ACI MATERIALS JOURNAL TECHNICAL PAPER

Title No. 118-M47

JSCE-SF6 Limitations for Shear Tests and ASTM D5607

Shear Tests on Fiber-Reinforced Concrete
by Fulvio Tonon
Although several JSCE-SF6 tests have established the strength
improvements caused by fiber reinforcement, it is unknown whether
the measured strength improvements are caused by shear strength
improvements (that is, Mode II), or by tensile or flexural strength
improvement (Mode I). This paper provides a basic understanding of
the state of stress: 1) in a JSCE-SF6 test; and 2) in an ASTM D5607
direct shear test, which has never been used earlier to investigate
the shear strength of fiber-reinforced concrete (FRC). It was found
that the JSCE-SF6 standard does not test the shear strength of a
material, and it cannot be used to obtain shear characteristics of
FRC. On the other hand, the ASTM D5607 standard can be used
effectively to obtain shear characteristics of FRC because it creates
a stress field very close to pure shear. The addition of steel fibers
to concrete does not increase the peak shear strength of concrete.

Keywords: fiber-reinforced concrete (FRC); glass fibers; shear capacity;

shear failure; shear strength; shear stress; steel fibers; synthetic fibers.

INTRODUCTION
The shear strength of fiber-reinforced concrete (FRC)
may be assessed at a structural level (tests on beams and
corbels) or at a materials level (direct shear tests). The direct
shear tests are required to understand the basic shear-transfer
mechanisms and behavior of FRC, while the tests on beams
and corbels are necessary to understand the behavior of
FRC structures.1 As for FRC, ACI 544.4R-18,2 Section 4.8,
only addresses “the design aspects of FRC for shear in flex-
ural members where longitudinal bars are used in conjunc-
tion with fibers as shear reinforcement”—that is, concrete
structures.
Direct shear tests for FRC have so far fallen into three cate-
gories3,4: Z-type pushoff specimen,1 push-through test devel- Fig. 1—(a) Schematic for JSCE-SF65 shear test8; and
oped by the Japan Society of Civil Engineers (JSCE) JSCE- (b) schematic of shear test used by Mostafazadeh and
SF65 (Fig. 1), and FIP standard method.6 Qasim4 presents Abolmaali.11
finite element models (FEM) for the Z-type pushoff test and
In the literature, it has been assumed that the JSCE-SF6
the FIP standard method and compares these model results to
test produces a pure shear state of stress (for example, Refer-
experimental results. Mostafazadeh7 presents an ABAQUS
ences 3 and 8 to 10), and that the test measures the FRC
FEM model for a JSCE-SF6 test, where the author intro-
shear strength at a materials level under a Mode II failure.
duces an interface at the notches under the knives: “8-node
For example, Mirsayah and Banthia stated: “More recently,
three-dimensional cohesive element (COH3D) with a small
the Japan Society of Civil Engineering (JSCE) has proposed
amount of viscosity has been used to simulate mode II frac-
a standard method SF-6, which is an improvement over
ture of concrete equipped with synthetic fiber.” However:
the Z-type specimens in that during the test, the stress field
• The introduction of an interface element breaks
remains substantially that of pure shear, and hence a more
the FRC continuity—it introduces an artifact into a
reproducible shear response is obtained.”8
homogeneous material where the failure mode is prede-
The research reported in this paper is intended to:
termined, and therefore the model is forced to fail along
the cohesive elements. ACI Materials Journal, V. 118, No. 3, May 2021.
• The author does not discuss the state of stress in a JSCE- MS No. M-2020-246, doi: 10.14359/51732599, received June 20, 2020, and
reviewed under Institute publication policies. Copyright © 2021, American Concrete
SF6 test but only provides a comparison between exper- Institute. All rights reserved, including the making of copies unless permission is
obtained from the copyright proprietors. Pertinent discussion including author’s
imental results and FEM model results. closure, if any, will be published ten months from this journal’s date if the discussion
is received within four months of the paper’s print publication.

ACI Materials Journal/May 2021 91

• Provide a basic understanding of the state of stress in a
JSCE-SF6 test before failure, which so far has received
little consideration in the literature;
• Provide a basic understanding of the state of stress in an
ASTM D5607 direct shear test, which so far has never
been used to investigate the shear strength of FRC; and
• Carry out a set of tests aimed at quantifying the mate-
rial properties of FRC in shear according to the ASTM
D5607 test.

RESEARCH SIGNIFICANCE
Although several JSCE-SF6 tests have established the
strength improvements caused by fiber reinforcement, it is
unknown whether the measured strength improvements are
caused by shear strength improvements (that is, Mode II),
or by tensile or flexural strength improvement (Mode I).
A need therefore exists to investigate the stress field in the
JSCE-SF6 test (used so far in research on FRC), and in other
existing standardized tests to determine which test develops
a stress field as close as possible to “pure shear.” Once such
a test has been found, FRC should be tested to determine the
actual fiber improvement to shear strength.

LIMITATIONS OF JSCE G-553 TEST

As shown in Fig. 1(a), in the push-through test devel-
oped by JSCE-SF6,5 the shear load is applied to a beam by
a loading block with two sharp loading knife edges 150 mm
apart. The beam is 600 mm long, with a 150 x 150 mm cross
section, and it is supported on another rigid block over a
pair of knife edges that are 155 mm apart. Thus, a narrow,
2.5 mm wide region of the beam in between the loading and
the supporting knife edges is subjected to a concentrated
stress, and this is where the shear failure is expected to
occur. Mirsayah and Banthia8 found that if the specimen had
a 15 mm deep all-around notch to predefine the crack plane,
the shear failure occurred in the prescribed plane of fracture
and the test produced valid results with significantly reduced
variability. Without such a notch, the failure plane often
deviated from this narrow region, which is the prescribed
plane for failure, and the test produced invalid results. A
variation to the JSCE-SF6 test is presented by Mostafazadeh
and Abolmaali11; as depicted in Fig. 1(b), the concrete spec-
imen rests on two steel blocks, where it is clamped down
so that horizontal and vertical displacements are prevented, Fig. 2—(a) FEM model of test setup in Fig. 1(a); dimen-
and the two portions A and C cannot rotate during and after sions in mm; (b) stresses from linearly elastic JSCE Model 1
the test. along Section G-H in (a); and (c) details of stresses plotted
To investigate the state of stress in the beam before in (b) near tip H.
cracking, a two-dimensional (2-D) plane strain finite element
model was prepared for the setup in Fig. 1(a); it consisted of distributions along Section G-H for a 100 kN load P applied
0.5 x 0.5 mm four-node quadrilateral elements. The model, to the upper knife (in Reference 12, plain concrete failed at
shown in Fig. 2(a), was constrained in the vertical direction an average load of 250 kN):
at the lower knives; only half of the beam was modeled, with • The stress distributions over G-H are far from
horizontal constraints along the symmetry plane. In JSCE homogeneous.
Model 1, the material was assumed to be homogeneous and • All authors have so far calculated the shear strength
linearly elastic with a Poisson’s ratio equal to 0.3, and a of the material (or the shear stress in the specimen) by
Young’s modulus equal to 37 GPa (per Table 3.1 in Refer- dividing the force P by the area of the specimen; this is
ence 12, and a cylindrical compressive strength of 47 MPa not correct because the shear stress distribution is very
as in Reference 8). Figures 2(b) and (c) show the stress far from constant over the specimen cross section even
at 1/2.5 of the failure load.

92 ACI Materials Journal/May 2021

• All normal stresses are compressive near the top
(point G), and tensile near the bottom (point H) of the
specimen.
• Over the central part of the G-H section (20 ≤ y ≤
100 mm), the principal stresses are sensibly horizontal
(tensile) and vertical (compressive). However, a “pure
shear” state of stress would imply that principal stresses
are at 45 degrees to the shear plane.
• The principal stresses are never symmetric about the
zero value, which means that nowhere is concrete
subject to a “pure shear” stress field, as assumed by all Fig. 3—Difference between horizontal crack width propaga-
researchers until now (for example, References 3 and 8 tion near top and bottom of specimen.3
to 10), not even in the elastic range of the material.
At 1/2.5 of the failure load, the shear stress is maximum
at the top (point G), where it reaches 34.5 MPa. At the
bottom (point H), the shear stress is less than a third—that
is, 10.9 MPa. If the material failed in shear, shear failure
should start at the top, where the shear stress is maximum.
However, the experimental evidence is exactly the oppo-
site, as shown in Fig. 3; Soetens and Matthys3 report that
“a linear elastic behavior is observed until a first crack is
induced in both shear planes near the bottom of the spec-
imen. Under increasing loads, this crack propagates towards
the upper side of the specimen until a maximum or peak load
is reached.” Soetens and Matthys attribute this behavior to
“limited bending [that] occurred before reaching the ulti- Fig. 4—Yielded elements and deformed shape for JSCE
mate shear stress at the interface and hence, undesirable Model 2 (Von Mises criterion).
compressive stresses are acting in the upper region of the finally, if it is on the cap of the yield surface. The Von Mises
shear interface.” Although Fig. 2(b) shows that the shear yield surface has no tip and no cap, so only red and blue dots
stress is maximum at the top, the “shear” crack initiates at may be displayed. Figure 4 shows that yielding only occurs
the bottom because Fig. 2(b) shows that at the bottom all under the top knife and over the bottom knife. The rest of the
normal stresses are tensile, whereas at the top, all normal specimen translates downwards in a rigid body motion; there
stresses are compressive. is no compression/tension failure, and limited unloading/
To conclude, in the JSCE-SF6 test, failure is initiated by reloading occurs. There are no yielded elements along the
pure tension (Mode I) at the bottom of the specimen, and cross section G-H. These results show that a material that
then it propagates upwards, like in a zipper, because at the only fails in shear does not fail along the cross section G-H
bottom of the uncracked cross section, all normal stresses as has occurred with all plain concrete or FRC specimens
are tensile. In the literature, the addition of steel fibers reported in the literature. Therefore, the JSCE-SF6 test is not
always leads to an increase in the “shear strength” (Mode II) a shear test.
of the composite material measured by the JSCE-SF6 test. To investigate the behavior of concrete at the begin-
In actuality, the observed increase in the “shear strength” is ning of cracking, JSCE Model 3 had the same geometry as
nothing else but the increase in the tensile strength (Mode I) JSCE Model 1, but the material was assumed to follow the
of the composite. smeared crack approach as implemented in Midas FEA NX.
To confirm that the JSCE-SF6 test has nothing to do The Cornelissen et al. model13 was used for tension with a
with the shear strength of the material, JSCE Model 2 had tensile strength, ft = 2.2 MPa; a tensile fracture energy, Gf =
the same geometry as JSCE Model 1 but the material was 6 N/mm; and a crack bandwidth, h = 15 mm. The Thoren-
assumed to yield according to a Von Mises failure criterion feldt et al. model14 was used in compression with a compres-
with a yield stress σyield = 10 MPa. In the principal stress sive strength Fc = 47 MPa. Finally, the Al-Mahaidi shear
space, the Von Mises criterion is a right circular cylinder retention function15 was adopted for shear behavior. Figure
whose axis is the hydrostatic axis, and whose radius is σyield; 5 confirms the experimental observation (Soetens and
therefore, it is a pressure-insensitive J2-plasticity criterion Matthys3) that damage starts at and progresses from the
with infinite tensile and compressive strengths (J2 = second lower tip of the cross section G-H where cracks open up
invariant of the deviatoric stress), where yielding may occur (blue symbols “P”). The displacements depicted in Fig.  6
only in shear. Figure 4 shows the deformed mesh for the illustrate that up to the last converged iteration, the spec-
JSCE Model 2 when a 5 mm displacement is imposed to imen is locally punched at the knives, and the deformation
the top knife. Also shown are the yielded elements, where is similar to that of an elastic material except at the tips of
different colors indicate whether the stress point is on cross section G-H. When a crack fully develops (Fig. 6(c)
the yield lateral surface, or if it has returned to the elastic and (d)), the localized deformations at the knives become
domain, or if it is at the tension tip of the yield surface, or negligible with respect to the Mode I tearing of the material

ACI Materials Journal/May 2021 93

Fig. 5—Yielded elements for JSCE Model 3 (smeared crack model): (a) to (e) first five converged increments (10 to 50%
loading); and (f) detail at lower tip (point H in Fig. 2(a)) at last converged increment (Fig. 5(e)).
at the lower tip (point H), and the overall clockwise rota- JSCE Model 3—that is, tensile stress equal to ft = 2.2 MPa
tion of portion A (refer to Fig. 1) that further opens up the at 0.0006 strain. The results did not change with respect to
lower tip (point H). The highly elongated finite elements in JSCE Model 3.
Fig.  6(d) lie right in the middle of the 2.5 mm slot width,
and they indicate that a Mode I fracture will develop orthog- LIMITATIONS OF MODIFIED JSCE G-553 TEST
onal to the elongated edges; adjacent elements show a clear This section analyzes the modified JSCE-SF6 setup in
tearing (Mode I) pattern with the zipper mentioned previ- Fig. 1(b). The concrete-steel interfaces are modeled using a
ously, indicated herein by a dashed line. The maximum Mohr-Coulomb contact with a 30-degree friction angle. The
displacement in Fig. 6(d) is 0.2 mm, so the deformed shapes analysis comprises two stages: in Stage 1, the self-weight
in the figure are highly exaggerated. Because no groups of and the bolt prestress are applied; in Stage 2, the knife
twisted (sheared) elements are shown in Fig. 6(b) and (d), loading is applied.
shear deformation is absent. These simulations confirm that In JSCE Model 5, the FRC specimen geometry, discreti-
the JSCE-SF6 test is not a shear test. zation, and material parameters are the same as in the JSCE
Finally, to consider if the strain-hardening behavior of the Model 1 (that is, the SFRC is linearly elastic). Model results
SFRC may alter the aforementioned results, JSCE Model 4 show that the friction at the plate-FRC interfaces ensures
was the same as JSCE Model 3, but the first cracking was that no slippage occurs at these interfaces, and therefore the
modeled at a normal strain equal to 0.0002,16 and a tensile SFRC specimen is now constrained in the horizontal (x-axis)
stress equal to 2/3ft = 1.47 MPa. Beyond this point, the direction. The stress distributions (Fig. 7) are not radically
material continued linearly to the peak already used in the different from the stress distributions in Fig. 2; the important

94 ACI Materials Journal/May 2021

Fig. 6—Displacements for JSCE Model 3 (smeared crack model): (a) and (b) last converged increment; and (c) and (d) prog-
ress in nonconverged increment. Figures 6(b) and (c) are close-up views of Fig. 6(a) and (c), respectively, at lower tip, H,
of specimen.
differences are mainly related to the constraint operated by
the steel plates in the horizontal (x-axis) direction:
• The major principal stress is now compressive in the
upper half of the slot (y > 60 mm). This makes the crack
initiation in the upper half of the slot even less probable
than in the original configuration.
• The horizontal normal stress σxx is now compressive
over the entire slot, except at the lower tip H.
• At the lower tip H, all normal stresses are tensile (as in
the original configuration). Their magnitudes, however,
are approximately half their corresponding values in the
original configuration.
Therefore, all the limitations noted for the original config-
uration remain valid for the modified configuration as well,
with the added limitation that cracks cannot initiate in the
upper half of the slot, which is now completely compressed.
To investigate the behavior of FRC at the beginning of
cracking, JSCE Model 6 had the same geometry as JSCE
Model 5, but the specimen material was the same as in JSCE
Model 3 (smeared crack). Convergence was now achieved
only up to increment 3 (30% loading). Figure 8 shows that
except for the lack of cracked elements at the lower knife,
the cracked elements are the same as those in Fig. 5 (original
test configuration); as a consequence, the same consider-
ations apply as in the previous section.
Although this paper does not address the post-peak
behavior of FRC, it must be noted that the modified JSCE-
SF6 setup in Fig. 1(b) constraints the horizontal displace-
ments of the specimen also in post-peak condition. The
measured shear toughness is overestimated because it
Fig. 7—FEM model of test setup in Fig. 1(b): (a) stresses
accounts for the energy needed to shear off the asperities
from linearly elastic JSCE Model 5 along Section G-H in
on plane G-H when the specimen is forced to shear at zero
Fig. 1(a); and (b) details of stresses plotted in (a) near tip H.
displacement normal to the shear plane G-H.

ACI Materials Journal/May 2021 95

Fig. 8—Yielded elements for JSCE Model 6 (smeared crack model): (a) to (c) first three converged increments (10, 20, and 30%
loading); and (d) detail at lower tip at last converged increment (Fig. 8(c)).
FEM ANALYSIS OF ASTM D5607 DIRECT ring and shear box (C) was 30 mm thick. The finite elements
SHEAR TESTS were the same (type and dimensions) as in the JSCE-SF6
The direct shear test described in ASTM D5607 (Fig. 9(a)) model in Fig. 2(a). Likewise, the specimen material prop-
was initially devised to test rock fractures, but it is also used erties were the same as used in Section 3 for the JSCE-SF6
to test intact rock, where the concept is the same as the direct models. The encapsulating mortar, B, was assumed to have
shear test on soils per ASTM D3080. The ASTM D5607 the same properties as the specimen, A.
test consists of encapsulating a specimen into a shear ring For a linearly elastic specimen (ASTM Model 1), Fig. 9(c)
(which is then inserted into a shear box), applying a normal shows the stress distributions along the shear plane for a
force, and increasing the shear force (or shear displacement) 240 kN/m shear load P (similar to the JSCE Model 1, this
to failure, and then to post-peak condition. The test limita- is approximately 2.5 times smaller than the failure load of
tions for testing soil or rock are well-known: plain concrete), and zero normal load.
• Predefined shear plane: In the case of FRC, this is not • Compared to Fig. 2(b), the stress distributions are far
a problem. Rather, it is an advantage because it allows more homogeneous.
the shear strength to be determined along the plane • The stresses range is (−8, 6.5) MPa as compared to
chosen by the designer. The JSCE-SF6 test also has a (−150, 44) MPa for the JSCE-SF6 test.
predefined “shear” plane. • The shear stress is higher at the tip closer to the point of
• Rotation of principal stress axes during the test: This load application.
is caused by the application of a normal stress to the • At no point are all normal stresses compressive or
shear plane. Because the original JSCE-SF6 test is not tensile. In the JSCE-SF6 test, all normal stresses are
meant to investigate the effect of a normal stress on the compressive near the top (point G), and tensile near the
shear plane, a zero normal stress is applied in the tests bottom (point H) of the specimen.
described herein. • Over the central part of the shear plane, the horizontal
• Lack of drainage control and pore pressure measure- normal stress is compressive (up to 1.6 MPa), and the
ment. This is not a consideration in FRC. vertical stress is nearly negligible (0.25 MPa, tensile).
Because the upper part of the shear box is constrained • The principal stresses are nearly symmetric about the
against rotation around horizontal axes, the plane-strain FEM zero value, with the shear stress nearly matching their
model in Fig. 9(b) was used to determine the stress distri- absolute values, which means that concrete is subject to
bution along the shear plane. To match the laboratory test a nearly “pure shear” stress field.
results described next, the specimen (A) was 100 x 100 mm To confirm that the ASTM D5607 test captures the shear
and had a 5 x 5 mm notch along the shear plane; the encap- strength of the material, ASTM Model 2 had the same geom-
sulating mortar (B) was 10 mm thick, and the steel shear etry as ASTM Model 1, but the material was assumed to
yield according to a Von Mises failure criterion with a yield

96 ACI Materials Journal/May 2021

 and cracking start at the notches of the concrete specimen
in a nearly symmetric manner with respect to the vertical
and horizontal axes of symmetry of the specimen. Figure 11
shows the deformed shape of the shear box, shear ring, and
specimen for the last converged iteration in ASTM Model 3.
When compared to Fig. 6, the displacement field is fairly
homogenous; there are no elongated elements (Mode I), and
the entire specimen is not subject to any rotation.

ASTM D5607 DIRECT SHEAR TESTS OF SFRC

After running tests according to the JSCE-SF6, many
researchers concluded that the addition of steel fibers
increases the peak shear strength of plain concrete.4,7-11 For
example, Mirsayah and Banthia8 found that the addition
of steel fibers with volume fractions between 0.25 and 2%
increased the peak shear strength from 19.5 to 88% when
measured per JSCE-SF6. Soetens and Matthys3 found that
the peak shear strength is, in MPa, approximately equal to
10 times the volume fraction (in percentage). However, the
previous part of this paper has shown that the JSCE-SF6 test
does not measure the shear strength of the material (Mode II),
and the ASTM D5607 test is much closer to testing the
shear strength of the material. To determine if steel fibers
really increase the peak shear strength of concrete, ASTM
D5607 tests were carried out on steel FRC (SFRC) and plain
concrete.
The mixture proportion is given in Table 1 and followed
preconstruction trial batches of SFRC tunnel liner segments
for the Seattle Brightwater Tunnel Project prepared by
CSI-Hanson JV. Slump was in the 5 to 10 cm range. Steel
fibers were cold-drawn wire fibers with hooked ends in the
amount of 41.5 kg/m3—that is, 0.55% volume fraction.
Besides being a real application for a major tunnel project,
this volume fraction is approximately the average of that
used in many academic studies, and is typical of precast
applications not only in tunneling but also in culverts.7,17
ASTM C39 cylinders and ASTM C1609 beams were cast,
and tests were carried out at 28 days. The specimens for
shear testing were cut out of the ASTM C1609 beams so that
the shear plane was either at 90 degrees or at 0 degrees to the
beam axis. Because steel fibers were found to be sensibly
parallel to the beam axis, this also means that the shear plane
was either at 90 degrees or at 0 degrees to the fiber direction.
Table 2 provides the results for the cylindrical compres-
Fig. 9—Direst shear test per ASTM D5607: (a) schematic
sive strength per ASTM C39, and the shear strength per
test setup—direct shear box with encapsulated specimen; (b)
ASTM D5607. The coefficients of variation for the compres-
FEM model of test; and (c) stresses along shear plane for
sive strengths are well within typical values for laboratory
linearly elastic model (ASTM Model 1).
mixtures. By comparing plain concrete with SFRC, the
stress σy = 10 MPa. If a material only yields in shear and it following may be noted:
is subjected to the ASTM D5607 shear test, then Fig. 10(a) • The differences in cylindrical compressive strength
shows that such a material only fails along the shear plane are negligible. This is in line with data in the literature
predefined by the shear box. This confirms that the ASTM (Chapter 2, Table 2.12).16
D5607 test captures the shear strength of the tested material. • The differences in shear strength are in the 2 to 5%
To confirm the correct behavior of ASTM D5607 at the range—that is, the same order of magnitude as differ-
beginning of cracking, ASTM Model 3 had the same geometry ences in cylindrical compressive strength, and within
as ASTM Model 1, but the material was assumed to follow the limits of material variability. Therefore, they are
the smeared crack approach as implemented in Midas FEA negligible. According to Soetens and Matthys,3 a
NX and with the material properties as in Section 3 (JSCE 5  MPa—that is, 67% shear strength increase—should
Model 3). Figures 10(b) and (c) demonstrate that yielding have been obtained for the volume fraction used herein.

ACI Materials Journal/May 2021 97

Fig. 10—(a) Yielded elements and deformed shape for ASTM Model 2 (Von Mises criterion); (b) crack status for Model 3
(smeared crack model); and (c) yielded elements for Model 3 (smeared crack model).
This demonstrates that the alleged increase in peak shear finding is in line with Eq. 4.82 that gives the shear capacity,
strength caused by the addition of steel fibers is not caused VFRC, of a beam with conventional longitudinal reinforce-
by an actual increase in the strength of the material under ment but without shear reinforcement
Mode II but due to the fact that the JSCE-SF6 test is actually
a Mode I test, where fibers do increase the peak strength of  1

the material (Reference 2, Eq. 4.5(a) and 4.5(b), and Refer-  0.18   f ut − FRC   3 
VFRC = k s 100ρ 1 + 7.5   + 0 .15σ cp  ⋅ b ⋅ d
ence 16, Chapter 4, Table 4.2). At a structural level, this  cγ   f t  
 

98 ACI Materials Journal/May 2021

Table 1—Concrete mixture proportions adopted If, as in the tests run in this section, there is no longitudinal
from Seattle Brightwater Tunnel Project reinforcement, ρ = 0, and if the average normal stress, σcp, is
Mixture ingredient Quantity
zero, then VFRC = 0.
Cement 341 kg/m3
CONCLUSIONS
Fine aggregate 872 kg/m3 1. The JSCE-SF6 standard does not test the shear strength
Coarse aggregate 1080 kg/m3 of a material and it cannot be used to obtain shear character-
Water-reducing admixture 1.577 L/m3
istics of FRC.
2. The alleged increase in peak shear strength caused by
Water 119 kg/m3
the addition of steel fibers is not caused by an actual increase
w/c 0.35 in the strength of the material under Mode II but is due to the

Table 2—Test results for plain concrete and SFRC: ASTM C39 cylindrical strength and ASTM D5607
shear strength
Shear specimen dimensions Peak shear load and strength
Compressive strength ASTM Shear strength difference with
Mixture C39 (standard deviation), MPa Length, mm Width, mm Load, kN Strength, MPa respect to plain concrete, %
Plain 55.8 (1.08) 100.01 63.10 44.3 7.14 0
Dramix 90 degrees (1) 96.42 50.93 36.1 7.35 3
Dramix 90 degrees (2) 96.41 58.78 41.4 7.30 2
56.6 (0.67)
Dramix 0 degrees (1) 102.50 59.40 45.7 7.51 5
Dramix 0 degrees (2) 97.25 49.79 35.8 7.39 4

Fig. 11—Displacements for ASTM Model 3 (smeared crack model), last converged increment: (a) general view of model; and
(b) close-up view of sheared specimen.

ACI Materials Journal/May 2021 99

fact that the JSCE-SF6 test is actually a Mode I test, where Postgraduate Engineering Conference, IOP Conference Series: Mate-
rials Science and Engineering, Baghdad, Iraq, V. 745, 2020, doi:
fibers do increase the peak strength of the material. 10.1088/1757-899X/745/1/012176
3. The ASTM D5607 standard can be used effectively 5. Japan Society of Civil Engineers, “Method of Test for Shear Strength
to obtain shear characteristics of FRC because it creates a of Steel Fiber Reinforced Concrete (SFRC) (JSCE-SF6),” JSCE, Tokyo,
Japan, 2005.
stress field very close to pure shear. 6. Fédération Internationale de la Précontrainte, “Shear at the Interface of
4. When tested according to the ASTM D5607 standard, Precast and In-Situ Concrete,” FIP, Lausanne, Switzerland, 1978.
the addition of steel fibers to concrete does not increase the 7. Mostafazadeh, M., “Shear Behavior of Synthetic Fiber Reinforced
Concrete Box Culverts,” PhD dissertation, Department of Civil Engi-
peak shear strength of the material. neering, The University of Texas at Arlington, Arlington, TX, 2017.
8. Mirsayah, A. A., and Banthia, N., “Shear Strength of Steel Fiber-Re-
AUTHOR BIO inforced Concrete,” ACI Materials Journal, V. 99, No. 5, Sept.-Oct. 2002,
Fulvio Tonon is an Associate Professor at the University of Udine, Udine, pp. 473-479.
Italy. He received his BS and MS in civil engineering from the University 9. Malatesta, S. C., and Contreras, M. C., “Shear Behavior of Steel Fiber
of Padua, Padua, Italy, in 1995, and his PhD in civil engineering from the Reinforced Concretes,” Revista de Ingeniería de Construcción, V. 24,
University of Colorado Boulder, Boulder, CO, in 2000. His research inter- No. 1, 2009, pp. 74-94.
ests include rock mechanics and engineering, underground excavations, 10. Carmona, S., and Maripangui, F., “Comportamiento del Hormigón
and fiber-reinforced shotcrete and concrete. Reforzado con Fibras Sometido a Esfuerzo de Corte,” Anales de las XVI
Jornadas Chilenas del Hormigón, Talca, Chile, 2007, 10 pp.
11. Mostafazadeh, M., and Abolmaali, A., “Shear Behavior of Synthetic
NOTATION Fiber Reinforced Concrete,” Advances in Civil Engineering Materials, V. 5,
b = section width, mm
No. 1, 2016, pp. 371-386. doi: 10.1520/ACEM20160005
d = depth of reinforcement, mm
12. European Committee for Standardization, “Eurocode 2: Design of
fc = compressive strength for plain concrete, MPa
Concrete Structures—Part 1-1: General Rules and Rules for Buildings (EN
ft = tensile strength for plain concrete, MPa
1992-1-1),” CEN, Brussels, Belgium, 2005.
fut-FRC = FRC ultimate tensile residual strength, MPa
13. Cornelissen, H. A. W.; Hordijk, D. A.; and Reinhardt, H. W., “Exper-
ks = factor for size effect in shear calculations
imental Determination of Crack Softening Characteristics of Normalweight
γc = partial safety factor for plain concrete
and Lightweight Concrete,” HERON, V. 31, No. 2, 1986, pp. 45-56.
σmax = major principal stress
14. Thorenfeldt, E.; Tomaszewicz, A.; and Jensen, J. J., “Mechanical
σmin = minor principal stress
Properties of High Strength Concrete and Application to Design,” Proceed-
σxx = normal stress on plane normal to x-axis
ings, Symposium: Utilization of High Strength Concrete, Stavanger,
σxy = shear stress on x-y plane
Norway, 1987, pp. 149-159.
σyy = normal stress on plane normal to y-axis
15. Al-Mahaidi, R. S. H., “Nonlinear Finite Element Analysis of Rein-
σzz = normal stress on plane normal to z-axis
forced Concrete Deep Members,” PhD thesis, Technical Report 79-1,
Department of Structural Engineering, Cornell University, Ithaca, NY,
REFERENCES 1979.
1. Valle, M., and Buyukozturk, O., “Behavior of Fiber Reinforced High- 16. Naaman, A. E., and Reinhardt, H. W., “Proposed Classification of
Strength Concrete under Direct Shear,” ACI Materials Journal, V. 90, HPFRC Composites Based on Their Tensile Response,” Materials and
No. 2, Mar.-Apr. 1993, pp. 122-133. Structures, V. 39, No. 5, 2006, pp. 547-555. doi: 10.1617/s11527-006-9103-2
2. ACI Committee 544, “Guide to Design with Fiber-Reinforced 17. Mostafazadeh, M.; Abolmaali, A.; and Ghahremannejad, M., “Shear
Concrete (ACI 544.4R-18),” American Concrete Institute, Farmington Strength of Synthetic Fiber-Reinforced Concrete Box Culverts,” Journal of
Hills, MI, 2018, 4 pp. Bridge Engineering, ASCE, V. 24, No. 6, 2019, p. 04019039. doi: 10.1061/
3. Soetens, T., and Matthys, S., “Shear-Stress Transfer across a Crack in (ASCE)BE.1943-5592.0001402
Steel Fibre-Reinforced Concrete,” Cement and Concrete Composites, V. 82, 18. Kovács, I., and Balázs, G. L., “Structural Performance of Steel Fibre
Sept. 2017, pp. 1-13. doi: 10.1016/j.cemconcomp.2017.05.010 Reinforced Concrete,” Proceedings, Second International Conference on
4. Qasim, O. A., “Comparison between the Behaviors of Different Fibre Reinforced Concrete—From Research to Practice, Budapest Univer-
Concrete Types in Various Shear Tests,” Proceedings, The Fourth sity of Technology and Economics, Hungary, Nov. 19, 2004.

100 ACI Materials Journal/May 2021