement yields when the ultimate strength is reached
(partially under-reinforced); and
4.
t
<
yield
and
L
<
yield
. Here the ultimate capacity is gov-
erned by crushing of the concrete before yielding in any of the
reinforcement (over-reinforced).
USE OF CURVES
An advantage of the proposed method is the simplicity of the
predictions of the shear capacity and the mode of failure. To
calculate the ultimate shear strength, three steps are followed:
1. Calculate
L
and
t
using Eq. (3) and (4);
2. Enter Fig. 3 with the
L
and
t
values and obtain
u
/ f
c
; and
3. I f
u
is smaller than the cracking stress, take
u
equal to
this cracking stress.
The relative location of the (
L
and
t
) point with respect to the
yield curves in Fig. 3 indicates if the transverse and/ or longitudi-
nal reinforcement yield when the ultimate strength is reached.
COMPARISON WITH EXPERIMENTAL RESULTS
The experimental results from 46 panels tested under pure
shear at the University of Toronto
4,13-17
and the University of
Houston
18-20
are used to check the accuracy of the method.
The observed shear capacities of the 46 panels are compared
with the calculations of the proposed method, and the results
are shown in Table 1. These test specimens had a wide range
of concrete strength and amounts of longitudinal and trans-
verse reinforcement. The concrete strength ranged from 11.6
to 103.1 MPa (1680 to 15,000 psi), while the amounts of rein-
forcement ranged between 2.7 and 70% for
t
and
L
.
Vecchio and Collins
4
tested seven 890 x 890 x 70 mm rein-
forced concrete panels containing a relatively large amount of
longitudinal reinforcement (
L
= 0.01785) and varying
amounts of transverse reinforcement (
t
= 0.00315 to 0.01785).
The concrete strength for these specimens ranged from 16 to
20.5 MPa, with an average of 19.1 MPa. Table 1 shows details
of the amount and strength of the six specimens. The mean
L
value for these specimens was 42.4%, and
t
ranged from 6.66
to 38.5%.
Figure 4(a) shows a comparison between the observed values
of the normalized ultimate stress
u
/ f
c
and those calculated us-
ing Fig. 3 (based on
L
= 43%). A good agreement is observed.
Fig. 3Normalized shear strength
u
/ f
c
for reinforced concretemembraneelements.
ACI Structural Journal/January-February 2000 89
Table 1Experimental verification of method
Specimen
f
c
, MPa
L
, %
t
, %
Experiment Proposed method

exp
/
calc

exp
, MPa
Mode of failure
*

calc
, MPa
Mode of failure
*
Proposed method General method ACI Code
PV18
4
19.5 39.50 6.67 3.04 TY 2.30 TY 1.32 1.20 1.18
PV12
4
16.0 52.30 7.50 3.13 TY 2.30 TY 1.36 1.33 1.32
PV19
4
19.0 43.00 11.20 3.95 TY 3.42 TY 1.15 1.12 1.16
PV20
4
19.6 42.00 13.40 4.26 TY 3.92 TY 1.09 1.04 1.09
PV21
4
19.5 42.00 20.00 5.03 TY 4.88 TY 1.03 1.03 1.19
PV22
4
19.6 41.70 32.70 6.07 CC 5.98 CC 1.02 1.23 1.43
PV27
4
20.5 38.50 38.50 6.35 CC 6.31 CC 1.01 1.23 1.46
PV16
4
21.7 8.70 8.70 2.14 LY, TY 1.89 LY, TY 1.13 1.11 0.66
PV4
4
26.6 9.60 9.60 2.89 LY, TY 2.55 LY, TY 1.13 1.10 0.71
PV3
4
26.6 12.00 12.00 3.07 LY, TY 3.19 LY, TY 0.96 0.93 0.65
PV6
4
29.8 16.00 16.00 4.55 LY, TY 4.77 LY, TY 0.95 0.90 0.87
PV9
4
11.6 70.00 70.00 3.74 CC 3.94 CC 0.95 1.29 1.15
PV10
4
14.5 34.00 19.00 3.97 TY 3.45 TY 1.15 1.09 1.09
PV11
4
15.6 26.90 19.70 3.56 LY, TY 3.63 LY, TY 0.98 0.98 0.94
J A1
13
49.9 17.30 8.60 6.34 NA, TY 6.09 LY, TY 1.04 1.03 1.00
J A2
13
43.0 20.00 10.0 6.22 NA, TY 6.02 LY, TY 1.03 1.02 1.00
PHS2
13
66.1 29.60 3.20 6.66 TY 5.29 TY 1.26 1.49 1.48
PHS3
13
58.4 33.50 7.30 8.19 TY 8.06 TY 1.02 1.21 1.26
PHS8
13
55.9 35.00 11.60 10.80 TY 10.10 TY 1.07 1.12 1.51
*
LY = longitudinal steel yields; TY: transversesteel yields; CC: concretecrushing; and NA: not available.
Fig. 4Comparison between observed and predicted strength and modeof failurefor normal strength and high-strength specimens.
90
The proposed method predicts that in specimens with
t
<
26%, the transverse reinforcement yields before the concrete
crushes. This is again in agreement with the observed mode of
failure of the specimens. The calculations of the ACI
2
and the
general method
8-10
are also shown. For this series of tests,
both the ACI and the general method accurately predict the
strengths, but underestimate the load that causes concrete
crushing in the panels.
Vecchio and Collins
4
also tested five reinforced concrete
panels where the amounts and strength of the longitudinal and
transverse reinforcement remained the same in every specimen
but varied amongst different specimens. The reinforcement
ranged from
L
=
t
= 0.00740 to 0.01785. The concrete
strength for these specimens ranged from 11.6 to 29.8 MPa;
(Table 1). Figure 4(b) shows a comparison between the ob-
served and the calculated shear capacities for the five speci-
mens. A good agreement is observed, but the method slightly
overestimated the crushing strength of the over-reinforced el-
ement. The new method also predicts that in specimens with

L
=
t
< 0.27, the reinforcement yields before the concrete
crushes. This is again in agreement with the observed mode of
failure of the five specimens. Again the general method accu-
rately predicted the behavior but slightly underestimates the
crushing stress. Since the concrete strength varied significant-
ly among the five specimens, the ACI calculations, which are
based on f
c
, are not shown in Fig. 4(b).
Table 1 also shows details of two specimens (PV10 and
PV11) tested by Vecchio and Collins,
4
and five specimens (J A1,
J A2, PHS2, PHS3, and PHS8) tested by Vecchio, Collins, and
Aspiotis,
13
who tested panels of relatively higher concrete
strength (43 to 66.1 MPa). Khalifa and Kirschner,
14,15
Bieder-
mann,
16
and Porasz
17
tested, respectively, two elements (SE1,
SE5), one element (SE11), and three elements (SE12, SE13,
and SE14), which were 1500 x 1500 mm. The depth of these el-
ements ranged between 285 and 310 mm, and the concrete
strength ranged from 25.9 to 80.5 MPa; refer to Table 1. Table 1
also shows details of twenty-one 1397 x 1397 x 178 mm spec-
imens (VA0 to VA4, VB1 to VB4, B1 to B6, HB1, HB3, HB4,
and A2 to A4) tested by Hsu and Zhang.
18-20
The concrete
strength ranged from 41.3 to 103.1 MPa, and the amounts of
reinforcement ranged from = 0.0023 and 0.0524.
The concrete strength of Specimens VA0 to VA4 tested by
Hsu and Zhang
18
averaged 98 MPa. The amounts and
strength of the longitudinal and transverse reinforcement re-
mained the same in each individual specimen but varied among
different specimens. The reinforcement ratios ranged from
L
Table 1 (cont.)Experimental verification of method
Specimen
f
c
, MPa
L
, %
t
, %
Experiment Proposed method

exp
/
calc

exp
, MPa
Mode of failure
*

calc
, MPa
Mode of failure
*
Proposed method General method ACI Code
SE1
14,15
42.5 34.00 11.00 6.77 TY 7.43 TY 0.91 0.94 1.08
SE5
14,15
25.9 55.70 54.20 8.10 CC 8.60 CC 0.94 1.25 1.66
SE11
16
70.1 20.00 6.30 6.60 TY 7.85 LY, TY 0.84 0.96 0.96
SE12
17
75.9 17.90 5.70 7.40 TY 7.59 LY, TY 0.98 1.09 1.08
SE13
17
80.5 39.40 11.2 12.00 TY 14.50 TY 0.83 0.88 1.40
SE14
17
60.4 39.30 39.3 16.90 CC 18.70 CC 0.90 1.11 2.27
VA0
18
98.8 2.70 2.70 3.35 LY, TY 2.67 LY, TY 1.26 1.25 0.60
VA1
18
95.1 5.61 5.61 6.16 LY, TY 5.34 LY, TY 1.15 1.15 0.75
VA2
18
98.2 9.95 9.95 9.73 LY, TY 9.78 LY, TY 1.00 0.97 1.02
VA3
18
94.6 17.30 17.30 15.10 LY, TY 16.30 LY, TY 0.92 0.88 1.62
VA4
18
103.1 23.90 23.90 21.40 CC 24.60 LY, TY 0.87 0.85 2.20
VB1
18
98.2 5.40 9.95 7.50 LY, TY 7.07 LY, TY 1.06 1.06 0.79
VB2
18
97.6 5.50 16.70 9.14 LY, TY 8.88 LY, TY 1.03 1.12 0.97
VB3
18
102.3 5.20 27.50 9.71 TY 11.00 TY 0.88 1.03 1.00
VB4
18
96.9 2.75 8.45 4.86 LY, TY 4.85 LY, TY 1.00 1.03 0.52
B1
19
45.3 12.20 5.90 3.97 LY, TY 3.85 LY, TY 1.03 1.08 0.86
B2
19
44.1 18.10 12.60 6.13 LY, TY 6.61 LY, TY 0.93 0.91 0.96
B3
19
44.9 17.80 5.90 4.37 LY, TY 4.58 LY, TY 0.95 1.01 0.95
B4
19
44.8 31.20 6.00 5.08 TY 5.38 TY 0.94 1.06 1.09
B5
19
42.9 32.60 12.90 7.17 TY 8.15 TY 0.88 0.87 1.14
B6
19
43.0 32.60 18.60 9.15 TY 10.00 TY 0.91 0.90 1.46
HB1
19
66.5 7.40 4.00 4.32 LY, TY 3.60 LY, TY 0.83 1.17 0.86
HB3
19
66.8 12.00 4.00 4.89 LY, TY 4.60 LY, TY 1.06 1.09 0.97
HB4
19
62.9 22.30 4.24 5.33 TY 5.84 TY 0.91 1.07 1.07
A2
20
41.3 13.30 13.30 4.92 LY, TY 5.49 LY, TY 0.90 0.86 0.80
A3
20
41.7 19.20 19.20 6.99 LY, TY 8.00 LY, TY 0.87 0.83 1.13
A4
20
42.5 33.00 33.00 11.38 NA, NA 12.40 CC 0.91 1.07 1.82
Average 1.01 1.07 1.13
Coefficient of variation, % 12.5 13.3 32.9
*
LY = longitudinal steel yields; TY: transversesteel yields; CC: concretecrushing; and NA: not available.
ACI Structural Journal/January-February 2000 91
=
t
= 0.0060 to 0.0524 and the concrete strength ranged from
95.1 to 103.1 MPa; refer to Table 1.
Figure 4(c) shows a comparison between the observed and the
calculated shear capacities for the five specimens. A good agree-
ment is observed. The proposed method also predicts that in
specimens with
L
=
t
< 0.27, the reinforcement yields before
the concrete crushes. This is again in agreement with the ob-
served mode of failure of four specimens, but was slightly off for
specimen VA4. Again, the general method accurately predicted
the behavior and gave relatively similar results to the proposed
method. The ACI method overestimated the strength of lightly
reinforced elements and underestimated the strength of heavily
reinforced elements. Figure 4(c) clearly shows the inadequacy of
the ACI equations for high-strength concrete.
For the 46 specimens, the average experimental-to-calculated
ultimate shear stress (
exp
/
calc
) for the proposed method is
1.01, and the coefficient of variation is 12.5%; refer to Table 1.
The corresponding values are 1.13 and 32.9% for the ACI
method, and 1.07 and 13.3% for the general method. I t can be
concluded that while the simplicity of the proposed method is
comparable to that of the ACI method, its accuracy is compara-
ble to that of the general method. Both the proposed method
and the general method are significantly more accurate than
the ACI method.
Figure 5 shows three plots between (
exp
/
calc
) calculated
using the three methods versus the concrete strength. The pro-
posed method and the general method accurately calculate the
shear capacity of normal strength as well as high-strength con-
crete. The ACI results showed a larger scatter, especially at
high concrete strength.
The mode of failure (MOF) of 43 out of the 46 specimens was
available. Figure 6 shows a comparison between the observed
MOF and the results of the proposed method for the 40 speci-
mens. The position of a point with coordinates
t
and
L
with
respect to the yield lines indicates if the longitudinal and/ or
transverse steel yield. The observed results fit accurately with-
in the four regions bounded by the yield curves of the proposed
method (shown in dark solid line). The proposed method cor-
rectly predicts the MOF of 40 of the 43 specimens. Not shown
in Fig. 6, the general method correctly predicts the MOF of 35
out of the 43 specimens. ACI does not provide a check on the
yielding of the longitudinal reinforcement, and hence, a general
MOF cannot be obtained.
The state of transverse reinforcement for 45 of the 46 spec-
imens was also available (yielding or nonyielding). The ACI
method correctly predicts the state of the transverse rein-
forcement in 27 out of 45 specimens. The proposed method
and the general method correctly predict the state of trans-
verse reinforcement in 44 and 42 specimens, respectively.
Again, the results of the proposed method and the general
method are significantly more reliable than those of the ACI
method. The proposed method is, however, much simpler
than the general method.
SYMMETRICAL REINFORCEMENT
Figure 3 shows that for under-reinforced membranes with
equal transverse and longitudinal reinforcement (
L
=
t
), the
normalized shear stress
u
/ f
c
is equal to the reinforcement index
Fig. 6Observed and predicted modeof failureof 43 test specimens.
(Note: TY: transversesteel yielding; LY: longitudinal steel yielding;
and CC: concretecrushing.)
Fig. 5Ratio (
exp
/
calc
) versus f
c
for 46 test specimens.
92
(5)
that can also be presented as
(6)
As an example, for
L
=
t
= 0.15, the term
u
/ f
c
= 0.15. This
relationship applies to reinforcement ratios smaller or equal to
0.27, after which the element is considered over-reinforced.
Sixteen of the test results shown in Table 1 were symmetri-
cally reinforced, and 10 of them failed after the steel had yielded
(under-reinforced). Figure 7 shows the ratio of
exp
/ f
c
plotted
against the reinforcement index for these 10 specimens. The
results from Eq. (5) and (6) are in very good agreement with
the experimental trends.
This equality is due to the fact that, for symmetrically rein-
forced membranes, the angle of inclination of the diagonal
struts is equal to 45 degrees. Consequently, the steel stress
contribution typically taken as
t
f
yt
cot can be taken as
t
f
yt
.
I n under-reinforced sections, failure occurs due to sliding
across the cracks, which cannot be controlled by increased steel
stresses at crack locations due to the yielding of the reinforce-
ment. Under these conditions, the concrete contribution di-
minishes to zero, and the resistance comes solely from the steel,
given by
t
f
yt
(Eq. (6)).
MAXIMUM REINFORCEMENT
An upper limit on the amount of reinforcement or on the ap-
plied shear stress ensures yielding of the reinforcement before
crushing of the concrete. Figure 6 shows the maximum values
of the reinforcement indexes
L
and
t
required to satisfy this
condition.
To satisfy serviceability requirements related to limiting the
crack width at service load, the ACI equations, which are based
on the modified 45 degree truss model, set an upper limit on
s
equal to 0.67f
c
. This limit is also a safeguard against concrete
crushing before steel yielding, which in turn justifies the use of
the yield stress in the steel contribution expression
t
f
yt
. This
upper limit plots as a horizontal line in Fig. 6. Two limits are
shown for f
c
= 20 MPa and for f
c
= 50 MPa.
The general method sets an upper limit by
u
/ f
c
= 0.25. Col-
lins and Rahal
11
give the background of this 0.25 value. This
limit is not shown in Fig. 6.

u
f
c
-----
L

t
= =

u

t
f
yt

L
f
y L
= =
I n light of the good correlation between the MOF observed
in the tests and that predicted using the proposed model, it can
be concluded from Fig. 6 that the ACI upper limit is consider-
ably conservative, especially at relatively high concrete
strength. From Fig. 6, limiting the transverse steel index
t
(which is equal to
s
/ f
c
) appears to be a suitable means of
avoiding over-reinforcement
(7)
which is conservative for the large range of
L
values. Due to
the symmetry in the shear problem, a similar limit is set on
L
.
I t is noted that the limit set by the general method
5,8-10
is on
the total shear stress where
u
/ f
c
= 0.25, which is practical in
a design procedure. Comparison against the experimental re-
sults suggests that limiting the steel shear stress or the total
shear stress to a percentage of f
c
seems more appropriate than
limiting it to a percentage of f
c
.
CONCLUSIONS
This paper presented a simple noniterative method to predict
the ultimate shear strength and the mode of failure of reinforced
concrete membrane elements. The method is based on the re-
sults of the modified compression field theory, which is the basis
of one alternative method for shear design in AASHTO and the
Canadian CSA-A23.2 codes. This method takes into account the
effects of the concrete strength, and the amounts of transverse as
well as longitudinal reinforcement on the shear strength.
Comparison between the results of the method and experi-
mental results on 46 panels subjected to pure shear showed a
very good agreement. The method accurately accounted for
variation in the concrete strength and amount of longitudinal
and transverse reinforcement.
The proposed method suggests setting
s
/ f
c
< 1.25
L
+ 0.14
0.25 as a suitable limit to avoid excessive transverse reinforce-
ment amounts which will not yield before concrete crushing.
Results from the ACI method and AASHTOs general method
were also presented, and it was found that the accuracy of the
proposed method and the general method was superior to that
of the ACI method. I t was also found that the general method
and the proposed method gave similar results. This was ex-
pected because both are simplifications of the MCFT. The pro-
posed method, however, is easier to use than the general
method.
A companion paper extends the applicability of the proposed
method to beams subjected to shear combined with axial load
(tension or compression) and bending moment.
ACKNOWLEDGMENTS
The research reported in this paper was made possible by a grant from
the Kuwait University Research Administration, Project No. EV-098. This
support is gratefully acknowledged. The comments of Daniel Kuchma of
the University of I llinois at Urbana-Champaign are also gratefully ac-
knowledged.
NOTATIONS
f
c
= specified compressive strength of concrete
f
cr
= cracking stress of concrete in tension
f
yL
= yield strength of longitudinal steel
f
yt
= yield strength of transverse steel
v
c
= contribution of concrete to shear resistance
v
s
= contribution of transverse reinforcement to shear resistance

L
= nondimensional longitudinal reinforcement index

t
= nondimensional transverse reinforcement index

L
= strain in longitudinal reinforcement

t
= strain in transverse reinforcement

yield
= yield strain in reinforcement

L
= ratio of total longitudinal reinforcing steel

t
= ratio of transverse reinforcing steel

t
1.25
L
0.14 0. 25 + =
Fig. 7Strength of symmetrically under-reinforced elements.
93 ACI Structural Journal/January-February 2000
REFERENCES
1. Ritter, W., DieBauweiseHennebique, SchweiserischeBauzeitung, Switzer-
land, 1899.
2. ACI Committee 318, Building Code Requirements for Structural Con-
crete (ACI 318-95) and Commentary (318R-95), American Concrete I nsti-
tute, Farmington Hills, Mich., 1995, 369 pp.
3. ACI -ASCE Committee 426, Suggested Revisions to Shear Provisions
for Building Codes (ACI 426.1R-77), American Concrete I nstitute, Farm-
ington Hills, Mich., 1979, 82 pp.
4. Vecchio, F. J ., and Collins, M. P., The Response of Reinforced Concrete
to I n-Plane Shear and Normal Stresses, Publication No. 80-03, the Univer-
sity of Toronto, 1982, 332 pp.
5. Collins, M. P., and Mitchell, D., Prestressed ConcreteStructures, Prentice
Hall, Englewood Cliffs, N.J ., 1991, 766 pp.
6. Vecchio, F. J ., and Collins, M. P., Modified Compression Field Theory
for Reinforced Concrete Elements Subjected to Shear, ACI J OURNAL, Pro-
ceedingsV. 83, No. 2, Mar.-Apr. 1986, pp. 219-231.
7. Rahal, K. N., and Collins, M. P., Analysis of Sections Subjected to Com-
bined Shear and TorsionA Theoretical Model, ACI Structural J ournal,V.
92, No. 4, J uly-Aug. 1995, pp. 459-469.
8. CSA Standard, Design of Concrete Structures (A23.3-94), Canadian
Standards Association, Rexdale, Ontario, Canada, 1994, 199 pp.
9. American Association of State Highway and Transportation Officials,
AASHTO LRFD BridgeDesign Specifications and Commentary, 1st Edition,
Washington D.C., 1994, 1091 pp.
10. Collins, M. P.; Mitchell, D.; Adebar, P.; and Vecchio, F. J ., A Gen-
eral Shear Design Method, ACI Structural J ournal, V. 93, No. 1, J an.-Feb.
1996, pp. 36-45.
11. Collins, M. P., and Rahal, K. N., Background of 1994 CSA-A23.3
General Method of Shear Design,Canadian J ournal of Civil Engineering, V.
26, No. 6, Dec. 1999, pp. 827-839.
12. Rahal, K., Shear Strength of Reinforced Concrete: Part I I Beams
Subjected to Shear, Bending Moment, and Axial Load, ACI Structural J our-
nal. (accepted for publication)
13. Vecchio, F. J .; Collins, M. P.; and Aspiotis, J ., High-Strength Concrete
Elements Subjected to Shear, ACI Structural J ournal, V. 91, No. 4, J uly-Aug.
1994, pp. 423-433.
14. Khalifa, J ., Limit Analysis and Design of Reinforced Concrete Shell
Elements, PhD thesis, University of Toronto, 1986.
15. Kirschner, U., I nvestigating Behavior of Reinforced Concrete Shell
Elements, PhD thesis, University of Toronto, 1986.
16. Biedermann, J ., The Design of Reinforced Concrete Shell Elements,
MASc thesis, University of Toronto, 1987.
17. Porasz, A., An I nvestigation of Stress-Strain Characteristics of High
Strength in Shear, MASc thesis, University of Toronto, 1989.
18. Hsu, T. T. C., and Zhang, L., Behavior and Analysis of 100 MPa Con-
crete Membrane Elements, J ournal of theStructural Division, V. 124, No. 1,
J an. 1998, pp. 24-34.
19. Hsu, T. T. C., and Zhang, L., Nonlinear Analysis of Membrane Ele-
ments by Fixed Angle Softened-Truss Model, ACI Structural J ournal, V. 94,
No. 5, Sept.-Oct. 1997, pp. 483-492.
20. Hsu, T. T. C., and Zhang, L., Tension Stiffening in Reinforced Con-
crete Membrane Elements, ACI Structural J ournal, V. 93, No. 1, Jan.-Feb.
1996, pp. 108-115.

Das könnte Ihnen auch gefallen