Engineering Structures 127 (2016) 101–116

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Shear strength prediction for steel fiber reinforced concrete beams

without stirrups
Fasheng Zhang a,b, Yining Ding b,⇑, Jing Xu c, Yulin Zhang d, Weiqing Zhu e, Yunxing Shi a
a
China State Construction Engineering Corporation Technical Center, Beijing 101300, China
b
State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology, Dalian 116023, China
c
College of Water Resources and Civil Engineering, China Agricultural University, Beijing 100083, China
d
Center of Mathematics, University of Minho, Braga 4700-052, Portugal
e
School of Highway, Chang’an University, Xi’an 710064, China

a r t i c l e i n f o a b s t r a c t

Article history: This study proposes a theoretical approach based on modified compression field theory to predict the
Received 21 October 2014 shear strength of steel fiber-reinforced concrete (SFRC) beams without stirrups. The tensile contribution
Revised 3 August 2016 of the steel fibers is considered in the tensile stress–strain constitutive equations, which take into account
Accepted 8 August 2016
the distribution of the fibers. The proposed strength model is verified by comparing it with 139 shear fail-
Available online 2 September 2016
ure tests previously conducted on SFRC and reinforced concrete beams without stirrups. The influences of
concrete strength, fiber volume, shear span-to-depth ratio, and longitudinal steel ratio on predicted shear
Keywords:
strength are also discussed. Comparisons between the predicted and experimental results show that the
Steel fiber
Beam
proposed model can estimate shear strength accurately.
Modified compression field theory Ó 2016 Elsevier Ltd. All rights reserved.
Stress transfer
Shear strength

1. Introduction beams is independent is still prevalent today. On the other hand,

Previous studies have shown that adding steel fibers to concrete
increases tensile strength, post-cracking toughness, and ductility of
concrete [1,2]. One of the most important functions of steel fibers
in concrete is transferring stress across cracks, thus providing post-
cracking diagonal tension resistance to RC beams. Fibers have been
found to control crack development, prevent large crack widths,
increase ultimate shear strength and stiffness, and reduce deflec-
tions of concrete beams, thus also enhancing aggregate interlock.
Furthermore, using steel fibers transforms the failure mode from
brittle shear into ductile flexural [3–8].
Numerous experimental and theoretical investigations on shear
problems of various fibrous concrete beams without transverse
reinforcement have been conducted during the past three decades
[3–29]. These studies have indicated that adding fibers signifi-
cantly improves the shear capacity of beams. Based on the experi-
mental results, numerous researchers have proposed equations to
predict the ultimate shear strength of SFRC beams. A summary of
several models is presented in Table A1 in the Appendix. In the
existing models, the idea that the contribution of fibers and con-
crete in the shear strength of steel fiber reinforced concrete (SFRC)
⇑ Corresponding author.
E-mail address: ynding@hotmail.com (Y. Ding).
other studies hold the view that the contribution of fibers and con-
crete in shear strength is coupled.
Mansur et al. [6] performed twenty-four simply supported
beams under two symmetrical point loads and proposed an equa-
tion to calculate the ultimate shear at failure. In the model the
effect of fibers were included by considering a uniform stress block
of average stress along the tension crack, providing good predic-
tions of the ultimate strength. However, for beams with short shear
spans, the predicted shear capacities were highly conservative.
Sharma [4] performed seven tests on SFRC beams and proposed
an equation to calculate the ultimate shear at failure. The proposed
equation is a function of the concrete tensile strength and the shear
span-depth ratio. The equation was validated with 41 other tests
on SFRC beams. However, the equation ignored some important
parameters that contribute to the shear strength, such as fiber vol-
ume, aspect ratio, and tensile reinforcement ratio. Narayanan and
Darwish [9] proposed the cracking and ultimate shear equations
for SFRC deep and slender beams based on 33 tests on SFRC beams.
The proposed equations considered the bond stresses in the fiber
matrix. The combined effect of fiber volume and aspect ratio is also
considered as fiber factor. Kwak et al. [10] proposed two empirical
equations for ultimate shear strength of SFRC deep and slender
beams based on the Zsutty’s equation [11], considering the anchor-
ing action in addition to fiber factor. Swamy et al. [12] performed

http://dx.doi.org/10.1016/j.engstruct.2016.08.012
0141-0296/Ó 2016 Elsevier Ltd. All rights reserved.
102 F. Zhang et al. / Engineering Structures 127 (2016) 101–116
Nomenclature
lf fiber length
df equivalent diameter
qx, qz the reinforcement ratios in the longitudinal and trans-
verse direction
f0c cylinder compressive strength
da maximum aggregate size
fsx, fsz average stresses in the longitudinal and transverse rein-
forcement
f1 average tensile stress
f2 average compressive stress
h inclination of the diagonal compressive
m shear strength
ex average longitudinal strain
ez
average transverse strain
e2 principal compressive strain
e1 principal tensile strain
m0ci shear stress on the crack face for SFRC
mci shear stress on the crack face for RC
w average crack width
sh average crack spacing
an experiment program on SFRC beams and proposed a shear
equation considering the post-cracking tensile strength of fiber
concrete. Li et al. [13] proposed two equations to predict the shear
strength of deep and slender SFRC beams. In their model, both the
flexural and splitting strengths were important in predicting the
ultimate shear strength. Khuntia et al. [14] proposed a simple
equation for shear equations for SFRC beams also considering the
post-cracking tensile behavior of FRC beams. They added one addi-
tional term in the ACI building code equation to include the contri-
bution of fibers, which was derived from the equilibrium condition
of the forces in the diagonal crack, assuming 45° for the shear
cracking angle. The proposed model can reflect arch action, fiber
volume fraction, fiber orientation, and effective length of fiber at
crack. Dinh [15] proposed a simple model to estimate the shear
strength of SFRC beams without stirrup reinforcement. In their
approach, actual tension stress distribution is replaced by an
equivalent uniform tensile stress with the same tensile force resul-
tant. The contribution of fiber reinforcement to the shear strength
of the beam is directly linked to the material performance obtained
through a standard ASTM 1609 four-point bending test. Yakoub
[16] developed two equations to predict the contribution of steel
fiber to the shear strength of SFRC. These equations were used to
modify the CSA A23.3-04 general shear design method and the
equations of Bažant and Kim [17]. Yakoub introduced the contribu-
tion of steel fiber to the shear strength of SFRC by considering fiber
distribution. Slater et al. [18] proposed six different empirical
equations based on regression analysis. The equations were devel-
oped to predict the shear strength of SFRC beams based on span-
depth ratio, concrete compressive strength and fiber shape. How-
ever, it is not desirable in the design perspective to use six different
equations to predict the shear strength. Shahnewaz and Alam [19]
carried out a parametric study to evaluate the contribution of dif-
ferent parameters on the shear strength of SFRC beams. They
developed several analytical equations to predict the shear
strength of SFRC beams by genetic algorithm (GA). Compared with
previous test results, the proposed equations produced less scatter
with high accuracy. However, the proposed shear equations did not
attempt to model the physical reality of actual resisting shear
mechanisms in SFRC beams due to the complex nature.
In the previous studies mentioned above, most models agree
well with the test results from which it is derived, but the model
sx, sz indicators of the crack control characteristics of the x-
reinforcement and the z-reinforcement
rf stress transfer capability perpendicular to the crack
a the aspect ratio (a = lf/df. lf, df represent the length and
diameter of fibers, respectively)
sf the average bond strength
Vf the volume fraction of fibers
x2 the degree of planar orientation
w the average crack width
f the coefficient of friction between concrete and fiber
sheared over the crack edge
Pf average value of the load carried by the fibers that inter-
sect the crack

number of fibers per unit of area for 3D N A ¼ p2 2f
V
NA
df
a arbitrary angle between the loading direction and the fi-
ber
fci compressive stress on the crack surface
mf load component of Pf parallel to crack (mf ¼ rf =ð1 þ f Þ)
does not show good agreement with the other test results. The
main reason for that is most existing shear strength models for
treating shear still rely on empirical or semi-empirical formulas
because of the difficulty in understanding the complex shear trans-
fer mechanism. However, several recent approaches, such as the
models of plasticity theory [30] or modified compression field the-
ory [31,32] (MCFT), can be regarded as rational because they reveal
the physical mechanisms involved. Until now, however, the pro-
posed shear strength model based on MCFT just adds the contribu-
tion of steel fibers to the shear strength provided by the RC beams.
Such as the equations suggested by Yakoub [16], in which the
shear strength provided by concrete was given by the CSA A23.3-
04 (2004) [33] shear strength code. While the fiber contribution
was a separate part, which was determined by compressive
strength, the aspect ratio, volume fraction and fiber geometry,
without explaining how the bridging effect of fiber transfer the
shear stress for the SFRC beam.
The current study proposes a model based on MCFT to estimate
the ultimate shear strength of SFRC. Fiber action is introduced into
cracked concrete, and the cracked SFRC is regarded as a new mate-
rial. Compared with previous models, the proposed model consid-
ers the bridging effect of fiber based on equilibrium equations. The
study clarifies the physical reality of actual resisting shear mecha-
nisms in SFRC beams. The important parameters that affect the
shear strength were all considered though mechanical model.
The proposed model examines the feasibility of applying MCFT to
assess shear resistance of SFRC beams without stirrups rationally.
The model is then applied to existing test results available in liter-
ature to evaluate its accuracy. The results show that the proposed
strength model accurately predicts the test results of SFRC beams
without stirrups.
2. Review of previous experimental data
In the past few decades, a large number of experimental studies
have been conducted to investigate the shear behavior of SFRC
beams. A complete list of the tests used, along with their refer-
ences, is presented in Table A1. This list summarizes the details
of the test specimens and the results of previously performed
experiments to determine the effects of steel fibers on the shear
behavior of concrete beams, which have been used for the shear
 F. Zhang et al. / Engineering Structures 127 (2016) 101–116 103

Fig. 1. Comparison of local stresses at a crack with calculated average stresses of SFRC and stress state of a single fiber.

strength prediction models of SFRC in the present study. The tested
beams are typically designed with the following variables: steel
fiber volume fraction (Vf), shear span-to-depth ratio (a/dv), per-
centage of longitudinal reinforcement (q), and concrete strength
0 3. The proposed model based on MCFT
(f c ). In the current study, the test results for 139 beams are
recorded based on previously conducted shear failure tests on SFRC
beams and RC beams without stirrups. In some tests, failures are
caused by flexure or the combined effects of shear and flexure,
which are not considered in this study. That is because this study
only addresses shear failures in order to facilitate SFRC ultimate
shear strength prediction. The database contained a wide range
of a/d—from 1 to 6.0 for deep and slender beams. The database also
contained both normal- and high-strength concrete. The concrete
0
compressive strength (f c ) varied from 25 to 100 MPa. Similarly,
the longitudinal reinforcement ratio (q), aspect ratio of fibers
(lf/df), and fiber volume (Vf) varied from 1.03 to 4.58%, 45 to 100,
and 0.25 to 2.0%, respectively. Different types of steel fiber, such
as hooked, crimped, and straight, have also been used in the
experiments.
In SFRC, the principal benefits of the fibers are effective after
matrix cracking occurs because the fibers that cross the crack guar-
antee a certain level of stress transfer between both faces of the
crack. This process provides residual strength to the composite,
the magnitude of which depends on the properties of the fiber
and concrete matrix. The established theory of geometrical proba-
bility (integral geometry and stereology) offers a straightforward
approach to the spatial modeling of fibers that reinforce a leading
crack in concrete. Stroeven [34,35] introduced, discussed, and
applied these theoretical principles to SFRC under tension, thus
yielding general expressions for fiber contributions to stress trans-
fer across the leading crack for the ultimate and post-ultimate
104 F. Zhang et al. / Engineering Structures 127 (2016) 101–116

domains. Stroeven [35] suggested the following simplified concept k = 8 or 4, for a straight or hooked-end fiber, respectively; and
for the anisotropic contributions of steel fiber to post-peak f = the coefficient of friction between concrete and fiber sheared
strength: over the crack edge.



1 1 kw
rf ¼ asf V f ð1 þ f Þ 1 þ x2 1 ð1Þ Recently, compression field theory [36] (CFT) and MCFT have
3 2 lf
provided a new method to study the shear capacity of SFRC. The
where development of CFT is a significant step toward a more rational
rf = stress transfer capability perpendicular to the crack; theory for shear. Unlike traditional models, CFT uses strain condi-
a = the aspect ratio (a = lf/df. lf, df represent the length and tions in the web to determine the inclination of diagonal compres-
diameter of fibers, respectively); sive stresses. Extensive research has resulted in the development
sf = the average bond strength; of MCFT, which provides more accurate constitutive relationships
Vf = the volume fraction of fibers; than CFT, and includes an average concrete tension component of
x2 = the degree of planar orientation; concrete stress–strain relationships.
w = the average crack width;

Fig. 2. Flow chart for analysis procedure of the proposed model.

 F. Zhang et al. / Engineering Structures 127 (2016) 101–116 105

The equations of the MCFT can be used to predict the shear regarded as new materials with their own stress–strain character-
strength of a beam subjected to shear, moment, and axial loads. istics. Interestingly, SFRC is more suitable than plain concrete for
Simplified MCFT has been proposed to calculate the shear strength applying MCFT because of its flatter stress–strain relationship
of RC elements [37,38]. In MCFT, the cracked plain concrete is within the post-peak range under tension. The contribution of

18 18
16 Mean:1.25 16 Mean:1.30
SD:0.26 SD:0.30
14 COV:0.21 14 COV:0.23
12 12
νexp.(MPa)

νexp.(MPa)
10 10
8 8
6 6
4 4
2 2
0 0
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
νpred.(MPa) νpred..(MPa)

(a) Ashour et al. Model (b) Yakoub Model

18 18
16 Mean:1.06
16 Mean:1.17
SD:0.22 SD:0.23
14 14 COV:0.20
COV:0.21
12 12
νexp.(MPa)

νexp.(MPa)

10 10
8 8

6 6

4 4
2
2
0
0 0 2 4 6 8 10 12 14 16 18
0 2 4 6 8 10 12 14 16 18
νpred.(MPa) νpred.(MPa)

(c) Kwak et al. Model (d) Narayanan and Darwish Model

18 18
16 Mean:1.23 16 Mean:1.31
14 SD:0.29 SD:0.32
COV:0.23
14
COV:0.24
12 12
νexp.(MPa)

νexp.(MPa)

10 10
8 8
6 6
4 4
2 2
0 0
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
νpred.(MPa) νpred.(MPa)

(e) The Sharma Model (f) The Khuntia et al. Model

18 18
16 Mean:1.06 16 Mean:1.07
14 SD:0.23 SD:0.22
COV:0.22
14 COV:0.21
12 12
νexp.(MPa)

νexp.(MPa)

10 10
8 8
6 6
4 4
2 2
0 0
0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18
νpred.(MPa) νpred.(MPa)
(g) The Shahnewaz and Alam Model (h) The Proposed Model
Fig. 3. Shear strength predictions by different strength models.
106 F. Zhang et al. / Engineering Structures 127 (2016) 101–116

fibers to the shear capacity of SFRC can be considered by improving tions used in MCFT for SFRC are summarized in Table A2 in the
the tensile strength and residual strength of concrete based on Appendix.
MCFT. A model has been developed for the ability of a crack to resist
Failure of RC or SFRC elements may be governed by local stres- interfacial shear stress, fiber action, or aggregate interlock. The
ses occurring at a crack rather than by average stresses [36,39]. The relationships of the proposed theory presented in Table A2 can
load component that is transferred perpendicularly to the crack is be used to predict the shear strength of an SFRC beam. Assuming
caused by the resistance of the pullout friction and the shearing that the shear stress in the web is equal to the shear force divided
over the crack at the intersection of the steel fiber and the leading by the effective shear area bwd, and that qz = 0 for members with-
crack (Fig. 1(a)). A free-body diagram of a part of the SFRC element out stirrups, then the equilibrium Eqs. (6) and (10) can also be rear-
is shown in Fig. 1. This figure compares the calculated average ranged to provide the following expression for the shear stress m of
stresses (Plane 1 in Fig. 1(b)) with the actual local stresses occur- the section:
ring at a crack (Plane 2 in Fig. 1(c)). The critical crack direction is
m ¼ f 1 coth ð21Þ
assumed to be normal to the principal tensile strain direction.
Although the calculated average shear stress on Plane 1 is zero,
local shear stresses may be present on Plane 2.
m ¼ m0ci ð22Þ
The applied external stresses are fixed, and thus, the two sets of In this case, fiber distribution is assumed to be 3D (spatial portions),
stresses shown in Fig. 1(b) and (c) should be statically equivalent. that is,
Assuming a unit area for both Planes 1 and 2, the requirement in
which the two sets of stresses produce the same force in the z- 2 qffiffiffiffi 3
0
0:33 f c
direction is given by m ¼ 4 pffiffiffiffiffiffiffiffiffiffiffiffiffi ð1  V f Þ þ rf 5 cot h ð23Þ
1 þ 500e1
qz f sz cos h þ f 1 cos h ¼ qz f szcr cos h  f ci cos h þ mci sin h þ Pf sin aNA sin h
þ P f ðcos a þ f sin aÞN A cosh ð2Þ Substituting Eq. (1) into Eq. (23) yields the following expression
of m:
qffiffiffiffi

The load component of Pf parallel to crack mf = Pf sina NA and 1 kw
perpendicular to crack rf = Pf (cosa + fsina) NA can be obtained by m ¼ b f 0c þ asf V f ð1 þ f Þ 1  cot h ð24Þ
3 lf
analyzing the stress transfer between the fiber and the concrete
(Fig. 1(a)), and then Eq. (2) can be transformed into: where b ¼ pffiffiffiffiffiffiffiffiffi
0:33 cot h
ð1  V f Þ.
1þ 500e1

The behavior of a beam, particularly a short one, can be repre-

qz f sz cos h þ f 1 cos h ¼ qz f szcr cos h  f ci cos h þ mci sin h þ mf sin h sented partly by the arch action and partly by the beam action.
þ rf cos h ð3Þ Resolving the shear transfer mechanism of the beam in the beam
action and the arch action is rational [14,40]. For short beams (a/
dv < 2.5), the arch action of concrete compression struts should
The requirement that the two sets of stresses on Plane 1 pro-
be considered.
duce the same force in the x-direction is:  qffiffiffiffi
 
1 kw
m ¼ 2:5dv =a b f 0c þ asf V f ð1 þ f Þ 1  cot h a=dv
qx f sx sin h þ f 1 sin h ¼ qx f sxcr sin h  f ci sin h  mci cos h  mf sin h 3 lf
þ rf cos h ð4Þ < 2:5 ð25Þ

qffiffiffiffi

1 kw
where m ¼ b f 0c þ asf V f ð1 þ f Þ 1  cot h a=dv P 2:5 ð26Þ
3 lf
Pf = average value of the load carried by the fibers that intersect
the crack; For simplification, kw/lf can be regarded as zero. In this case,


NA = number of fibers per unit of area for 3D N A ¼ p2 d2f ;
V Eqs. (25) and (26) can be respectively rewritten as follows:
f qffiffiffiffi
1
a = arbitrary angle between the loading direction and the fiber; m ¼ 2:5dv =a ½b f 0c þ asf V f ð1 þ f Þ cot h a=dv < 2:5 ð27Þ
fci = compressive stress on the crack surface; 3
mci = shear stress on the crack face of the RC; and qffiffiffiffi
1
mf = load component of Pf parallel to crack (mf ¼ rf =ð1 þ f Þ). m ¼ b f 0c þ asf V f ð1 þ f Þ cot h a=dv P 2:5 ð28Þ
3
When the force on the x- and z-directions satisfies equilibrium When the two sets of stresses (the calculated stresses and the
conditions, then fci = 0 [36]. Given that the element will be used to actual local stresses occurring at a crack) in Fig. 1(b) and (c) are
model a section in the shear flexural region of a beam, the clamp- considered to be equivalent statically, b and h can be determined
ing stresses fz are assumed to be negligibly small. Then, Eq. (5) can
be obtained when qz = 0, as follows:
Table 1
f 1 ¼ m0ci tan h ð5Þ Comparison between different shear strength models.

Investigator Average value SD CV

where
v 0
v m r
ci ¼ ci þ f þ f cot h; and
The proposed model 1.07 0.22 0.21
Narayanan and Darwish [9] 1.17 0.23 0.20
m 0
ci = shear stress on the crack face of the SFRC.
Ashour et al. [20] 1.25 0.26 0.21
Kwak et al. [10] 1.06 0.22 0.21
The equilibrium, compatibility, and constitutive relationships Yakoub [16] 1.30 0.30 0.23
formulated in terms of average stresses and strains for SFRC can Sharma [4] 1.23 0.29 0.23
Khuntia et al. [14] 1.31 0.32 0.24
be obtained by introducing the mechanism of load transfer at the
Shahnewaz and Alam [19] 1.06 0.23 0.22
intersection of the steel fiber and the leading crack. Relevant equa-
 F. Zhang et al. / Engineering Structures 127 (2016) 101–116 107

by using the 2004 CSA general method [33,41]. Fig. 2 shows the
required flowchart for the analysis procedures to evaluate b, h
and the shear strength of SFRC members in detail by using the
equilibrium equations, compatibility equations, material constitu-
tive equations, and limit conditions discussed in the previous
0
sections. For given values of ex, concrete strength f c , and average
crack spacing sh, Fig. 2 shows how the 15 equations of MCFT for
SFRC in Table A2 can be solved through an iterative procedure.
The variable of longitudinal strain ex is defined as the strain at
mid-depth of the cross section at the flexural failure, i.e., ex = fy/2E

2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.
1.0 1.0

0.5 0.5

0.0 0.0
20 30 40 50 60 70 80 90 100 110 20 30 40 50 60 70 80 90 100 110
fc' (MPa) fc' (MPa)
(a) Ashour et al. Model (b) Yakoub Model
2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
20 30 40 50 60 70 80 90 100 110 20 30 40 50 60 70 80 90 100 110
fc' (MPa) fc' (MPa)
(c) Kwak et al. Model (d) Narayanan and Darwish Model
2.5 2.5

2.0 2.0
Vexp./Vpred.

1.5
Vexp./Vpred.

1.5

1.0 1.0

0.5 0.5

0.0 0.0
20 30 40 50 60 70 80 90 100 110 20 30 40 50 60 70 80 90 100 110
fc' (MPa) fc' (MPa)

(e) The Sharma Model (f) The Khuntia et al. Model

2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
20 30 40 50 60 70 80 90 100 110 20 30 40 50 60 70 80 90 100 110
fc' (MPa) fc' (MPa)
(g) The Shahnewaz and Alam Model (h) The Proposed Model
0
Fig. 4. Effect of f c on the predicted shear strength by different models.
108 F. Zhang et al. / Engineering Structures 127 (2016) 101–116

2.0 2.0

1.5 1.5

Vexp./Vpred.
Vexp./Vpred.
1.0 1.0

0.5 0.5

0.0 0.0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7

a/dv a/dv
(b) Yakoub Model
(a) Ashour et al. Model
2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.
1.0 1.0

0.5 0.5

0.0 0.0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
a/dv a/dv
(c) Kwak et al. Model (d) Narayanan and Darwish Model
2.5 2.5

2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
a/dv a/dv
(e) The Sharma Model (f) The Khuntia et al. Model
2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
a/dv a/dv
(g) The Shahnewaz and Alam Model (h) The Proposed Model
Fig. 5. Effect of a/dv on the predicted shear strength by different models.

[41]. For shear failure, the predicted shear capacity calculated by
ex = fy/2E is conservative. To reflect the real state of the longitudinal
strain used in MCFT for SFRC, the strain at mid-depth of the cross
section at the mid-shear span is selected for analysis, i.e., ex = fy/4E
(the strain at mid-depth of the cross section ex is assumed to have a
linear relationship with the bending moment at the same cross
section).
f typically changes with the sliding increment of fiber pullout
and the inclined angle of fiber with respect to cracking, and thus,
this variable is difficult to determine accurately. For convenience,
 F. Zhang et al. / Engineering Structures 127 (2016) 101–116 109

2.0 2.0

1.5 1.5

Vexp./Vpred.
Vexp./Vpred.
1.0 1.0

0.5 0.5

0.0 0.0
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 -0.5 0.0 0.5 1.0 1.5 2.0 2.5

Vf (%)
Vf (%)

(a) Ashour et al. Model (b) Yakoub Model

2.0 2.0

1.5 1.5

Vexp./Vpred.
Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 -0.5 0.0 0.5 1.0 1.5 2.0 2.5

Vf (%) Vf (%)

(c) Kwak et al. Model (d) Narayanan and Darwish Model

2.5 2.5

2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 -0.5 0.0 0.5 1.0 1.5 2.0 2.5
Vf (%) Vf (%)

(e) The Sharma Model (f) The Khuntia et al. Model

2 .0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 -0.5 0.0 0.5 1.0 1.5 2.0 2.5
Vf (%) Vf (%)
(g) The Shahnewaz and Alam Model (h) The Proposed Model
Fig. 6. Effect of Vf on the predicted shear strength by different models.

f can be regarded as zero in the following shear strength prediction, the second consideration of the influence of fiber shape on bond
qffiffiffiffi
although the contribution of fibers will be underestimated to a cer-
strength. In this study, the average bond strength s ¼ 0:68 f c ,
0

tain extent. Similar to previous studies [6,9,11], the current study

does not consider the shape of the steel fiber as a factor to avoid proposed by Khuntia et al. [14], is used.
110 F. Zhang et al. / Engineering Structures 127 (2016) 101–116

2.0 2.0

1.5 1.5

Vexp./Vpred.
Vexp./Vpred.
1.0 1.0

0.5 0.5

0.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
ρ (%) ρ (%)
(a) Ashour et al. Model (b) Yakoub Model
2.0
2.0

1.5
1.5
Vexp./Vpred.

Vexp./Vpred.
1.0
1.0

0.5
0.5

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
ρ (%) ρ (%)
(c) Kwak et al.Model (d) Narayanan and Darwish Model
2.5 2.5

2.0 2.0

1.5 1.5
Vexp./Vpred.
Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

ρ (%) ρ (%)

(e) The Sharma Model (f) The Khuntia et al. Model

2.0 2.0

1.5 1.5
Vexp./Vpred.

Vexp./Vpred.

1.0 1.0

0.5 0.5

0.0 0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
ρ (%) ρ (%)
(g) The Shahnewaz and Alam Model (h) The Proposed Model
Fig. 7. Effect of q on the predicted shear strength by different models.

4. Analytical and experimental results: comparison and the seven other models from literature are applied to 139 tests of
discussion SFRC and RC beams (see Table A3 in the Appendix). The seven other
models used to calculate the ultimate shear capacity for compar-
4.1. Comparisons between test results and predicted results ison are as follows:

To evaluate and compare the accuracy of the proposed model The Narayanan and Darwish model [9],
that predict the shear strength of SFRC, the proposed model and The modified ACI Building Code equation of Ashour et al. [20],
 F. Zhang et al. / Engineering Structures 127 (2016) 101–116
The Kwak et al. model [10],
The Yakoub modified CSA A23.3-04 model [16],
The Sharma model [4],
The Khuntia et al. model [14], and
The model developed by genetic algorithm of Shahnewaz and
Alam [19].
Fig. 3 and Table 1 show the comparison of the experimental
measured shear strength and predicted shear strength calculated
by different models. The analysis results shows that all the equa-
tions proposed by different investigators are over-predicting the
calculated shear strength. Khuntia et al. [14] is the most conserva-
tive of all the models considered, with a mean and standard devi-
ation (SD) for the average ratio of experimental shear strength to
predicted shear strength values of 1.31 and 0.32, respectively.
The equation proposed by Narayanan and Darwish [9] exhibits a
mean of 1.17, an SD of 0.23, and a coefficient of variation (CV) of
0.20. The equation of Ashour et al. [20] exhibits a mean of 1.25,
an SD of 0.26, and a CV of 0.21. The mean ratio (1.07) between
the experimental and predicted strength values of the proposed
model is lower than those of the models of Narayanan and Darwish
[9], Yakoub [16], and Ashour et al. [20], but greater than the Kwak
et al. [10] and Shahnewaz and Alam [19] models. Among all the
models, the proposed model produced better results, where the
average ratio of experimental shear strength to predicted shear
strength close to 1 with low SD (0.22) and COV (0.20).
4.2. Effects of certain parameters on the predicted shear strength
In this section, the proposed model is considered to investigate
further the effects of several variables on predicting the shear
strength values of SFRC beams. Although numerous important
variables affect the shear behavior of SFRC beams, only concrete
compressive strength, percentages of reinforcement in the longitu-
dinal direction, shear-span-to-depth ratio, and the volume fraction
of steel fiber are considered in the present study to assess their
effects on the predicted shear strength values.
Fig. 4(a)–(g) show the ratios of the experimental shear strength
to the predicted shear strength versus the concrete compressive
strength of the considered methods. These figures show that the
ratio fluctuates at 1 when the concrete strength is between
30 MPa and 50 MPa. The mean values of the ratios of the five mod-
els are all higher than 1. All the ratios at concrete strength between
60 MPa and 70 MPa are higher compared with those at concrete
strength between 30 MPa and 50 MPa. The mean values of the
ratios at concrete strength between 90 MPa and 100 MPa are
approximately 1, except for that of the Kwak model [10], which
is lower than 1. The Khuntia et al. [14], Ashour et al. [20] and Yak-
oub [16] methods yield conservative and scattered results for the
range of concrete strength values being investigated. The values
of the experimental shear strength to the predicted shear strength
ratios by using the Kwak et al. [10], Narayanan and Darwish [9] and
Shahnewaz and Alam [19] models display less scatter compared
with those of the other models. The ratios obtained using the pro-
posed model indicate consistent accuracy for the range of concrete
strength values being investigated, as shown in Fig. 4(h).
Fig. 5 shows the variations in the ratio of the experimental shear
strength to the predicted shear strength with a/dv values. The
results obtained using the Kwak et al. [10] and Shahnewaz and
Alam [19] models exhibit less scatter compared with those
obtained by the Yakoub [16], Khuntia et al. [14], Ashour et al.
[20], and Narayanan and Darwish [9] methods, thus resulting in
significant variations in all a/dv values being investigated. As can
be seen in Fig. 5, the ratio of the experimental shear strength to
111
the predicted shear strength for beams with 2.0 < a/dv < 4.0 is
higher compared with those with 1.0 6 a/dv 6 2.0, and a/dv > 4
for all models, except for that of Yakoub [16] and Khuntia et al.
[14]. The results obtained from the proposed model exhibit consis-
tency and accuracy for all a/dv values being investigated.
Fig. 6 shows the effects of fiber volume fractions on the pre-
dicted results. The Ashour et al. [20], Khuntia et al. [14] and Yakoub
[16] models exhibit a mean value of the ratio that is higher than 1.
The other three models, including the proposed model, exhibit a
mean value close to 1, which is significantly lower compared with
those of the Ashour et al. [20], Khuntia et al. [14] and Yakoub [16]
models. As expected, the results predicted using the proposed
model is in good agreement with all fractions, except for 2%.
Fig. 7 shows the variations in the experimental shear strength
to the predicted shear strength ratio as a function of the longitudi-
nal reinforcement ratio for all models. The ratios obtained using
the proposed model, the Shahnewaz and Alam [19] model and
the Kwak et al. [10] model exhibit less scatter compared with those
obtain using the other models when the longitudinal reinforce-
ment ratio is within the range of 1–3%. The results of the Yakoub
[16], Sharma [4] and Khuntia et al. [14]models exhibit more scatter
compared with those of the Kwak et al. [10], Shahnewaz and Alam
[19] and the proposed models.
5. Conclusions
This study presented a simple model based on MCFT to predict
the shear strength of SFRC beams. While most previous studies on
shear strength of SFRC beams proposed empirical equations based
on test results, the current study proposed the shear strength
model, in which the strain compatibility conditions, force equilib-
rium conditions, and constitutive equations of materials were
modified appropriately for SFRC. From the comparison of the
experimental and the predicted strength values using the proposed
shear strength model, as well as the other four existing models, the
following conclusions can be drawn.
1. Equations based on the concept of MCFT were proposed to pre-
dict shear capacity. In these equations, the cracked SFRC was
regarded as a new material with its own stress–strain charac-
teristics. Fiber contribution to the tensile strength of concrete
was introduced by a theoretical method.
2. The proposed model was used to predict the shear strength of
SFRC beams without shear reinforcement. The predicted shear
strength is in good agreement with the test results for SFRC
beams.
3. The results of the proposed model exhibited consistency and
accuracy within the range of the studied parameters, which
were also investigated by other researchers.
Acknowledgments
The research was supported by the National Natural Science
Foundation of China (Grant No.: 51578109), the National Natural
Science Foundation of China (Grant No.: 51121005), and the Por-
tuguese Funds through FCT-Fundação para a Ciência e a Tecnologia
within the Project UID/MAT/00013/2013. The authors wish to
express their gratitude for the financial support.
Appendix A
See Tables A1–A3.
112 F. Zhang et al. / Engineering Structures 127 (2016) 101–116

Table A1
Summary of proposed shear strength equations for fiber reinforced concrete (FRC) beams by different investigators.

Investigator Shear strength model, MPa

Sharma [4] d0:25
mfrc ¼ kft a
k ¼ 1 if f t is obtained by direct tension test;
k ¼ 2=3 if f t is obtained by indirect tension test;
k ¼ 4=9 if f t is obtained using modulus of rupture:
Mansur et al. [6]

qffiffiffiffiffi 
mfrc ¼ 0:41 sV f dlff þ 0:16 f 0c =bd þ 17:2 qMVd
h i
Narayanan and Darwish [9] mfrc ¼ e 0:24f cp þ 80q da þ 0:41sF
e = 1 for a/dv > 2.8; else e = 2.8 dv/a
Kwak et al. [10]
1=3
mfrc ¼ 3:7ef 2=3
spfc
q dav þ 0:8mb
qffiffiffiffiffi
Swamy et al. [12]
mfrc ¼ 0:37s f c V f d þ mc
0 lf
f

factor of 2 dv/a to multiply mc for a/dv < 2

 

Li et al. [13] mfrc ¼ 9:16 ðf t Þ2=3 ðqÞ1=3 da for a

6 2:5
d
 1=3

mfrc ¼ 1:25 þ 4:68 ðf t f sp Þ3=4 þ q da ðdÞ1=3 for a

d
> 2:5


qffiffiffiffiffi
Khuntia et al. [14]
mfrc ¼ 0:167 2:5 dav þ 0:25F f 0c for dav 6 2:5
qffiffiffiffiffi
mfrc ¼ ð0:167 þ 0:25FÞ f 0c for dav > 2:5
Dinh [15] mfrc ¼ 0:13As f y =bd þ ðrt Þarg ðd  cÞ cot a=dv
qffiffiffiffiffi

Yakoub [16]
mfrc ¼ 2:5b f 0c 1 þ 0:70V f DLff Rg dv
a for a
dv
6 2:5
qffiffiffiffiffi

0
mfrc ¼ b f c 1 þ 0:70V f
Lf
Df Rg for a
dv
> 2:5
Shahnewaz and Alam [19] mfrc ¼ 0:2 þ 0:034f 0c þ 19q0:087  5:8ða=dÞ1=2 þ 3:4V f0:4  800ðlf =df Þ1:6
0:05 1:4 2:12
12ðða=dÞV f Þ  197ðða=dÞðlf =df ÞÞ þ 105ðV f ðlf =df ÞÞ for a
d
6 2:5
0:85
mfrc ¼ 0:2 þ 0:072ðf 0c Þ þ 12:5q0:084  24ða=dÞ
0:07
þ 13:5V 0:07
f
2
þ 450ðlf =df Þ
3:9 0:84 2:69
0:0002ðða=dÞV f Þ  27:69ðða=dÞðlf =df ÞÞ þ 1181ðV f ðlf =df ÞÞ
0:9
21:89ðða=dÞV f Þðlf =df Þ for a
d
> 2:5
qffiffiffiffiffi
Ashour et al. [20]
mfrc ¼ ð0:7 f 0c þ 7FÞ dav þ 17:2qs dav

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 
Voo et al. [30]  2
mfrc ¼ 12 f c 1 þ hx  hx

mcr ¼ 12 f t h ah
2
þx2

mfrc can be determined when x is obtained by mfrc ¼ mcr (for non-prestressed FRC beams)

Table A2
Equations of modified compression field theory for fiber reinforced concrete.

Equilibrium: Compatibility conditions: Stress-strain relationships:

Average stresses: Average strains: Reinforcement:
f x ¼ qx f sx þ f 1  v cot h (6) tan2 h ¼ eexz þ
þ e2
e2
(11) f sx ¼ Es ex 6 f yx (16)
f z ¼ qz f sz þ f 1  v tan h (7) e1 ¼ ex þ ez þ e2 (12) f sz ¼ Es ez 6 f yz (17)
v ¼ ðf 1 þ f 2 Þ=ðtan h þ cot hÞ (8) cxz ¼ 2ðex þ e2 Þ cot h (13) Concrete:

0
2  (18)
fc 2e2 e2
f 2 ¼ 0:8þ170 e1 e0  e0
pffiffiffi0
0:33 f c
ffi ð1  V f Þ þ rf (19)
f 1 ¼ pffiffiffiffiffiffiffiffiffi
1þ 500e1

Stresses at cracks: Crack widths: Shear stress on crack:

pffiffiffi0ffi
f sxcr ¼ ðf x þ v cot h þ v 0ci cot hÞqx (9) w ¼ e1 sh (14) (20)

v 0ci 6 v ci þ mf þ rf coth ¼ 0:31þ 24wc þ mf þ rf cot h
0:18 f

f szcr ¼ ðf z þ v tanh  v 0ci tan hÞ=qz (10) sh ¼ 1= sinh (15) da þ16

sx þ sz
cosh

Where qx and qz = the reinforcement ratios in the longitudinal and transverse direction; f0c = cylinder compressive strength; da = maximum aggregate size; fsx, fsz = average
stresses in the longitudinal and transverse reinforcement; f1 = average tensile stress; f2 = average compressive stress; h = inclination of the diagonal compressive; m = shear
strength; ex = average longitudinal strain; ez = average transverse strain; e2 = principal compressive strain; e1 = principal tensile strain; m0ci = shear stress on the crack face for
SFRC; mci = shear stress on the crack face for RC; w = average crack width; sh = average crack spacing; sx, sz = indicators of the crack control characteristics of the x-rein-
forcement and the z-reinforcement.
Table A3
Summary of beam details and comparison of experimental and predicted shear strengths.

Investigator Beam no. Fiber Vf lf lf/df fc 0 bw d a/dv q (%) fy da vexp. vexp./vpred.

type (%) (mm) (MPa) (mm) (mm) MPa (mm) (MPa)
(1) (2) (3) (4) (5) (6) (7) (8)
Ashour et al. [20] B-1-0.5-A Hooked 0.5 60 75 99 125 204 1.0 2.84 460 10 9.09 0.78 0.87 0.63 1.26 2.07 1.73 1.08 0.95
B-2-0.5-A 0.5 60 75 99.1 125 204 2.0 2.84 460 10 4.82 1.05 1.19 0.80 1.33 0.89 1.16 0.95 1.01
B-4-0.5-A 0.5 60 75 95.4 125 204 4.0 2.84 460 10 2.27 0.80 0.96 0.75 0.80 1.02 1.43 1.02 0.86
B-6-0.5-A 0.5 60 75 95.83 125 204 6.0 2.84 460 10 1.95 0.74 0.95 0.72 0.69 1.51 1.08 1.12 0.74
B-1-1.0-A 1.0 60 75 95.3 125 204 1.0 2.84 460 10 12.74 1.04 0.96 0.87 1.49 0.96 0.99 1.03 1.11
B-2-1.0-A 1.0 60 75 95.3 125 204 2.0 2.84 460 10 6.06 1.17 1.18 0.95 1.42 1.51 2.32 1.13 1.05
B-4-1.0-A 1.0 60 75 97.53 125 204 4.0 2.84 460 10 3.17 0.90 1.10 0.89 0.92 1.22 1.63 1.08 0.97
B-1-1.5-A 1.5 60 75 96.4 125 204 1.0 2.84 460 10 13.95 1.07 0.86 0.92 1.38 1.13 1.58 0.96 1.02
B-2-1.5-A 1.5 60 75 96.6 125 204 2.0 2.84 460 10 7.21 1.22 1.15 1.04 1.43 0.71 1.05 1.19 1.06
B-4-1.5-A 1.5 60 75 97.1 125 204 4.0 2.84 460 10 3.51 0.84 1.04 0.86 0.87 0.78 1.08 1.07 0.90
B-2-1.0-M 1.0 60 75 94.5 125 201 2.0 4.58 470 10 6.73 1.10 1.14 0.92 1.58 1.21 1.49 1.08 1.17
B-4-1.0-M 1.0 60 75 93.8 125 201 4.0 4.58 470 10 3.88 1.02 1.16 0.98 1.14 1.05 1.51 0.96 1.20
B-6-1.0-M 1.0 60 75 95.0 125 201 6.0 4.58 470 10 2.93 0.83 1.00 0.82 0.86 1.05 1.54 1.00 0.90
Cucchiara et al. [22] B00 Hooked 0 – – 41.2 150 219 2 1.92 435 10 1.51 0.69 0.78 0.51 0.82 1.28 1.42 0.97 0.63

F. Zhang et al. / Engineering Structures 127 (2016) 101–116

A00 0 – – 41.2 150 219 2.8 1.92 435 10 1.23 0.92 0.89 0.65 0.83 1.03 1.30 0.98 0.72
B20 2 30 60 43.2 150 219 2 1.92 435 10 3.53 0.82 0.67 0.76 1.01 1.22 1.51 1.03 0.61
B10 1 30 60 40.9 150 219 2 1.92 435 10 3.50 1.09 0.98 0.93 1.34 1.01 1.56 1.01 0.87
A10 1 30 60 40.9 150 219 2.8 1.92 435 10 2.93 1.24 1.35 1.09 1.40 1.05 1.61 0.98 1.01
A20 2 30 60 43.2 150 219 2.8 1.92 435 10 3.14 0.91 1.05 0.88 1.13 1.23 1.75 1.06 0.76
Dupont and Vandewalle 13 Hooked 0 – – 40.3 200 260 1.5 1.81 560 14 4.04 1.24 1.46 0.96 1.66 1.35 1.37 0.89 1.28
[27] 25 0 – – 29.4 200 262 2.5 1.15 560 14 1.34 1.22 1.23 0.86 1.07 1.29 1.24 1.01 0.83
28 0 – – 29.4 200 260 2.5 1.81 560 14 1.63 1.23 1.30 0.90 1.31 1.40 1.31 1.06 1.02
Dupont and Vandewalle 1 Hooked 0 – – 43.5 200 260 3.5 3.55 560 14 1.74 1.07 1.08 0.96 1.15 1.24 1.18 1.01 1.29
[27] 22 0 – – 40.3 200 260 4 1.81 560 14 1.44 1.27 1.21 1.08 0.99 1.16 1.12 1.05 1.11
14 0.25 60 65 40.7 200 260 1.5 1.81 560 14 5.38 1.52 1.54 1.21 1.98 1.13 1.10 0.93 1.43
15 0.75 60 65 42.4 200 260 1.5 1.81 560 14 5.77 1.39 1.17 1.16 1.72 1.91 2.11 1.56 1.15
41 0.57 60 80 34.4 200 305 2.5 1.03 590 14 2.66 1.38 1.65 1.18 1.49 1.50 1.65 1.22 0.96
17 0.25 60 65 39.1 200 262 2.5 1.15 560 14 1.57 1.02 1.14 0.78 0.98 1.43 1.82 1.32 0.71
18 0.75 60 65 38.6 200 262 2.5 1.15 560 14 2.06 0.98 1.18 0.84 1.08 1.67 2.08 1.38 0.70
26 0.25 50 45 26.5 200 262 2.5 1.15 560 14 1.91 1.54 1.62 1.17 1.49 1.23 1.72 1.00 1.08
31 0.5 60 65 47.4 200 262 2.5 1.15 560 14 2.48 1.26 1.51 1.02 1.28 1.43 1.38 1.43 0.89
27 0.75 50 45 27.2 200 262 2.5 1.15 560 14 2.29 1.40 1.58 1.17 1.54 1.75 1.64 1.54 1.00
33 0.5 50 80 45.4 200 262 2.5 1.15 560 14 2.81 1.37 1.64 1.13 1.42 1.49 1.34 0.98 0.97
20 0.25 60 65 39.1 200 260 2.5 1.81 560 14 2.08 1.17 1.29 0.90 1.30 1.12 1.08 0.94 0.93
29 0.25 50 45 26.5 200 260 2.5 1.81 560 14 1.92 1.30 1.40 1.02 1.51 1.48 1.28 1.21 1.09
30 0.75 50 45 27.2 200 260 2.5 1.81 560 14 2.31 1.23 1.37 1.05 1.56 1.20 1.08 1.10 1.01
21 0.75 60 65 38.6 200 260 2.5 1.81 560 14 2.77 1.19 1.36 1.01 1.44 0.95 0.82 0.63 0.94
32 0.5 60 65 46.8 200 260 2.5 1.81 560 14 3.03 1.38 1.58 1.11 1.57 0.87 0.83 0.72 1.10
2 0.25 60 65 46.4 200 260 3.5 3.55 560 14 2.12 1.08 1.12 1.01 1.21 0.98 1.01 0.61 1.27
3 0.5 60 65 43.2 200 260 3.5 3.55 560 14 2.31 1.05 1.10 1.02 1.24 1.15 1.34 0.87 1.22
4 0.75 60 65 47.6 200 260 3.5 3.55 560 14 2.98 1.17 1.25 1.16 1.40 1.09 1.40 1.01 1.34
23 0.25 60 65 40.7 200 260 4 1.81 560 14 1.59 1.12 1.14 1.02 0.97 0.80 1.13 1.00 1.01
24 0.75 60 65 42.4 200 260 4 1.81 560 14 2.25 1.12 1.26 1.11 1.12 0.96 0.72 0.96 1.05
Kwak et al. [10] FHB1-2 Hooked 0 – – 62.6 125 212 2 1.48 442 10 3.02 1.27 1.48 0.89 1.33 1.00 0.74 0.69 1.02
FHB1-3 0 – – 62.6 125 212 3 1.48 442 10 2.53 1.68 1.77 1.28 1.39 0.98 1.18 0.92 1.56
FHB1-4 0 – – 62.6 125 212 4 1.48 442 10 1.98 1.41 1.53 1.25 1.09 0.84 1.22 1.05 1.22
FHB2-2 0.5 50 63 63.8 125 212 2 1.48 442 10 5.09 1.72 1.78 1.32 1.82 1.04 1.30 0.98 1.32
FHB2-2 0.5 50 63 30.8 125 212 2 1.48 442 10 4.04 1.79 1.67 1.46 2.08 1.05 1.31 0.99 1.38
FHB3-2 0.75 50 63 68.6 125 212 2 1.48 442 10 5.44 1.63 1.64 1.29 1.72 1.25 1.50 1.17 1.23
FHB2-3 0.5 50 63 63.8 125 212 3 1.48 442 10 3.09 1.49 1.71 1.27 1.38 1.11 1.61 1.11 1.43
FHB2-3 0.5 50 63 30.8 125 212 3 1.48 442 10 2.55 1.63 1.70 1.43 1.64 1.00 1.25 0.96 1.54

113
(continued on next page)
 114
Table A3 (continued)

Investigator Beam no. Fiber Vf lf lf/df fc 0 bw d a/dv q (%) fy da vexp. vexp./vpred.

type (%) (mm) (MPa) (mm) (mm) MPa (mm) (MPa)
(1) (2) (3) (4) (5) (6) (7) (8)
FHB3-3 0.75 50 63 68.6 125 212 3 1.48 442 10 3.40 1.40 1.67 1.25 1.34 1.36 1.72 1.31 1.37
FHB2-4 0.5 50 63 63.8 125 212 4 1.48 442 10 2.41 1.22 1.47 1.19 1.08 1.19 1.50 1.15 1.12
FHB2-4 0.5 50 63 30.8 125 212 4 1.48 442 10 2.00 1.36 1.47 1.32 1.28 1.22 1.53 1.17 1.21
FHB3-4 0.75 50 63 68.6 125 212 4 1.48 442 10 2.74 1.18 1.48 1.18 1.08 1.10 1.26 0.95 1.10
Para-Montesinos et al. [23] 6 Hooked 0 – – 42.8 152 381 3.4 2.67 410 10 1.08 0.75 0.74 0.65 0.72 1.03 1.33 1.01 0.66
5 0 – – 42.8 152 381 3.4 2.67 410 10 1.12 0.78 0.76 0.68 0.74 0.97 0.80 0.86 0.69
11 1 60 80 49.2 152 381 3.4 2.67 410 10 2.97 1.01 1.12 1.03 1.18 0.84 0.69 0.74 0.92
7 1.5 30 60 31 152 381 3.4 2.67 410 10 2.56 0.91 0.99 0.97 1.22 1.41 1.49 1.02 0.84
10 1.5 30 60 44.9 152 381 3.4 2.67 410 10 3.26 1.07 1.19 1.12 1.30 1.13 1.42 1.05 0.98
9 1.5 30 60 44.9 152 381 3.4 2.67 410 10 3.28 1.08 1.20 1.12 1.31 1.03 0.92 0.97 0.99
12 1 60 80 49.2 152 381 3.4 2.67 410 10 3.76 1.29 1.42 1.31 1.50 1.39 1.38 1.02 1.17
8 1.5 30 60 31 152 381 3.4 2.67 410 10 3.37 1.20 1.31 1.28 1.61 1.09 1.43 1.06 1.11
4 1 30 60 38.1 152 381 3.5 2.67 410 10 2.53 1.06 1.14 1.07 1.25 0.95 0.87 0.90 0.96
3 1 30 60 38.1 152 381 3.5 2.67 410 10 3.46 1.45 1.56 1.47 1.72 1.22 1.58 1.17 1.31
1 1 30 60 38.1 152 381 3.5 1.96 410 10 3.03 1.36 1.51 1.37 1.50 1.20 1.14 1.20 1.15
2 1 30 60 38.1 152 381 3.5 1.96 410 10 3.09 1.39 1.54 1.40 1.53 0.90 0.86 0.90 1.17

F. Zhang et al. / Engineering Structures 127 (2016) 101–116

Dinh et al. [29] B 18-0a Hooked 0 – – 42.8 152 381 3.43 2.7 410 14 1.1 0.74 0.74 0.59 0.77 1.30 1.65 1.27 0.72
B 18-0b 0 – – 42.8 152 381 3.43 2.7 410 14 1.1 0.77 0.80 0.65 0.86 1.36 1.58 1.22 0.74
B 18-1a 0.75 30 55 44.8 152 381 3.43 2 410 14 2.9 0.79 0.84 0.70 0.91 0.89 1.03 0.80 0.75
B 18-1b 0.75 30 55 44.8 152 381 3.43 2 410 14 2.8 0.81 0.79 0.70 0.77 1.43 1.27 1.43 0.71
B 18-2a 1 30 55 38.1 152 381 3.5 2 410 10 3 0.81 0.79 0.70 0.77 1.00 0.88 1.00 0.71
B 18-2b 1 30 55 38.1 152 381 3.5 2 410 10 3.1 0.90 0.98 0.87 0.92 1.38 1.14 1.28 0.75
B 18-2c 1 30 55 38.1 152 381 3.5 2.7 410 10 3.5 0.90 0.98 0.87 0.92 0.87 0.77 0.87 0.75
B 18-2d 1 30 55 38.1 152 381 3.5 2.7 410 10 2.6 0.91 0.99 0.90 0.99 1.30 0.98 1.10 0.76
B 18-3a 1.5 30 55 31 152 381 3.43 2.7 410 10 2.6 0.91 0.99 0.90 0.99 1.45 1.96 1.41 0.76
B 18-3b 1.5 30 55 31 152 381 3.43 2.7 410 10 3.4 0.84 0.90 0.85 0.99 1.43 1.74 1.14 0.76
B 18-3c 1.5 30 55 44.9 152 381 3.43 2.7 410 10 3.3 0.84 0.90 0.85 0.99 1.01 1.27 0.92 0.76
B 18-3d 1.5 30 55 44.9 152 381 3.43 2.7 410 10 3.3 0.84 0.92 0.89 1.12 1.12 1.43 0.93 0.78
B 18-5a 1 60 80 49.2 152 381 3.43 2.7 410 10 3 0.84 0.92 0.89 1.12 1.01 0.97 0.70 0.78
B 18-5b 1 60 80 49.2 152 381 3.43 2.7 410 10 3.8 0.85 0.93 0.87 1.01 0.71 1.06 0.69 0.77
B 18-7a 0.75 30 79 43.3 152 381 3.43 2 410 10 3.3 0.85 0.93 0.87 1.01 1.64 1.52 1.46 0.77
B 18-7b 0.75 30 79 43.3 152 381 3.43 2 410 10 3.3 0.85 0.94 0.86 0.99 1.68 0.99 1.50 0.77
B 27-1a 0.75 30 55 50.8 203 610 3.5 2 410 10 2.9 0.85 0.94 0.86 0.99 1.74 1.02 1.55 0.77
B 27-1b 0.75 30 55 50.8 203 610 3.5 2 410 10 2.7 0.90 1.00 0.90 0.97 1.77 1.04 1.58 0.76
B 27-2a 0.75 60 80 28.7 203 610 3.5 2 410 10 2.8 0.90 1.00 0.90 0.97 1.25 1.32 1.25 0.76
B 27-2b 0.75 60 80 28.7 203 610 3.5 2 410 10 2.8 0.90 0.99 0.87 0.90 1.27 1.25 1.20 0.75
B 27-3a 0.75 30 55 42.3 203 610 3.5 1.6 410 10 2.7 0.90 0.99 0.87 0.90 1.37 1.11 1.31 0.75
B 27-3b 0.75 30 55 42.3 203 610 3.5 1.6 410 10 2.8 0.90 0.98 0.92 1.07 1.32 1.49 1.17 0.77
B 27-4a 0.75 60 80 29.6 203 610 3.5 1.6 410 10 2.1 0.90 0.98 0.92 1.07 1.46 1.50 1.38 0.77
B 27-4b 0.75 60 80 29.6 203 610 3.5 1.6 410 10 1.8 0.95 1.05 0.93 0.93 0.93 0.80 0.93 0.75
Dinh et al. [29] B 27-5 Hooked 1.5 30 55 44.4 203 610 3.5 2.1 410 10 3.5 0.89 1.02 0.92 1.02 1.26 1.22 1.12 0.77
B 27-6 1.5 60 80 42.8 203 610 3.5 2.1 410 10 3.4 0.89 1.05 0.96 1.09 1.06 0.96 0.99 0.78
B 27-7 0 – – 37 203 610 3.5 1.6 410 10 1.3 0.99 0.93 0.84 0.77 1.05 1.25 1.05 0.72
Swamy et al. [12] 1TL-1 0 – – 35.4 55 265 2 4.31 460 10 3.37 0.99 1.40 0.94 1.97 1.67 1.97 1.52 1.52
2TL-1 0 – – 36.5 55 265 2 2.76 460 10 2.49 0.98 1.19 0.80 1.43 1.22 1.43 1.11 1.11
3TL-1 0 – – 37.4 55 265 2 1.55 460 10 2.15 1.14 1.23 0.82 1.22 0.99 1.22 0.94 0.94
1TL-2 0 – – 33.6 55 265 3.4 4.31 460 10 1.30 0.77 0.82 0.76 0.97 1.32 0.97 1.09 1.09
2TL-2 0 – – 33.4 55 265 3.4 2.76 460 10 1.22 0.93 0.89 0.83 0.92 1.24 0.92 1.03 1.03
3TL-2 0 – – 32.8 55 265 3.4 1.55 460 10 1.02 1.00 0.91 0.85 0.78 1.18 0.78 0.88 0.88
1TL-3 0 – – 34.1 55 265 4.9 4.31 460 10 1.19 0.86 0.84 0.78 0.88 1.33 0.88 0.99 0.99
2TL-3 0 – – 36.1 55 265 4.9 2.76 460 10 1.06 0.91 0.85 0.78 0.76 0.86 0.76 0.86 0.86
3TL-3 0 – – 33.8 55 265 4.9 1.55 460 10 1.03 1.11 1.03 0.95 0.77 0.87 0.77 0.87 0.87
Table A3 (continued)

Investigator Beam no. Fiber Vf lf lf/df fc 0 bw d a/dv q (%) fy da vexp. vexp./vpred.

type (%) (mm) (MPa) (mm) (mm) MPa (mm) (MPa)
(1) (2) (3) (4) (5) (6) (7) (8)
1TLF-1 1 50 100 35.6 55 265 2 4.31 460 10 5.48 1.07 0.96 1.11 1.88 1.19 1.88 1.10 1.10
2TLF-1 1 50 100 37.8 55 265 2 2.76 460 10 4.91 1.14 0.97 1.08 1.63 1.61 1.63 0.97 0.97
3TLF-1 1 50 100 35.7 55 265 2 1.55 460 10 4.63 1.29 1.08 1.18 1.59 1.50 1.59 0.93 0.93
1TLF-2 1 50 100 40.9 55 265 3.4 4.31 460 10 4.03 1.14 1.21 1.24 1.61 1.51 1.61 1.35 1.35
2TLF-2 1 50 100 33.1 55 265 3.4 2.76 460 10 3.11 1.02 1.12 1.10 1.38 1.13 1.38 1.09 1.09
3TLF-2 1 50 100 34.5 55 265 3.4 1.55 460 10 2.83 1.02 1.23 1.09 1.23 1.19 1.23 0.99 0.99
1TLF-3 1 50 100 36 55 265 4.9 4.31 460 10 2.90 0.92 1.01 0.99 1.24 1.21 1.24 1.00 1.00
2TLF-3 1 50 100 35.9 55 265 4.9 2.76 460 10 2.92 1.01 1.18 1.08 1.25 1.09 1.25 1.01 1.01
3TLF-3 1 50 100 32.5 55 265 4.9 1.55 460 10 2.01 0.76 1.00 0.83 0.90 0.77 0.90 0.71 0.71
Rosenbusch and Teutsch 2.2/1 Hooked 0 – – 40.8 200 260 1.5 1.81 500 14 4.03 1.24 1.45 0.96 1.65 1.19 0.73 0.89 1.27
[28] 2.4/1 0 – – 40.1 200 260 2.5 1.81 500 14 2.30 1.52 1.65 1.09 1.58 1.19 0.73 0.89 1.22
2.3/1 0 – – 40.1 200 260 2.5 1.15 500 14 1.50 1.18 1.25 0.82 1.03 1.41 1.46 1.23 0.80
T15⁄100-plain- 0 – – 32.1 200 460 3.4 2.8 500 14 1.65 1.26 1.22 1.14 1.27 1.36 1.41 1.19 1.43
1
20⁄30-plain-1 0 – – 32.1 200 260 3.5 2.83 500 10 1.15 0.89 0.85 0.80 0.88 1.25 1.53 1.20 1.00
1.2/1 0 – – 44 200 260 3.5 3.56 500 10 1.74 1.06 1.08 0.95 1.14 1.29 1.58 1.24 1.28

F. Zhang et al. / Engineering Structures 127 (2016) 101–116

20⁄60-plain-1 0 – – 32.1 200 540 3.5 2.73 500 10 1.00 0.79 0.75 0.70 0.77 1.46 1.78 1.40 0.87
2.6/1 0 – – 40.8 200 260 4 1.81 500 10 1.44 1.26 1.20 1.07 0.98 1.08 1.32 1.04 1.10
2.2/2 0.25 60 67 41.2 200 260 1.5 1.81 500 10 5.38 1.51 1.53 1.20 1.96 1.18 1.29 0.89 1.41
2.2/3 0.76 60 67 40.3 200 260 1.5 1.81 500 10 5.76 1.39 1.15 1.18 1.74 1.30 1.68 1.17 1.14
2.4/2 0.25 60 67 40 200 260 2.5 1.81 500 10 2.07 1.15 1.28 0.88 1.27 1.08 1.36 1.04 0.92
2.4/3 0.76 60 67 38.7 200 260 2.5 1.81 500 10 2.77 1.17 1.34 1.00 1.43 1.08 1.36 1.04 0.93
2.3/2 0.25 60 67 40 200 260 2.5 1.15 500 10 1.57 1.00 1.12 0.77 0.97 0.97 1.19 0.94 0.70
2.3/3 0.76 60 67 38.7 200 260 2.5 1.15 500 10 2.06 0.97 1.16 0.83 1.06 1.23 1.51 1.20 0.69
T15⁄100-SFRC- 0.5 60 67 37.7 200 460 3.4 2.8 500 10 2.65 1.34 1.39 1.31 1.52 1.33 1.54 0.90 1.46
2
T23⁄50-SFRC-2 0.5 60 67 38.8 200 460 3.4 2.8 500 10 2.74 1.38 1.43 1.33 1.55 1.33 1.54 1.23 1.50
T15⁄75-SFRC-2 0.5 60 67 37.7 200 460 3.4 2.8 500 10 2.81 1.42 1.48 1.38 1.61 2.14 1.37 1.02 1.55
T15⁄50-SFRC-1 0.5 60 67 37.7 200 460 3.4 2.8 500 10 2.86 1.45 1.50 1.41 1.64 2.00 1.28 0.95 1.58
1.2/2 0.25 60 67 46.9 200 260 3.5 3.56 500 10 2.11 1.07 1.11 0.99 1.20 1.20 1.60 1.06 1.25
1.2/3 0.51 60 67 43.7 200 260 3.5 3.56 500 10 2.31 1.04 1.09 1.01 1.23 1.20 1.60 1.06 1.20
1.2/4 0.76 60 67 48.3 200 260 3.5 3.56 500 10 2.98 1.15 1.22 1.14 1.37 1.17 1.40 1.17 1.31
20⁄30-SFRC-1 0.5 60 67 37.7 200 260 3.5 2.83 500 10 2.13 1.09 1.13 1.05 1.22 1.21 1.45 1.21 1.17
20⁄30-SFRC-2 0.5 60 67 38.8 200 260 3.5 2.83 500 10 2.53 1.28 1.33 1.24 1.43 1.56 1.18 0.86 1.38
Rosenbusch and Teutsch 20⁄60-SFRC-1 Hooked 0.25 60 67 37.7 200 540 3.5 2.73 500 10 1.42 0.86 0.87 0.80 0.90 1.33 1.01 0.73 0.93
[28] 20⁄60-SFRC-2 0.5 60 67 38.8 200 560 3.5 2.73 500 10 2.05 1.05 1.09 1.01 1.16 1.15 1.45 0.96 1.12
2.6/2 0.25 60 67 41.2 200 260 4 1.81 500 10 1.58 1.10 1.13 1.00 0.96 1.15 1.23 0.88 0.99
2.6/3 0.76 60 67 40.3 200 260 4 1.81 500 10 2.25 1.12 1.26 1.11 1.14 2.56 1.00 0.86 1.05

Note 1: (1) Narayanan and Darwish; (2) Ashour et al.; (3) Kwak et al.; (4) Yakoub; (5) Sharma; (6) Khuntia et al.; (7) Shahnewaz and Alam; (8) The proposed model.

115
116 F. Zhang et al. / Engineering Structures 127 (2016) 101–116

References [22] Cucchiara C, Mendola LL, Papia M. Effectiveness of stirrups and steel fibres as
shear reinforcement. Cement Concr Compos 2004;26(7):777–86.
[23] Para-Montesinos G, Wight JK, Dinh H, Libbrecht A, Padilla C. Shear strength of
[1] ACI Committee 544. State-of-the-art report on fiber reinforced concrete (ACI
fiber reinforced concrete beams without stirrups. Ann Arbor (MI): University
544.IR-96). Farmington Hills (Mich): American Concrete Institute; 1997. p.
of Michigan; 2006. p. 39. Report No. UMCEE 06-04.
1–60.
[24] Choi KK, Park HG, Wight JK. Shear strength of steel fiber-reinforced concrete
[2] Ding YN, Kusterle W. Comparative study of steel fibre-reinforced concrete and
beams without web reinforcement. ACI Struct J 2007;104(1):12–21.
steel mesh-reinforced concrete at early ages in panel tests. Cem Concr Res
[25] Aoude H, Belghiti M, Cook WD, Mitchell D. Response of steel fiber-reinforced
1999;29(11):1827–34.
concrete beams with and without stirrups. ACI Struct J 2012;109(3):359–67.
[3] Lim DH, Oh BH. Experimental and theoretical investigation on the shear of
[26] Ding YN, You ZG, Jalali S. Hybrid fiber influence on strength and toughness of
steel fibre reinforced concrete beams. Eng Struct 1999;21(10):937–44.
RC beams. Compos Struct 2010;92(9):2083–9.
[4] Sharma AK. Shear strength of steel fiber reinforced concrete beams. ACI J Proc
[27] Dupont D, Vandewalle L. Shear capacity of concrete beams containing
1986;83(4):624–8.
longitudinal reinforcement and steel fibers. In: Banthia N, Criswell M,
[5] Ding YN, You ZG, Jalali S. The composite effect of steel fibres and stirrups on
Tatnall P, Folliard K, editors. Innovations in Fiber Reinforced Concrete for
the shear behaviour of beams using self-consolidating concrete. Eng Struct
Value, vol. SP-216. Farmington Hills (MI): American Concrete Institute; 2003.
2011;33(1):107–17.
p. 79–94.
[6] Mansur MA, Ong KCG, Paramsivam P. Shear strength of fibrous concrete beams
[28] Rosenbusch J, Teutsch M. Trial beams in shear Brite/Euram Project 97-4163
without stirrups. J Struct Eng 1986;112(9):2066–79.
Final Rep., Sub Task 4.2. Braunschweig (Germany): Technical Univ. of
[7] Yazdanbakhsh A, Altoubat S, Rieder KA. Analytical study on shear strength of
Braunschweig; 2003. p. 105–17.
macro synthetic fiber reinforced concrete beams. Eng Struct 2015;100
[29] Dinh H, Parra-Montesinos G, Wight JK. Shear strength model for steel fiber
(1):622–32.
reinforced concrete beams without stirrup reinforcement. J Struct Eng
[8] Amin A, Foster SJ. Shear strength of steel fibre reinforced concrete beams with
2011;137(10):1039–51.
stirrups. Eng Struct 2016;111(3):323–32.
[30] Voo Y, Poon W, Foster S. Shear strength of steel fiber-reinforced ultrahigh-
[9] Narayanan R, Darwish IYS. Use of steel fibers as shear reinforcement. ACI Struct
performance concrete beams without stirrups. J Struct Eng 2010;136
J 1987;84(3):216–27.
(11):1393–400.
[10] Kwak Y, Eberhard MO, Kim WS, Kim J. Shear strength of steel fiber-reinforced
[31] Vecchio FJ, Collins MP. The modified compression field theory for reinforced
concrete beams without stirrups. ACI Struct J 2002;99(4):530–8.
concrete elements subjected to shear. ACI J 1986;83(2):219–31.
[11] Zsutty T. Shear strength prediction for separate categories of simple beam
[32] Minelli F, Vecchio FJ. Compression field modeling of fiber-reinforced concrete
tests. ACI J Proc 1971;68(2):138–43.
members under shear loading. ACI Struct J 2006;103(2):244–52.
[12] Swamy RN, Jones R, Chiam ATP. Influence of steel fibers on the shear resistance
[33] CSA Committee A23.3. Design of concrete structures (CSA A23.3-
of lightweight concrete I-beams. ACI Struct J 1993;90(1):103–14.
04). Mississauga: Canadian Standards Association; 2004.
[13] Li VC, Ward R, Hamza AM. Steel and synthetic fibers as shear reinforcement.
[34] Stroeven P, Hu J. Effectiveness near boundaries of fiber reinforcement in
ACI Mater J 1992;89(5):499–508.
concrete. Mater Struct 2006;39(10):1001–13.
[14] Khuntia M, Stojadinovic B, Goel S. Shear strength of normal and high-strength
[35] Stroeven P. Stereological principles of spatial modeling applied to steel fiber-
fiber-reinforced concrete beams without stirrups. ACI Struct J 1999;96
reinforced concrete in tension. ACI Mater J 2009;106(3):213–22.
(2):282–90.
[36] Collins MP. Towards a rational theory for RC members in shear. J Struct Div
[15] Dinh H. Shear behavior of steel fiber reinforced concrete beams without
1978;104(4):649–66.
stirrup reinforcement PhD thesis. University of Michigan; 2009.
[37] Collins MP, Mitchell D, Adebar P, Vecchio FJ. A general shear design method.
[16] Yakoub HE. Shear stress prediction: steel fiber-reinforced concrete beams
ACI Struct J 1996;93(1):36–45.
without stirrups. ACI Struct J 2011;108(3):304–14.
[38] Bentz EC, Vecchio FJ, Collins MP. Simplified modified compression field theory
[17] Bažant ZP, Kim JK. Size effect in shear failure of longitudinally reinforced
for calculating shear strength of reinforced concrete elements. ACI Struct J
beams. ACI J Proc 1984;81(5):456–68.
2006;103(4):614–24.
[18] Slater E, Moni M, Alam MS. Predicting the shear strength of steel fiber
[39] Collins MP, Mitchell D. Prestressed concrete structures. Englewood
reinforced concrete beams. Constr Build Mater 2012;26(1):423–36.
Cliffs: Prentice Hall; 1991.
[19] Shahnewaz M, Alam MS. Improved shear equations for steel fiber-reinforced
[40] Kim W, Jeong J. Decoupling of arch action in shear-critical reinforced concrete
concrete deep and slender beams. ACI Struct J 2014;111(4):851–60.
beams. ACI Struct J 2011;108(4):395–404.
[20] Ashour SA, Hasanain GS, Wafa FF. Shear behaviour of high-strength fiber-
[41] Bentz EC, Collins MP. Development of the 2004 CSA A23.3 shear provisions for
reinforced concrete beams. ACI Struct J 1992;89(2):176–84.
reinforced concrete. Can J Civ Eng 2006;33(5):521–34.
[21] Adebar P, Mindess S, Pierre DS, Olund B. Shear tests of fiber concrete beams
without stirrups. ACI Struct J 1997;94(1):68–76.