You are on page 1of 6

PERFORMANCE OF CODE EQUATIONS COMPARED TO EXPERIMENTAL

DATA FOR SHEAR CAPACITY OF FRP-RC BEAMS


A. Said
1
and H. El Chabib
2

1
Dept. of Civil and Env. Eng., University of Nevada, Las Vegas, Las Vegas, NV 89154, USA
2
Civil. Eng. and Construction, Bradley University, Peoria, Il 61625, USA



ABSTRACT

Current approaches for estimation of shear capacity of concrete beams reinforced with fiber-reinforced polymer
(FRP) are generally based on existing semi-empirical shear design equations for steel-reinforced concrete (S-RC).
These equations were primarily evaluated based on experimental data generated on concrete beams with steel
reinforcement. However, FRP materials have different mechanical properties and accordingly exhibit different
modes of failure than steel, making the extension of existing shear design equations for S-RC beams to cover
concrete beams reinforced with FRP somehow inaccurate. Currently available methods include ACI 440-06,
J SCE-97, CSA S806-02, and ISIS Canada-01. Availability of FRP reinforcement products varies in terms of
capacity and modulus of elasticity, which can result in a significant change in behavior. An experimental
database of 150 FRP-reinforced concrete (FRP-RC) beams was developped from published literature.
Subsequently, this database was used to assess the validity of these four main existing shear design methods for
FRP-RC beams. This research investigates the performance of the abovementioned design methods to estimate
the nominal shear capacity, V
n
of steel-free concrete beams reinforced with FRP bars. Results show that current
design guidelines provide a shear strength underestimation in the case of beams without shear reinforcement and
a shear strength overestimation for beams with shear reinforcement.

KEYWORDS

Shear strength, FRP-reinforced, beams, code, fiber reinforced polymers, rigidity.

INTRODUCTION

The use of advanced composite materials to substitute traditional reinforcing steel has been an increasing trend.
These materials have high strength to weight ratio, high durability and superior resistance to corrosion (ACI 440,
2006). However, the behavior and mode of failure of FRP are significantly different from those of steel.
Accordingly, research is needed to further investigate its behavior as a reinforcement material to facilitate a more
extensive, safer, and economical implementation in major infrastructure.
Previous experimental and analytical studies extended the fundamental principles of flexural theories of concrete
beams reinforced with steel bars to beams reinforced with FRP. Studies also recommended that beam flexure
design to be aimed at a compression failure, since it exhibits more warning before collapse, which somehow
affects the force path in the beam. However, due to its brittle mode of failure, a more conservative material
reduction factor should be used for design purposes when FRP is considered for reinforcement. Furthermore, the
longitudinal reinforcement contributes significantly to the nominal shear capacity, V
n
of S-RC beams. Such
contribution depends mostly on the axial rigidity and tensile strength of the reinforcing material. Due to the
lower axial rigidity of FRP compared to that of steel, it is expected that the amount of shear resisted by concrete,
V
cf
, would be lower and needs to be properly evaluated.
Existing shear design procedures and standards assume that the nominal shear capacity of conventional RC
beams with shear reinforcement (stirrups), V
n
, is simply the superposition of the concrete contribution to shear
resistance, V
c
, and that of the shear reinforcement, V
s
:
s c n
V V V + =
(1)
V
c
is deemed as the shear resistance of a similar reinforced concrete beam but without shear reinforcement, and
is usually calculated using a semi-empirical equation. V
s
is evaluated based on the parallel truss model with 45
o

constant inclination diagonal shear cracks which is conservative for the case of conventional steel reinforcement.
Despite the numerous research programs that investigated shear in RC members, accurately predicting the shear
capacity of conventionally reinforced concrete beams by itself remains a complex challenge that has not been

369
fully resolved (ACI 445, 1998). For instance, the axial stiffness of FRP is lower than that of steel, thus a larger
tensile strain in FRP bars is expected when it is used as longitudinal reinforcement, leading to a reduction in the
depth of the neutral axis, wider shear cracks, and a reduction in the overall contribution to shear resistance
mechanisms. Accordingly, when FRP is used as shear reinforcement, a reduction factor must be applied to its
capacity. Similarly, a reduction in the ultimate capacity of FRP stirrups by up to 79% of the guaranteed tensile
strength parallel to the fibers is applied.
The lack of detailed and reliable design standards and the fact that FRP materials do exhibit any plasticity
compared to steel reinforcement (brittle sudden failure) gives current guidelines a tendency to highly
underestimate the shear capacity of FRP-RC members. Consequently, excessive amount of FRP needed to resist
shear could be both costly and likely to create reinforcement congestion problems (Razaqpur et al., 2004).
Studies examined the ability of various shear design guidelines to calculate the contribution of concrete and/or
FRP stirrups to the nominal shear capacity of FRP-RC beams (Razaqpur et al., 2004). It was observed that
current shear design guidelines either underestimate or overestimate shear capacity of FRP-RC beams.
Accordingly, the objective of this paper is to evaluate the capacity of four current design specifications to predict
the shear strength of FRP-RC beams, namely, the provisions of the American Concrete Institute (ACI 440.1R-
06), the J apan Society of Civil Engineers (J SCE-97), the Canadian Standards Association (CSA S806-02) and
the Canadian Network of Centers of Excellence on Intelligent Sensing for Innovative Structures (ISIS Canada-
01).

METHODS EVALUATED IN THIS STUDY

Current shear design recommendations for FRP-RC beams in design codes and standards adopt a similar
approach to that used to calculate the shear capacity of S-RC beams, provided that the axial stiffness of FRP
longitudinal reinforcement is accounted for. Such an approach is based on the parallel truss model with 45
o

constant inclination diagonal shear cracks, and simply superimposes the shear capacity resisted by concrete, V
cf

to that resisted by FRP stirrups, V
f
. Previous research (Russo and Puleri, 1997) concluded that simply adding V
c

and V
s
to calculate V
n
leads to underestimating the shear strength of S-RC beams, but since it was always on the
safer side, it went overlooked. Furthermore, due to the brittle failure of FRP materials, current design guidelines
tend to further underestimate the shear capacity of FRP-RC members to achieve a more conservative design. The
various shear design guidelines and shear prediction equations used in this study are briefly discussed below.
Note that all strength reduction factors used in the following equations for design purposes are set to one for
comparison.

ACI Design Approach

ACI Committee 440 recommendations assume that equation (1) is also applicable for the calculation of nominal
shear capacity, V
n
of FRP-RC beams provided that shear cracks are adequately controlled. Accordingly,
accounting for the axial rigidity of the longitudinal reinforcement, ACI 440-06 evaluates the shear capacity of
FRP-RC beams without shear reinforcement using the following equation:
C b
f
V
w
c
cf
5
2
'
= (2)
where
'
c
f is the compressive strength of concrete cylinder (MPa); Kd C = ,
f fl f fl f fl
n n n K ρ ρ ρ − + =
2
) ( 2 ,
and
c fl f
E E n / = .
ACI 440-06 limits the stress level in the FRP shear reinforcement to control the width of shear cracks and avoid
failure at bends of stirrups. Thus, the shear capacity of FRP shear reinforcement is calculated as recommended
by ACI 440-06 using:
d b f V
w fv fv fv
ρ = where f
fv
=0.004 x E
fv
≤ f
fb
, (3)
where f
bend
, E
fv
, and ρ
fv
, are the strength at the bend portions, modulus of elasticity, and ratio of FRP stirrups,
respectively.

JSCE Design Approach

The J SCE-97 guidelines for the design and construction of FRP-RC structures are also a modified version of
specifications for steel-reinforced members. The J SCE-97 recommends the following equations for FRP-RC
flexural members to calculate the shear capacity carried by concrete, V
cf
:
d b
f
V
w
b
vcd n p d
cf
γ
β β β
= (4a)
370
5 . 1
100
3 ≤ =
s
fl fl
p
E
E ρ
β (4a)
5 . 1
1000
4
≤ =
d
d
β (4b)
72 . 0 2 . 0
3
'
≤ =
cd vcd
f f (MPa) (4d)
where
n
β =1.0 for members without axial force,
b
γ is a strength reduction factor,
'
cd
f =
'
c
f /
c
γ is the design
compressive strength, and
c
γ is the compressive strength material factor =1.3 for
'
c
f <50 MPa and 1.5
otherwise.
The contribution of the FRP shear reinforcement, V
f
, in non pre-stressed flexural members is given by:
b
s s fv fv fv
f
jd
s
E A
V
γ
α α ε
(
¸
(

¸
+
=
) cos (sin
(5a)
fv
bend
fv fv
fl fl mcd
fv
E
f
E
E f
≤ =
ρ
ρ
ε
'
0001 . 0 (5b)
mfb
fuv
b
b
bend
f
d
r
f
γ
|
|
.
|

\
|
+ = 3 . 0 05 . 0 (5c)
10 / 1
' '
300

|
.
|

\
|
=
h
f f
cd mcd
(5d)
where
fv
ε is the design value to limit the maximum strain in the FRP shear reinforcement, jd =d/1.15,
s
α is the
angle between the shear reinforcement and the member axis, h is the member’s height,
mfb
γ =1.3 (safety factor
for the bend portions of FRP bars), and r
b
and d
b
are the bend’s radius and bar’s diameter, respectively.

CSA Design Approach

The nominal shear capacity of FRP-RC flexural members not subjected to significant axial tension can be
calculated based on CSA S806-02 using the following equation:
d b f V V V V
w c cf f cf n
'
6 . 0 + ≤ + = (6)
In a similar way to that of ACI recommendations, the concrete contribution to the shear capacity, V
cf
and the
FRP stirrups contribution, V
f
in FRP-RC beams with minimum shear reinforcement is calculated using equations
(7) and (8 and 9), respectively.
d b
M
d V
E f V
w
f
f
fl fl c cf
3 / 1
'
035 . 0
|
|
.
|

\
|
= ρ for d ≤ 300 mm(7a)
d b f V d b f
w c cf w c
' '
2 . 0 1 . 0 ≤ ≤ (7b)
d b f d b f
d
V
w c w c cf
' '
08 . 0
1000
130
</ |
.
|

\
|
+
= for d >300 mm (8)
d b f d b f V
w c w fuv fv f
'
6 . 0 4 . 0 ≤ = ρ (9)
fuv
f is the ultimate strength of FRP bars parallel to the direction of fibers.

ISIS Canada Design Approach

ISIS Canada adopted similar principals used by CSA A23.3-94 (1994) for shear design of steel-reinforced
concrete in its design manual for FRP-RC structures with minimal modifications. The ISIS Canada method
calculates V
cf
and V
f
using the following equations:
s
fl
w c cf
E
E
d b f V
'
2 . 0 = for d ≤ 300 mm (10)
s
fl
w c
s
fl
w c cf
E
E
d b f
E
E
d b f
d
V
' '
1 . 0
1000
260
≥ |
.
|

\
|
+
= for d >300 mm (11)
371
s
fv c
w w fuv fv f
E
E f
d b d b f V
'
8 . 0 ≤ = χ ρ (χ is assumed = 0.4) (12)

Experimental Database

In the current study, shear strength results for 150 simply supported, rectangular concrete beams (50 of which are
without shear reinforcement) were collected from published literature. Beams without shear reinforcement had
FRP bars as longitudinal reinforcement. For beams with shear reinforcement, FRP stirrups were used, 72 of
which had FRP bars. Table 1 provides the range and average values of all parameters used in the database.

Table 1. Range of shear design parameters and V
n
for beams used in database
Parameters
Without shear reinf. (50 beams) With shear reinf. (100 beams)
Minimum Maximum Average COV Minimum Maximum Average COV
D (mm) 150.0 360.0 253.1 26 210.0 500.0 291.2 22
b
w
(mm) 150.0 1000.0 360.5 82 135.0 300.0 200.8 20
a/d 1.8 6.5 3.7 36 1.2 4.3 2.7 27
ρ
fl
E
fl
(GPa) 0.3 3.2 0.9 60 0.3 9.6 1.1 0.53
ρ
fv
f
fv
(MPa) -- -- -- -- 0.7 20.3 6.0 0.95
f

c
(MPa) 22.7 49.0 38.5 17 22.5 84.2 39.5 33
V
n
(kN) 31.5 190.0 87.8 54 56.2 375.5 164.1 47

DISCUSSION OF CODE EQUATIONS PERFORMANCE

The predicted shear capacities in comparison to those calculated using ACI 440.1R-06, J SCE-97, CSA S806-02,
and ISIS Canada-01 were then compared to the experimentally measured values. The performance of each
method was assessed based on both the ratio of measured to calculated shear strength (V
m
/V
p
), and the average
absolute error (AAE) calculated using equation (13):
100
1
x
V
V V
n
AAE
m
p m


= (13)
The average, standard deviation (STDV), and coefficient of variation (COV) for V
m
/V
p
, and the average absolute
error (AAE) of the investigated shear calculation methods are listed in Table 2.

Table 2. Performance of shear design methods in predicting the shear strength of FRP-RC beams.
Method
Without shear reinf. (50 beams) With shear reinf. (100 beams)
AAE (%)
V
measured
/ V
calculated

AAE (%)
V
measured
/ V
calculated

Average STDV COV (%) Average STDV COV (%)
ACI 440.1R-03 68.35 4.02 2.11 52 45.57 1.90 0.91 47
ACI 440.1R-06 29.74 1.51 0.69 45 31.33 1.22 0.48 39
CSA S806-02 33.13 1.68 0.66 39 17.49 1.13 0.27 24
J SCE-97 33.16 1.69 0.71 41 50.05 2.22 0.73 32
ISIS Canada-01 34.57 1.69 0.81 48 29.52 1.06 0.33 31

Beams without Shear Reinforcement
Figure 1 displays the shear strength of FRP-RC beams without shear reinforcement calculated using current
shear design provisions are plotted against the experimentally measured values. It is clear that shear design
equation provided by the latest version of ACI guidelines (ACI 440-06) is considerably improved from that of
ACI 440-03 and better estimates the shear capacity of FRP-reinforced concrete beams without stirrups with an
average (V
m
/V
cal
) of 1.51 (4.02 for ACI 440-03) (Figure 1). The 2003 version of ACI guidelines assume that the
shear strength of FRP-reinforced concrete beams increases linearly with the axial rigidity, ρ
fl
E
fl
of the
longitudinal reinforcement and decreases as the compressive strength of concrete increases. The first assumption
leads to overestimating the shear capacity of highly reinforced concrete beams, especially beams reinforced with
carbon fiber-reinforced polymer CFRP (high E
fl
), whereas the latter assumption yields very conservative shear
strength for FRP-RC beams without stirrups. The improved equations of ACI 440-06 properly assume that the
372
shear strength, V
cf
, is a function of compressive strength of concrete, f

c
, longitudinal reinforcement ratio, ρ
fl
, and
the modular ratio (E
fl
/E
s
).


0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 10 20 30 40 50 60
Beams
V
m
/
V
C
a
l
ACI 440-03 ACI 440-06

0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 10 20 30 40 50 60
Beams
V
m
/ V
C
a
l
CSA S806

0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 10 20 30 40 50 60
Beams
V
m
/
V
C
a
l
JSCE

0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0 10 20 30 40 50 60
Beams
V
m
/
V
c
a
l
ISIS

Figure 1. Measured versus calculated shear capacity of FRP-RC beams without shear reinforcement.

Similarly, shear design equations adopted by J SCE-97, CSA S806-02, and ISIS Canada-01 generally yielded
conservative results even without applying reduction factors (Figure 1) with an average V
m
/V
p
of 1.69.
Nonetheless, ISIS shear equations (eqs. 10 and 11) displayed a more scattered pattern than those of J SCE and
CSA with a COV of 48 % compared to 41 % and 39 % for J SCE and CSA, respectively.
Current guidelines and recommendations show an incoherent approach to modify the existing steel-based shear
design guidelines in order to adjust them to FRP-RC beams. For instance, the axial rigidity of the longitudinal
reinforcement (E
f
/E
s
) is square root for ISIS-01, and cubic root for J SCE-97and CSA S806, while mixed for
ACI-440. This is mainly caused by the limited verification with experimental results by each standard,
something that the current study aimed at exposing. Due to different availability of FRP materials in various
markets, researchers may be less exposed to other types of products and subsequently derive conclusion limited
to the range of tested parameters. Moreover, ISIS guidelines does not account for other important shear design
parameters on V
cf
, such as the longitudinal steel ratio, ρ
l
, and the shear span to depth ratio, a/d. This could be the
reason for the very conservative values of shear strength calculated using ISIS equations for the case of slightly
reinforced slender concrete beams (ρ
l
<0.5 % and a/d >2.5).

Beams with Shear Reinforcement
Figure 2 illustrates the calculated values of shear strength for concrete beams reinforced with FRP stirrups using
the design guidelines discussed earlier in comparison to the experimentally measured values. Despite the
conservative nature of shear equations proposed by ACI 440 in calculating V
cf
(Figure 1), these equations, in
some cases, offered unsafe predictions of the nominal shear strength V
n
=V
cf
+V
f
(Figure 2). This indicates that
ACI 440 tends to highly overestimate the V
f
component in the equation. A closer look at results reveals that the
majority of the beams for which ACI 440 overestimated the nominal shear strength had either a high shear
reinforcement ratio, ρ
fv
or were reinforced with stirrups having a high modulus of elasticity (E
f
). This indicates
yet again the need to revisit these two parameters (ρ
fv
, E
f
/E
s
) and their powers in the equations. Similar
observations are valid for shear equations proposed by ISIS (Figure 2), and to a lower extent for equations
proposed by CSA S806 (Figure 2).
Shear provisions of ACI 440, CSA S806, and ISIS adopted the rationale of equation (1) in calculating the
nominal shear capacity, V
n
, assuming that stirrups yield at failure and the capacity of FRP shear reinforcement
varies linearly with ρ
fv
E
fv
(in the case of ACI) or with ρ
fv
f
fuv
(for CSA and ISIS). These assumptions result in
overestimating the shear strength of concrete beams reinforced with FRP materials of high ultimate strength or
high modulus of elasticity. Moreover, CSA S806 (eq. 10) and ISIS (eq. 13) assume the dependency of shear
reinforcement capacity on the strength of FRP bars at the bend portion, which is estimated to be 40 % of the
ultimate tensile strength parallel to the direction of FRP fibers. However, an experimental study conducted by
Duranovic et al. (1997) to investigate the shear strength and mode of failure of concrete beams reinforced with
FRP stirrups lead to a conclusion that stresses in FRP stirrups at failure never exceeded 65 % of their bend’s
capacity.

373

0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
0 10 20 30 40 50 60 70 80 90 100 110
Beams
V
m
/

V
C
a
l
ACI 440-03 ACI 440-06

0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
0 10 20 30 40 50 60 70 80 90 100 110
Beams
V
m
/

V
C
a
l
CSA-S806

0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
0 10 20 30 40 50 60 70 80 90 100 110
Beams
V
m

/

V
C
a
l
JSCE

0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
0 10 20 30 40 50 60 70 80 90 100 110
Beams
V
m

/

V
C
a
l
ISIS

Figure 2. Measured versus calculated shear capacity for RC beams with FRP shear reinforcement.

CONCLUSIONS

This study verified the accuracy of several existing shear design methods to predict the nominal shear strength of
FRP-RC beams, and compared such predictions with actual experimental values. Several conclusions can be
drawn:
• Existing shear provisions considered in this study provided highly conservative results in estimating the
shear strength of FRP-RC beams without shear reinforcement.
• The ACI 440 equation for the FRP stirrups contribution to shear capacity, V
fv
is more for FRP materials
having low modulus of elasticity. However, it overestimates the capacity of FRP stirrups with high modulus
of elasticity.
• Shear standards of J SCE-97 provided good estimation of the shear capacity supplied by FRP shear
reinforcement for low shear capacity beams. However, such standards are highly conservative for beams
having high shear capacity.
• The CSA S806 method reasonably estimated shear strength of concrete beams having relatively low shear
reinforcement ratio, yet overestimated the shear capacity of highly reinforced beams with FRP stirrups of
high tensile strength.
• The effect of the axial rigidity of FRP longitudinal reinforcement on the shear capacity of FRP-RC beams is
better captured by a cubic root.

REFERENCES

American Concrete Institute (2006). “Guide for the design and construction of concrete reinforced with FRP
bars.” Technical Committee Document No. 440.1R-06, Farmington Hills, Mich.
Canadian Standards (CSA S806-02), “Design and construction of building components with fiber-reinforced
polymers,” Canadian Standards Association, Rexdale, Ontario, Canada, 2002, 116 p.
CSA A23.3-94, “Design of concrete structures,” Canadian Standards Association, Rexdale, Ontario, Canada,
1994, 220 p.
Duranavic, N., Pilakoutas, K., and Waldron, P. (1997) “Test on concrete beams reinforced with glass fiber
reinforced plastic bars,” Proceeding of FRPRCS-3, Sapporo, J apan, pp. 479-486.
ISIS Canada (2001) “Reinforcing concrete structures with fiber-reinforced polymers,” The Canadian Network of
Centres of Excellence, Design Manual No. 3, Zukewich, J ., editor, University of Manitoba, Winnipeg,
Manitoba, Canada, 133 p.
J apan Society of Civil Engineers (J SCE-97), “Recommendations for design and construction of concrete
structures using continuous fiber-reinforced materials,” Concrete Engineering Series 23, Machida, A.,
editor, 1997, 325 p.
J oint ACI-ASCE Committee 445 (1998) “Recent approaches to shear design of structural concrete,” J ournal of
Structural Engineering, V. 124, No. 12, pp. 1375-1417.
Razaqpur, A. G., Isgor, B. O., Greenaway, S., and Selley, A. (2004) “Concrete contribution to the shear
resistance of fiber-reinforced polymer reinforced concrete members,” ASCE J ournal of Composites for
Construction, Vol. 8, No. 5, pp. 452-460.
Russo, G., and Puleri, G., “Stirrups effectiveness in reinforced concrete beams under flexure and shear,” ACI
Structural J ournal, Vol. 94, No. 3, 1997, pp. 227-238.
374