ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title No. 115-S122

Cracking and Ductility Analysis of Steel Fiber-Reinforced

Prestressed Concrete Beams in Flexure
by Suhas S. Joshi, Nikesh Thammishetti, S. Suriya Prakash, and Saumitra Jain

The present study focuses on understanding the effect of steel-fiber Numerous previous works have demonstrated the use
dosage on cracking and ductile behavior of prestressed concrete of steel fibers as secondary reinforcement for concrete
beams (PCBs) under flexure using digital image correlation elements.3,4 Cuenca et al.3 and Cuenca and Serna4 used steel
(DIC). Seven prestressed concrete beams were cast and tested fibers to control the crack propagation in precast beams and
under a shear span-depth ratio (a/d) of 5 to simulate the flexure/
hollow-core slabs. They observed that the fibers and stirrups
flexure-shear dominant behavior. Full field strain measurement
had a synergic effect, resulting in better crack-control mecha-
with the DIC technique was used to understand the effectiveness
of steel fibers on the crack-bridging mechanisms. Three different nisms and enhanced tension stiffening. Fantilli et al.5 observed
volumetric fiber reinforcement ratios—0.35%, 0.70%, and 1.0%— the increase in ductility of fiber-reinforced concrete both
were considered. Test results revealed that the strain energy-based in tension and compression. They attributed the increase in
ductility increased with an increase in fiber dosage from 0.35% to ductility in compression to the passive confinement provided
1.0%. Post-cracking stiffness improved by 50% due to the addi- by the steel fibers. Padmarajaih and Ramaswamy6 predicted
tion of 1.0% volume of steel-fiber dosage. DIC measurements of the crack width in partially and fully prestressed concrete
displacements and strains were found to be in good agreement with beams with different dosages of steel fibers. They found that
the conventional linear variable displacement transducer (LVDT) addition of fibers restricted the crack initiation and propaga-
measurements. DIC results clearly established the effect of fibers tion, eventually leading to an increase in the ultimate flexural
on crack bridging and strain reduction. Concrete strains and the
strength. Harajli7 studied the bond behavior of steel fiber-
strain in the prestressing strand reduced due to the better crack
reinforced concrete zones under static and cyclic loading.
bridging of steel fibers.
The author noted that the addition of steel fibers improved
Keywords: crack bridging; digital image correlation (DIC); ductility; the bond strength, reduced the damage, and increased the
prestressed concrete beams; steel fibers. energy-dissipation capacity under cyclic loading. Sahoo et al.8
evaluated the effect of steel fibers on the behavior of concrete
INTRODUCTION beams with and without stirrups. They noted that the flexural
The use of fiber-reinforced concrete (FRC) has been capacity did not significantly improve when the fiber dosage
continuously increasing in the construction industry due was more than 0.5%. Singh9 presented the flexural modeling
to its various advantages such as improvement in post- of SFRC to determine the ultimate capacity, crack width of
cracking stiffness, flexural toughness, and ease of avail- rectangular sections using strain compatibility, and force equi-
ability at a competitive price. Steel FRC is mainly used in librium equations. Additionally, a normalized design chart for
seismic-resistant structures, tunnel construction, and blast- strength calculations considering the random distribution of
and impact-resistant structures where the post-cracking fibers and other fiber parameters was also presented.
behavior is a major concern. ACI 318-111 permits design Fantilli et al.10 proposed a unique function related to the
engineers to use steel fiber-reinforced concrete (SFRC) as a post-peak response of different cement-based materials
replacement to conventional shear reinforcement. However, for numerical studies. Tiberti et al.11 studied the influence
ACI documents1,2 mandate that SFRC beams are required to of concrete strength on the crack development of SFRC
have a minimum steel fiber dosage of 0.75% in volume and members and found that the usage of fibers led to reduced
compressive strength not greater than 42 MPa (6.091 ksi). mean crack spacing in high-strength concrete (HSC) when
This mandate is because current ACI provisions are primarily compared to normal-strength concrete (NSC). They also
based on experimental studies on non-prestressed concrete noted that the influence of HSC and the presence of fibers
beams with a compressive strength less than 42  MPa requires further research to establish a stabilized crack
(6.091  ksi). However, in a prestressed concrete beam, the pattern. Job and Ramaswamy12 developed an analytical
beneficial effect of prestressing forces could further relax model to predict crack width and crack spacing of partially
the minimum required fiber volume fraction and can make prestressed concrete with steel fibers considering the bond
the use of SFRC more economical. Thus, the focus of this slip of longitudinal reinforcement, and pullout of fibers.
investigation is to study the effect of steel fibers on cracking
and ductility behavior of prestressed concrete beams under ACI Structural Journal, V. 115, No. 6, November 2018.
flexure/flexure-shear using full field strain measurements MS No. S-2017-170.R4, doi: 10.14359/51706827, was received October 23, 2017,
and reviewed under Institute publication policies. Copyright ©  2018, American
with digital image correlation (DIC). Concrete Institute. All rights reserved, including the making of copies unless
permission is obtained from the copyright proprietors. Pertinent discussion including
author’s 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 Structural Journal/November 2018 1575

Cuenca and Serna13 observed from their experimental study Table 1—Details of test specimens
on hollow core slabs that the addition of fibers increased the Fiber content Compressive
ultimate loads as well as improved ductility when compared Specimen (volume fraction strength of cube
with the control specimens. Marazzini and Rosati14 tested No. Label in %) a/d (fck), MPa (ksi)
individual cut beams of hollow-core slabs with and without 1 SF00-1 0
steel fibers. They observed improved post-peak behavior due
2 SF00-2 0
to fiber addition. Campione15 presented the flexural behavior
of SFRC deep beams. They compared the RC deep beams 3 SF35-1 0.35
with and without fibers; SFRC beams exhibited improved 4 SF35-2 0.35 5 62 (8.99)
strength and ductility through crack bridging. Abbas and 5 SF70-1 0.7
Khan16 conducted fiber pullout tests to study fiber-matrix
6 SF70-2 0.7
interfacial behavior of hooked-end steel fiber-reinforced
concrete. They concluded that ultimate pullout load was 7 SF100-2 1.0
found to increase with an increase in fiber size and embed-
ment length. Gencturk et al.17 used DIC technique for under- Table 2—Mixture design details
standing the shear behavior of prestressed concrete I-beams. Material Quantity, kg/m3
They recorded the values of displacement and strain using
20 mm aggregate 754
both DIC and conventional measurements. They highlighted
the need for further research in using the DIC technique 10 mm aggregate 355
because of non-availability of standard algorithms to identify Crushed stone sand (CSS) 415
cracks, crack width, and loss of data in the region of excessive Natural river sand (NRS) 313
spalling. Rimkus et al.18 proposed an algorithm to identify the
Cement (OPC 53) 428
crack evolution and its spacing using digital images of rein-
forced concrete, which can eliminate the subjective judgment Fly ash 22
using traditional methods. The flexural cracking behavior of Water 165
reinforced concrete beams made up of plain concrete (NSC
Admixture 2.5
and HSC) and steel fiber-reinforced high-strength concrete
was studied using the DIC technique by Hamrat et al.19 The different steel fiber dosages. Four different series of beams
authors were able to detect the first crack with high precision corresponding to fiber dosages of 0, 0.35%, 0.70%, and 1.0%
and measure the crack width during the testing using DIC. were cast. Two beams were tested at each fiber dosage to
They have also observed an increase in the cracking load and ensure the consistency of test results. All these beams were
a reduction in the crack spacing and crack width due to the tested at a shear span-depth ratio (a/d) of 5 to ensure flexure/
presence of steel fibers. flexure shear behavior. A strain energy-based ductility
measurement was employed to study the influence of fibers.
RESEARCH SIGNIFICANCE This work is part of a larger research program investigating
The research carried out in the past has confirmed that the the effect of various types of fibers and their combinations on
addition of steel fibers contributes to improved post-cracking the performance improvement of prestressed concrete beams
behavior through increased ductility, multiple crack forma- under different a/d. Currently, work is in progress to under-
tion/crack distribution, reduced crack width, and enhanced stand the effect of fibers on shear behavior of prestressed
toughness properties.20-23 Various factors such as shear concrete beams at a low a/d of 2.5. However, only the results
span-depth ratio (a/d), compressive and tensile strengths pertaining to a/d = 5 are presented in this manuscript. All the
of concrete, and fiber dosage influence the performance of beams were cast on the same day and were water-cured for
fiber-reinforced concrete elements. However, only limited 28 days at room temperature. The overall test matrix with
results are available on the behavior of steel fiber-rein- study parameters is detailed in Table 1.
forced prestressed concrete members made of concrete with
strengths higher than 50 MPa (7250 psi). The broad objec- MATERIAL PROPERTIES AND BEHAVIOR
tive of this study is to investigate the influence of steel fibers Concrete
on the response of prestressed concrete beams and under- All specimens were cast in the precast plant using concrete
stand its effect on the crack propagation and post-cracking designed as per IS 10262-200925 to have a 28-day target
behavior using full-field strain measurements with the help compressive strength of 58 MPa (8.41 ksi). Blended coarse
of the DIC technique. In addition, the influence of steel fiber aggregates of size 10 and 20 mm (0.39 and 0.78 in.) along
dosage on aspects such as crack control, energy absorption with fine aggregates were used to obtain a uniform mixture.
capacity, and deflection is also presented. Mixture design details are given in Table 2. The unit weight
of concrete was 2400 kg/m3 (149.82 lb/ft3). The cube strength
EXPERIMENTAL INVESTIGATION of concrete was 62 MPa (8.99 ksi) with a standard devia-
The experimental program consists of casting and testing tion of 2.2 MPa (319.08 psi) at 28 days. The SFRC cylin-
of the full-scale prestressed concrete beams designed as ders were tested using servo-controlled compression testing
per IS 1343-2012,24 having a length of 3.5 m (11.48 ft) and machine for the complete stress-strain behavior (Fig. 1).
cross section of 200 x 300 mm (7.87 x 11.81 in.) containing

1576 ACI Structural Journal/November 2018

Fig. 1—Cylinder testing under compression.

Fig. 2—Testing of FRC specimen under pure tension.

Behavior of concrete in compression predominantly due to a single explicit crack at the center
Concrete cylinders were tested (Fig. 1(a)) to obtain the of the specimen. This observation indicates that the stress
stress-strain curves with and without steel fibers. The addi- concentration in the crack region reduces the post-cracking
tion of steel fibers was found to significantly improve the load resistance of the specimen. On the other hand, the FRC
strength and stiffness degradation in the post-peak region specimen showed a similar large crack at failure, but the
with better ductility (Fig. 1(b)). Previous work of authors crack opening was delayed due to the presence of fibers in
of this manuscript has also confirmed the same at both room the section. Similar observations have been made by Choun
and moderate temperature exposure.26-28 and Park,29 as shown in Fig. 2(b).

Concrete tests in tension Internal reinforcement—prestressing steel strands

The control and fiber-reinforced concrete (FRC) dog-bone Two 1/2 in. (12.7 mm diameter) strands containing seven
specimens were tested using a servo-controlled hydraulic low-relaxation wires with an effective area of 200 mm2
MTS fatigue testing machine under pure tension. The (0.31  in.2) were used for prestressing the beams. Coupon
loading was controlled at a rate of 10 μm/min up to a midspan specimens were prepared for tendons and tested under
displacement of 0.8 mm (0.03 in.) and then at 50 μm/min up tension using a servo-controlled machine. The ultimate
to 1.2 mm (0.047 in.) displacement followed by 1000 μm/ tensile strength and modulus of elasticity were found to be
min for the rest of the test. All the specimens with different 1860 MPa (269.77 ksi) and 196.5 GPa (28,500 ksi), respec-
fiber dosages were tested in uniaxial tension using rigid steel tively. Jacking force was applied to each of the strands to
plates on a servo-controlled closed-loop system, as shown subject them to an initial strain of 0.004.
in Fig. 2. The failure pattern in the control specimen was

ACI Structural Journal/November 2018 1577

Hooked-end steel fibers Digital image correlation (DIC)
Hooked-end steel fibers having length and diameter of 30 Digital image correlation (DIC) is a technique for
and 0.6 mm (1.18 and 0.023 in.), respectively (Fig. 3), were measuring the entire field strain and displacement of speci-
used in this investigation. The tensile strength and modulus of mens. DIC works by comparing two images of the specimen
elasticity of steel fibers were found to be 1000 MPa (145 ksi) coated with a random speckle pattern in an undeformed and
and 200 GPa (29,007 ksi), respectively. During casting, the deformed state.30 Images of the object’s surface before and
fibers were added gradually to prevent the balling effect and after deformation are recorded, digitized, and stored in the
to obtain a uniform fiber distribution. Uniform fiber distri- computer. These images are then analyzed to determine the
bution was verified by the inspection of the cracked faces of displacements by invoking a pattern-matching principle.
failed beams after testing. Because it is impossible to find matching points using a single
pixel, areas (called as subsets) containing multiple pixels are
Instrumentation used for the analysis.30 The size of subset varies with respect
Displacements in all the specimens were recorded using to the experimentation details. The step size controls the
linear variable displacement transducers (LVDTs) and DIC. density of analyzed data. For example, a step size of 5 will
LVDTs were positioned at specific locations along the length analyze every fifth point in each direction. A higher step size
of the beam to capture the entire curvature profile during gives faster results but coarser data. A smaller step size will
testing. Two 100 mm (3.94 in.) LVDTs were positioned at return more points but will take more computation time. For
the center of the span to capture the midspan deflection. Six the present experiments, a subset size of 35 and a step size
LVDTs were used to form a strain rosette (Fig. 4) to measure of 5 was chosen after a thorough sensitivity analysis. All the
surface strains near the loading point. Strain gauges of 5 mm specimens had a speckle pattern on the surface for capturing
(0.197 in.) gauge length were instrumented at the midspan of of DIC images. The surface was initially coated with nonre-
the prestressing strands to capture the strain variation in the flective white paint, and then black speckle was sprayed on
strands during testing. the white coat. Two halogen lights were placed at an angle to
the specimen as shown in Fig. 5 to illuminate the specimen.
The camera was placed in front of the specimen with its axis
normal to the specimen. Images were taken at regular inter-
vals and were processed using specialized software for strain
analysis as well as crack initiation and propagation.

Test setup and loading details

All beams were tested in a four-point bending configu-
ration by which a constant moment region at the midspan
was generated. Figure 6 illustrates the components used in
the test setup. A 250 kN (56.25 kip) MTS servo-hydraulic
actuator was used to apply the loads. The actuator load is
transferred to the concrete specimens through longitudinal
rigid steel spreader beam and thereafter to the I-beams, as
shown in Fig. 6. Loading was applied monotonically in
displacement control mode at a rate of 0.05 mm/s. Loading
was paused intermittently to observe and mark the cracks
and study the failure progression.

Fig. 3—Hooked-end steel fibers.

Fig. 4—Instrumentation details: LVDT and strain rosette arrangement.

1578 ACI Structural Journal/November 2018

 TEST RESULTS AND DISCUSSION Fig. 7. Two specimens were tested for each fiber dosage to
Load-deflection behavior ensure the consistency of the results. The load-displacement
All the beams were tested with a/d = 5 and the load- behavior of PSC beams without fibers (control specimen) are
versus-midspan deflection plots were obtained, as shown in illustrated in Fig. 7(a). The average cracking load observed
from the test was 60 kN (13.5 kip). After cracking, the beam
continued to resist the applied load until 142.5 kN (32.06 kip)
and the corresponding deflection was 40 mm (1.57 in.).
It is worth mentioning that the beam was heavily under-
reinforced, leading to the yielding of strands just before the
peak load.
The beam with 0.35% steel fibers exhibited better ductility
as compared to the control specimen. An approximately 13%
improvement in cracking load and 23% (average) improve-
ment in post-cracking stiffness was observed (Fig.  7(b)).
However, the ultimate load did not change. This beam
had better crack distribution with the delayed formation of
cracks as compared to the control specimen before reaching
its final failure load. Flexure-dominant failure was observed
Fig. 5—Schematic plan view of DIC setup. with 0.35% steel fiber dosage. The crack bridging effect was

Fig. 6—Experimental setup and instrumentation for prestressed concrete beams.

ACI Structural Journal/November 2018 1579

 dosage had honeycombing due to improper compaction and,
hence, it was excluded from comparisons.

Crack distribution and failure modes

The crack distribution of tested specimens is illustrated in
Fig. 8. For control specimens, the first flexural crack occurred
at the tension fiber in the constant moment region. Soon after
cracking, the formation of multiple cracks continued at the
midspan location below the loading point. All the specimens
failed in flexure mode except the control specimen, which
failed in the flexure-shear mode soon after the yielding of
prestressing strands. The failure mode in the control spec-
imen was sudden, due to the crushing of concrete just after
the yielding of strands. The presence of steel fibers resulted
in more distribution of cracks and converted sudden brittle
failure to a gradual, ductile failure mode.

Load versus strain behavior

Strain gauges were installed on the strand at the midlength
of the specimen to measure the strain variation during
pretensioning and testing. Strain gauges recorded the value
of around 4000 µm/m during pretensioning. Load-versus-
strain graphs presented in Fig. 9 show a trilinear behavior.
The first change in slope can be attributed to the cracking of
beam. The second change in slope can be correlated to the
beginning of crushing of concrete in compression. Figure 9
indicates that strain in the strands was nominal before cracking.
After cracking, as expected, the strain increased significantly,
indicating the higher contribution of strands in load resistance.
In general, the strain values at a particular load reduced with
an increase in fiber dosage, indicating the steel fiber contri-
bution in load resistance after cracking (Fig. 9). The strands
yielded in all the specimens before they could reach their
ultimate load. However, strain values were not recorded
after a certain load value level due to malfunctioning of the
Fig. 7—Comparison of load-deflection behavior. (Note: strain gauges.
1 mm = 0.039 in.)
evident through the formation of numerous smaller cracks Fiber distribution on fractured surface
and fewer major cracks. The specimen with 0.7% steel The fiber distribution on the cracked section was verified
fiber dosage (Fig. 7(c)) exhibited better ductility than the after testing of PSC beams to complete failure. Figure 10
specimen with 0.35% fiber dosage. However, there was no shows the distribution of steel fibers on the failure surface. It
substantial improvement in ultimate strength due to higher presents the random and uniform distribution of fibers along
fiber dosage. Additionally, significant improvement in the the depth of the section. Moreover, the elongation of fibers
post-cracking stiffness (50%), ductility, and ultimate load and fiber pullout (bond failure) illustrates the contribution of
(11.6%) was observed with the increase in fiber dosage from fibers in load resistance. Nevertheless, the fiber contribution
0% to 1.0%. The specimen had an ultimate load of 159 kN in crack bridging was evident from the fractured surfaces of
(35.77 kip) with the first crack forming at a load of 70 kN the failed beams.
(15.75 kip). The test was terminated at 100 mm (3.94 in.)
displacement due to a limitation in stroke capacity of the Energy absorption capacity
actuator. However, the specimen could have resisted loads The strain energy developed to resist the applied external
at even higher displacements, which is evident from the load was calculated for comparison. Energy absorption
overall load-displacement curves (Fig. 7(d)). It was evident capacity of PSC beams can be obtained from the area under
from the test results that the presence of steel fiber increases load-deflection curve. All the specimens were tested at same
the displacement corresponding to cracking load. The addi- a/d, but with different fiber dosages. Their energy absorption
tion of fibers also converted the brittle flexure-shear failure capacity is compared and is presented in Table 3. The addi-
of control beams into ductile flexure-dominant failure in tion of steel fibers resulted in an increase of energy absorp-
fiber-reinforced beams. One specimen with 1.0% steel fiber tion capacity due to better crack bridging. Energy capacity
doubled when 1.0% by volume of steel fibers were added to
the control specimen.

1580 ACI Structural Journal/November 2018

Fig. 8—Crack distribution and failure modes.

Fig. 9—Effect of steel fibers on strain variation of strands in PCB.

ACI Structural Journal/November 2018 1581

Fig. 10—Steel fiber distribution on failure surface.
ANALYSIS OF RESULTS USING DIGITAL IMAGE the tensile zone of all the tested beams. Figure 12(b) shows
CORRELATION the tensile strain variation along the line a-b at a load level
Load versus deflection analysis of 60 kN (13.48 kip), which is slightly less than the cracking
Displacements and strains from DIC analysis can be load. It indicates that the peak strain along the considered
correlated only until major crack formation. After significant horizontal line is same for beams of different series. This
cracking, the correlation breaks due to excessive cracking elucidates the fact that fibers do not play a vital role in
and spalling of concrete, leading to the difficulty in pattern load-carrying mechanisms before cracking. On the other
matching. The load-deflection response of the specimens hand, Fig. 12(c) shows the tensile strain variation along line
using DIC analysis is compared with the LVDT measure- a-b at a load of 70 kN (15.73 kip), which is greater than
ments for beams with different fiber dosages (Fig. 11). The the cracking load. It presents the decrease in peak strain
results from DIC analysis showed a close agreement with at same load level after cracking with the increase in fiber
the LVDT measurements. dosage. The reduction in strain indicates that fibers resist the
crack widening and localization, and helps to redistribute
Effect of fibers on tensile strain in concrete the stresses and increase the number of cracks. The extent
Full field strains on the surfaces were measured using the of reduction in strains mainly depends on the dosage of
specialized software. Strain distribution was evaluated to see steel fibers. In Fig. 12(a) to (c), a sudden jump in the strain
the effect of steel fiber contribution in resisting the crack contour indicates the presence of a crack in that region.
propagation. A horizontal line a-b was chosen (Fig. 12(a)) in Figure 13 illustrates the reduction in average strain along

1582 ACI Structural Journal/November 2018

Table 3—Summary of test results
Specimen SF00-1 SF00-2 SF35-1 SF35-2 SF70-1 SF70-2 SF100-2
Cracking load Pcr, kN (kip) 55 (12.36) 66 (14.84) 68 (15.28) 70 (15.73) 70 (15.73) 67 (15.06) 70.3 (15.80)
Deflection at cracking load ∆cr, mm
2.5 (0.098) 2.97 (0.117) 3.02 (0.119) 3.28 (0.13) 4.03 (0.158) 4.02 (0.158) 4.66 (0.183)
(in.)
Peak load Ppl, kN (kip) 141 (31.7) 144 (32.37) 143 (32.14) 143.5 (32.26) 148.6 (33.4) 146 (32.82) 159 (35.75)
Increase in Ppl, % — — 0.35 0.7 4.28 2.45 11.57
Deflection at peak load ∆pl, mm (in.) 36.65 (1.44) 42.2 (1.66) 34.4 (1.35) 29.6 (1.17) 45.0 (1.77) 32.4 (1.27) 36.6 (1.44)
Midspan deflection at failure, mm (in.) 65.0 (2.56) 53.3 (2.09) 97.5 (3.83) 78.8 (3.10) 99.0* (3.89) 97.8* (3.85) 99.7* (3.92)
Post-cracking stiffness, kN/mm (kip/in.) 2.57 (14.73) 2.59 (14.79) 2.903 (16.578) 3.457 (19.744) 3.076 (17.567) 3.335 (19.046) 3.871 (22.107)
Increase in post-cracking stiffness, % — — 12.54 34.03 19.26 29.3 50.07
Experimental cracking moment, kN-m
34.4 (25,372) 37.5 (27,658) 42.5 (31,346) 43.7 (32,231) 43.7 (32,231) 41.9 (30,903) 43.9 (32,378)
(lb-ft)
Experimental peak moment, kN-m
88.1 (64,979) 89.4 (65,938) 89.4 (65,938) 89.7 (66,159) 92.9 (68,519) 91.3 (67,339) 99.4 (73,313)
(lb-ft)
Cracking moment (RILEM), kN-m
34.7 (25,593) 39 (28,764) 43.3 (31,936) 46.7 (34,444)
(lb-ft)
Peak moment (RILEM), kN-m (lb-ft) 81 (59,742) 86 (63,430) 91 (67,118) 95 (70,068)
Strain energy, J 6332 6358 9309 10,989 11,534 11,774 13,239
Percentage increase in strain energy — — 46.7 73.2 81.8 85.6 108.7
*
Test was terminated due to limitation in stroke capacity of actuator.

Fig. 11—Comparison of DIC results with LVDT measurement.

ACI Structural Journal/November 2018 1583

Fig. 12—Longitudinal strain contours (DIC) on horizontal line a-b before and after cracking.
the level of crack tip at failure due to addition of fiber. Addi- SFRC. Later, the models developed are used to perform
tionally, Fig. 13 also indicates that the presence of steel fiber the moment-curvature analysis of the section considered in
reduces the maximum strain as much as by 50%. The strain this research.
values corresponding to the post-cracking loads for different
specimen highlights that the presence of steel fibers reduces Constitutive relationships for SFRC under
the strain at the crack locations (Fig. 14). compression
Several analytical models are available in the litera-
PREDICTIONS USING ANALYTICAL MODELS ture to predict the behavior of SFRC under compression. Ou
The behavior of SFRC under compressive and tensile loads et al.32 proposed an analytical model for the development
are predicted theoretically based on the models available in of compressive stress-strain relationship of SFRC, which is
literature.30-32 The following sections discuss the method- an extension of the stress-strain model previously proposed
ology followed to obtain the constitutive relationships of by Carreira and Chu31 for plain concrete under compression

1584 ACI Structural Journal/November 2018

 ε  εcf = εco + 0.0007(RIv) (3)
β c 
σc  ε cf  β = 0.71(RIv)2 – 2.00(RIv) + 3.05 (4)
= β
(1)
f cf′ ε 
β −1+  c  l
 ε cf  RI v = V f (5)
φ
where σc is compressive stress; fcf′ is the compressive
strength; εcf is the peak strain; εc is the compressive strain, where fc′ and εco are compressive strength (in MPa) and the
and β defines the shape of the stress-strain curve, which is a strain (mm/mm) at the peak stress of plain concrete, respec-
function of the reinforcement index (RIv) tively; and Vf, l, and  are volume fraction of fibers (in %),
and length and diameter of fibers, respectively.
fcf′ = fc′ + 2.35(RIv) (2) The cylinder compressive strength (43 MPa [6235 psi]) of
the control specimen is used for the development of stress-
strain behavior of SFRC with different volume fraction of
fibers employing the model defined by Eq. (1) to (5).

Constitutive relationships for SFRC under tension

The tensile behavior of SFRC usually depends on the
strength of the plain concrete, type of fiber, and fiber fraction.
Neocleous et al.33 developed an analytical model to predict
the behavior of SFRC under tension, adopting recommenda-
tions of RILEM34 for predicting the tensile strength of plain
concrete. Typical behavior of SFRC is shown in the Fig. 15.
Various salient points in the tensile constitutive relation-
ships for SFRC are computed as follows

fc′ = 0.8(fck) (6)

Fig. 13—Average strain along level of crack tip at failure.

Fig. 14—DIC images showing strain contours at various load levels.

ACI Structural Journal/November 2018 1585

Fig. 15—Stress-strain response of fiber-reinforced concrete in compression32 and tension.33

Fig. 16—Layer-by-layer approach for sectional analysis of prestressed section. (Note: 1 mm = 0.0394 in.)
fctm = 0.3(fc′)2/3 (7) Moment-curvature analysis
A layer-by-layer method of sectional analysis was carried
ffcm = fc′ + kxSp (8) out to establish the moment-curvature behavior of steel
fiber-reinforced prestressed concrete sections using an iter-
Ec = 9500(ffcm)2/3 (9) ative procedure. To maximize the accuracy of the results
and to minimize the computational effort, the cross section
(1600 − d ) and corresponding stress-strain curves were discretized into
σ1 = 0.7 × f ctm × (10)
1000 numerous layers of small thickness to get the uniform strain
throughout the thickness (Fig. 16). The constitutive relation-
σ1f = σ1 × (1 + 0.32 × Vfw) (11) ship for fiber-reinforced concrete in compression and tension
according to the model32,33 (Fig. 15) is used. A compressive
top fiber strain value is fixed, and the neutral axis depth is
σ 2 f = σ1 f × eξ ( ε1 − 0.002 )10
3
(12)
iterated to satisfy the force equilibrium. Moment resistance
is calculated from the moment equilibrium, and the corre-
σ3f = 0 (13)
sponding curvature can be derived using the assumed top
compressive strain and the neutral axis depth.
ε1f = σ1f /Ec (14)
The analytical results are compared with experimentally
measured moment-curvature response in Fig. 17. It can be
 2  observed that post-cracking stiffness and ultimate strength
ε 2 f = ε1 f +  (15)
 1000  increases with increase in fiber dosage. The cracking and
ultimate moment were found to be matching reasonably well
ε3f = 0.04 (16) with the test results (Fig. 17). However, after certain loading,
curvature was not measured due to instrumentation error.
where fck and fc′ are the characteristic compressive strength Hence, overall response was not presented. RILEM recom-
of cube and cylinder (in MPa) respectively; fctm, ffcm, and Ec mendations34 were also used to compare the experimental
are the characteristic tensile strength, mean tensile strength, results at ultimate capacity, as shown in Table 3.
and mean secant modulus of concrete (in MPa), respec-
tively; Vfw is the weight fraction of fibers (in %); σ1f, σ2f, and SUMMARY AND CONCLUSIONS
σ3f are the salient points of the tensile stress-strain model Pre-tensioned prestressed concrete beams with and
corresponding to strains of ε1f, ε2f, and ε3f, respectively. without steel fibers were cast and tested under four-point

1586 ACI Structural Journal/November 2018

Fig. 17—Comparison of analytical and experimental results. (Note: 1 kN-m = 0.736 kip-ft.)
bending to evaluate the role of steel fibers in crack bridging Nikesh Thammishetti is a Research Scholar of the structural engineering
division, Indian Institute of Technology Hyderabad. He received his BTech
and improvement in serviceability performance. Based on from Kakatiya Institute of Technology and Science Warangal, Telangana-
the test results of seven prestressed concrete beams presented India, and his master’s degree from Birla Institute of Technology Pilani
in this study, the following major conclusions can be drawn: (BITS Pilani), Pilani, India. His research interests include mechanics of
reinforced concrete, repair, and rehabilitation of reinforced concrete struc-
1. Though the addition of steel fibers showed marginal tures using innovative ductile materials.
improvement in the peak strength, it resulted in very
good improvement in strain energy-based ductility of the S. Suriya Prakash is an Associate Professor in the Department of Civil
Engineering, Indian Institute of Technology Hyderabad. He received his
prestressed concrete beams. The major contribution of steel PhD from Missouri S&T, Rolla, MO. His research interests include seismic
fibers was in bridging the cracked surfaces, which resulted in design, repair, and rehabilitation of reinforced concrete and masonry struc-
improved post-cracking stiffness and higher ultimate deflec- tures using advanced construction materials.
tion. The strain in the strand reduced due to the addition of Saumitra Jain is a Graduate Student of the structural engineering divi-
steel fibers in the post-cracking regime. sion in the Department of Civil Engineering, Indian Institute of Technology
2. Distribution of cracks increased with an increase in the Hyderabad. He received his BE from SGSITS Indore, Indore, India. His
research interests include the behavior of reinforced concrete members,
addition of steel fibers. More importantly, the failure mode structural health monitoring of concrete members, and strengthening of
changed from flexure-shear to flexure, resulting in higher concrete members using FRP composites.
ductility and ultimate deflections.
3. No increase in ultimate flexural strength was observed ACKNOWLEDGMENTS
at low fiber dosages. The maximum increase in load capacity This experimental work was carried out as part of the project funded
through Utchattar Avishkar Yojana Scheme of Indian Government. PRECA
was approximately 11.6% at a higher fiber dosage of 1.0%. India Pvt. Ltd. and Grenix India, Ltd., donated the materials required for
4. Energy dissipation capacity increased with an increase this research. PRECA provided the labor required for the casting of speci-
in fiber dosage. It doubled when 1.0% by volume of steel mens in their factory. All their support is duly acknowledged.
fibers was added.
5. DIC measurements correlated well with conventional REFERENCES
1. ACI Committee 318, “Building Code Requirements for Structural
measurements of displacement sensors. The reduction in Concrete (ACI 318-11) and Commentary,” American Concrete Institute,
peak strain in tensile fibers was up to 50% due to the crack Farmington Hills, MI, 2011, 503 pp.
bridging mechanism of steel fibers. 2. ACI Committee 544, “Report on Fiber-Reinforced Concrete (ACI
544.1R-96),” American Concrete Institute, Farmington Hills, MI, 1996, 66 pp.
3. Cuenca, E.; Echegaray-Oviedo, J.; and Serna, P., “Influence of Concrete
AUTHOR BIOS Matrix and Type of Fiber on the Shear Behavior of Self-Compacting Fiber
Suhas S. Joshi is a Graduate Student of the structural engineering divi- Reinforced Concrete Beams,” Composites. Part B, Engineering, V. 75,
sion in the Department of Civil Engineering, Indian Institute of Technology 2015, pp. 135-147. doi: 10.1016/j.compositesb.2015.01.037
Hyderabad, Telangana, India. His research interests include understanding 4. Cuenca, E., and Serna, P., “Shear Behavior of Prestressed Precast
the behavior of reinforced and prestressed concrete and use of fiber-rein- Beams Made of Self-Compacting Fiber Reinforced Concrete,” Construc-
forced polymer (FRP) composites in construction. tion and Building Materials, V. 45, 2013, pp. 145-156. doi: 10.1016/j.
conbuildmat.2013.03.096

ACI Structural Journal/November 2018 1587

 5. Fantilli, A. P.; Mihashi, H.; Vallini, P.; and Chiaia, B. M., “Equiva- nique,” Construction and Building Materials, V. 106, 2016, pp. 678-692.
lent Confinement in HPFRCC Columns Measured by Triaxial Test,” ACI doi: 10.1016/j.conbuildmat.2015.12.166
Materials Journal, V. 108, No. 2, Mar.-Apr. 2011, pp. 159-167. 20. Narayanan, R., and Darwish, I. Y. S., “Shear in Prestressed Concrete
6. Padmarajaiah, S. K., and Ramaswamy, A., “Flexural Strength Predic- Beams Containing Steel Fibres,” International Journal of Cement
tions of Steel Fiber Reinforced High-Strength Concrete in Fully/Partially Composites and Lightweight Concrete, V. 9, No. 2, 1987, pp. 81-90. doi:
Prestressed Beam Specimens,” Cement and Concrete Composites, V. 26, 10.1016/0262-5075(87)90023-6
No. 4, 2004, pp. 275-290. doi: 10.1016/S0958-9465(02)00121-X 21. Peaston, C.; Elliot, K.; and Paine, K., “Steel Fiber Reinforcement
7. Harajli, M. H., “Bond Behavior in Steel Fiber-Reinforced for Extruded Prestressed Hollow Core Slabs,” Structural Applications of
Concrete Zones under Static and Cyclic Loading: Experimental Eval- Fiber-Reinforced Concrete, SP-182, American Concrete Institute, Farm-
uations and Analytical Modelling,” Journal of Materials in Civil Engi- ington Hills, MI, 1999, pp. 87-108.
neering, ASCE, V.  22, No. 7, 2010, pp. 674-686. doi: 10.1061/(ASCE) 22. Barros, J. A., and Figueiras, J. A., “Flexural Behavior of SFRC: Testing
MT.1943-5533.0000067 and Modelling,” Journal of Materials in Civil Engineering, ASCE, V. 11,
8. Sahoo, D. R.; Bhagat, S.; and Reddy, T. C. V., “Experimental Study No. 4, 1999, pp. 331-339. doi: 10.1061/(ASCE)0899-1561(1999)11:4(331)
on Shear-Span to Effective-Depth Ratio of Steel Fiber Reinforced Concrete 23. Soulioti, D. V.; Barkoula, N. M.; Paipetis, A.; and Matikas, T. E.,
T-Beams,” Materials and Structures, V. 49, No. 9, 2016, pp. 3815-3830. “Effects of Fibre Geometry and Volume Fraction on the Flexural Behaviour
doi: 10.1617/s11527-015-0756-6 of Steel-Fibre Reinforced Concrete,” Strain, V. 47, 2011, pp. e535-e541.
9. Singh, H., “Flexural Modeling of Steel Fiber-Reinforced Concrete doi: 10.1111/j.1475-1305.2009.00652.x
Members: Analytical Investigations,” Practice Periodical on Structural 24. IS 1343, “Code of Practices for Prestressed Concrete,” Bureau of
Design and Construction, V. 20, No. 4, 2015, p. 04014046 doi: 10.1061/ Indian Standards, New Delhi, India, 2012, 62 pp.
(ASCE)SC.1943-5576.0000244 25. IS 10262 “Proportioning – Guideline, Indian Standard Concrete
10. Fantilli, A. P.; Mihashi, H.; and Vallini, P., “Post-Peak Behavior of Mix,” Bureau of Indian Standards, New Delhi, India, 2009.
Cement-Based Materials in Compression,” ACI Materials Journal, V. 104, 26. Rasheed, M. A., and Prakash, S. S., “Mechanical Behavior of Sustain-
No. 5, Sept.-Oct. 2007, pp. 501-510. able Hybrid-Synthetic Fiber Reinforced Cellular Lightweight Concrete for
11. Tiberti, G.; Minelli, F.; Plizzari, G. A.; and Vecchio, F. J., “Influ- Structural Applications of Masonry,” Construction and Building Materials,
ence of Concrete Strength on Crack Development in SFRC Members,” V. 98, 2015, pp. 631-640. doi: 10.1016/j.conbuildmat.2015.08.137
Cement and Concrete Composites, V. 45, 2014, pp. 176-185. doi: 10.1016/j. 27. Jain, S.; Prakash, S. S.; and Subramaniam, K. V. L., “Monitoring
cemconcomp.2013.10.004 of Concrete Cylinders with and without Steel Fibers under Compression
12. Job, T., and Ramaswamy, A., “Crack Width in Partially Prestressed Using Piezo-Ceramic Smart Aggregates,” Journal of Nondestructive Eval-
T-Beams Having Steel Fibers,” ACI Structural Journal, V. 103, No. 4, uation, V. 35, No. 4, 2016, p. 59 doi: 10.1007/s10921-016-0376-2
July-Aug. 2006, pp. 569-576. 28. Srikar, G.; Goudar, A. G.; and Prakash, S. S., “A Study on Residual
13. Cuenca, E., and Serna, P., “Failure Modes and Shear Design of Compression Behavior of Structural Fiber-Reinforced Concrete Exposed
Prestressed Hollow Core Slabs Made of Fiber-Reinforced Concrete,” to Moderate Temperature Using Digital Image Correlation,” Interna-
Composites. Part B, Engineering, V. 45, No. 1, 2013, pp. 952-964. doi: tional Journal of Concrete Structures and Materials, V. 10, No. 1, 2016,
10.1016/j.compositesb.2012.06.005 pp. 75-85. doi: 10.1007/s40069-016-0127-x
14. Marazzini, M., and Rosati, G., “Fiber-Reinforced High-Performance 29. Choun, Y. S., and Park, H. K., “Containment Performance Evaluation
Concrete Beams Material and Structural Behavior,” Structural Applications of Prestressed Concrete Containment Vessels with Fiber Reinforcement,”
of Fiber-Reinforced Concrete, SP-182, American Concrete Institute, Farm- Nuclear Engineering and Technology, V. 47, No. 7, 2015, pp. 884-894. doi:
ington Hills, MI, 1999, pp. 29-52. 10.1016/j.net.2015.07.003
15. Campione, G., “Flexural Behavior of Steel Fibrous Reinforced 30. Schreier, H.; Orteu, J. J.; and Sutton, M. A., Image Correlation
Concrete Deep Beams,” Journal of Structural Engineering, ASCE, V. 138, for Shape, Motion and Deformation Measurements, Springer, Berlin,
No. 2, 2012, pp. 235-246. doi: 10.1061/(ASCE)ST.1943-541X.0000442 Germany, 2009.
16. Abbas, M. Y., and Khan, M. I., “Fiber-Matrix Interfacial Behavior 31. Carreira, D. J., and Chu, K. H., “Stress-Strain Relationship for
of Hooked-End Steel Fiber-Reinforced Concrete,” Journal of Materials in Plain Concrete in Compression,” ACI Journal Proceedings, V. 82, No. 6,
Civil Engineering, ASCE, V. 28, No. 11, 2016, p. 04016115 doi: 10.1061/ Nov.-Dec. 1985, pp. 797-804.
(ASCE)MT.1943-5533.0001626 32. Ou, Y.; Tsai, M.; Liu, K.; and Chang, K., “Compressive Behavior of
17. Gencturk, B.; Hossain, K.; Kapadia, A.; Labib, E.; and Mo, Y. L., Steel-Fiber Reinforced Concrete with a High Reinforcing Index,” Journal
“Use of Digital Image Correlation Technique in Full-Scale Testing of of Materials in Civil Engineering, ASCE, V. 24, No. 2, 2012, pp. 207-215.
Prestressed Concrete Structures,” Measurement, V. 47, 2014, pp. 505-515. doi: 10.1061/(ASCE)MT.1943-5533.0000372
doi: 10.1016/j.measurement.2013.09.018 33. Neocleous, K.; Tlemat, H.; and Pilakoutas, K., “Design Issues for
18. Rimkus, A.; Podviezko, A.; and Gribniak, V., “Processing Digital Concrete Reinforced with Steel Fibers,” Journal of Materials in Civil
Images for Crack Localization in Reinforced Concrete Members,” Procedia Engineering, ASCE, V. 18, No. 5, 2006, pp. 677-685. doi: 10.1061/
Engineering, V. 122, 2015, pp. 239-243. doi: 10.1016/j.proeng.2015.10.031 (ASCE)0899-1561(2006)18:5(677)
19. Hamrat, M.; Boulekbache, B.; Chemrouk, M.; and Amziane, S., 34. RILEM TC 162-TDF, “Test and Design Methods for Steel Fibre
“Flexural Cracking Behavior of Normal Strength, High Strength and High Reinforced Concrete,” Materials and Structures, V. 35, No. 9, 2002,
Strength Fiber Concrete Beams, Using Digital Image Correlation Tech- pp. 579-582.

1588 ACI Structural Journal/November 2018