Journal of Constructional Steel Research 113 (2015) 115–126

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

Review

Static behavior of stud shear connectors in elastic concrete–steel

composite beams
Qinghua Han a,b, Yihong Wang a, Jie Xu a,b,⁎, Ying Xing a
a
School of Civil Engineering, Tianjin University, Tianjin 300072, China
b
Key Laboratory of Coast Civil Structure and Safety of Ministry of Education, Tianjin University, Tianjin 300072, China

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

Article history: Elastic concrete was first introduced into steel–concrete composite beams due to its superior deformability. The
Received 30 December 2014 static behavior of stud shear connectors embedded in elastic concrete is studied in this paper. Eighteen push-out
Received in revised form 19 May 2015 tests were conducted to evaluate the load-slip behavior, bearing capacity and ultimate slip of shear studs. Four
Accepted 15 June 2015
different rubber contents, 0%, 5%, 10% and 15% were taken into consideration. Test results show that the ductility
Available online 19 June 2015
of stud improves significantly with the increasing rubber content. Especially, when the rubber content reaches
Keywords:
10%, the shear stud has relatively high bearing capacity, better deformation and better ductility. In specimens
Elastic concrete with 5% rubber content elastic concrete, shear stud shows a more ductile behavior embedded in lower compres-
Composite beam sive strength elastic concrete and the diameter has little influence on ductility and stiffness of studs. The equa-
Stud shear connectors tions provided by AASHTO LRFD, Eurocode-4 and GB50017-2003 can still apply to shear studs embedded in
Push-out test elastic concrete. Compared with the experimentally obtained bearing capacities, AASHTO LRFD is confirmed to
Crumb rubber content be the closest one.
Ductility © 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
2. Material properties of elastic concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
2.1. Raw materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.2. Compressive strengths and elastic modulus of elastic concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.3. Compressive stress–strain curves of elastic concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
2.4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
3. Push-out test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.1. Test set-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.2. Test results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
3.2.1. Modes of failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
3.2.2. Bearing capacity and ultimate slip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.1. Bearing capacity, stiffness and ductility of the shear stud in elastic concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.1.1. Effect of rubber contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.1.2. Effect of elastic concrete compressive strengths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.1.3. Effect of stud diameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.2. Stress mechanism of studs in push-out test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.3. Comparison between design codes and test results on ultimate strength of studs . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

⁎ Corresponding author at: School of Civil Engineering, Tianjin University, Tianjin 300072, China.
E-mail address: jxu@tju.edu.cn (J. Xu).

http://dx.doi.org/10.1016/j.jcsr.2015.06.006
0143-974X/© 2015 Elsevier Ltd. All rights reserved.
116 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126

1. Introduction fatigue design formula. In 1971, a design formula of static behavior of

Steel–concrete composite beams have been widely used in the field
of bridge and building structures for decades [1]. As an important com-
ponent of steel–concrete beams, stud shear connectors transfer the lon-
gitudinal shear force on the surface between steel and concrete. The
deformation ability of shear stud is a decisive factor to evaluate its duc-
tility. Hence, it is of great importance to find a method to improve the
deformation capacity of shear studs without sacrificing its bearing
capacity. To this end, elastic concrete is introduced into composite
beams in this paper.
Elastic concrete, also called crumb rubber concrete or recycled tire
rubber-filled concrete, is a new environmental material in the last few
years. According to previous studies, the deformability of concrete in-
creases significantly after adding rubber crumb to it. Elastic concrete
has ductile failure and better crack resistance. Moreover, it has superior
acoustical properties, wearing resistance and aging resistance to ordi-
nary concrete [2–4]. Nowadays elastic concrete is used in the paving
of tennis courts and parking lots due to its superior properties, but it
has never been used in composite beams. Hence, it is significant to
study the performance of the elastic concrete and steel composite
beams.
This paper presents an investigation on the shear studs in the elastic
concrete and steel composite beams through push-out test. The push-
out test has proven to be an effective method of determining the ulti-
mate strength and deformation capacity of shear connectors. Since the
first push-out test was devised in Switzerland in the 1930s [5], many
experimental tests of shear connectors have been studied by numerous
researchers. Viest [6] conducted 12 push-out tests and proposed the
conception of “critical load”. He suggested that the shear bearing capac-
ity of studs should be the load value when load-slip curves just entered
nonlinear stage or the residual slip was 0.762 mm. Chapman [7] and
Johnson [8] measured the shear performance of studs and developed a
calculation model based on push-out tests. In 1966, Slutter and Fisher
[9] tested the fatigue behavior of shear connectors and proposed the
Table 1
Proportions of elastic concrete with different rubber contents.
shear studs was proposed by Ollgaard [10] based on the push-out test
results. The structure of the formula was adopted by most of national
codes for nominal strength of studs. Changsha Railway Institute [11]
conducted 15 push-out tests of studs based on Wuhu Bridge in China.
An [12] investigated the different behaviors of studs between normal
strength and high strength concrete through push-out tests. The results
showed that the concrete compressive strength significantly affected
the shear capacity of studs.
Recently a number of researchers have focused on the different as-
pects of studs. In 2004, Lee and Shim [13,14] investigated the static
and fatigue behavior of large stud shear connectors up to 30 mm in di-
ameter, which were beyond the limitation of current design codes. A
new stud system fastened with high strength pins was investigated ex-
perimentally by Tahir [15]. Pavlović [16] studied the different behaviors
between bolted shear connectors and headed studs. W Xue [17] con-
ducted 18 push-out tests to show the static behavior of single-stud.
D Xue [18] investigated the different behaviors between single-stud
and multi-stud connectors. According to the aforementioned research,
the shear bearing capacity of studs depends on many factors, including
the material and diameter of the stud itself and properties of the sur-
rounding concrete slab. These factors are all included in several national
codes [19–22].
People pay more attention to the bearing capacity of the shear studs,
however few researchers have addressed the relationship between the
deformation capacity of shear studs and the properties of the surround-
ing concrete. This paper will focus on this point. Moreover, the optimum
efficiency of rubber content can be found after the experimental study.
The remainder of the paper is organized as follows. In Section 2, we
present an introduction to the material properties of elastic concrete
used in push-out tests. Section 3 details a total of 18 push-out tests an-
alyzing failure modes, bearing capacity and ultimate deformation of
studs. Three influence factors on the static behavior of studs, including
rubber contents, elastic concrete compressive strengths and stud diam-
eters are discussed in Section 4. The actual stress mechanism of studs in
push-out tests and the comparison between the test results and three
national codes regarding nominal strength of studs are also presented
in this section. Section 5 presents some concluding remarks of this
current paper.

Group Rubber Crumb Cement Stone Sand Water Water

content rubber (kg) (kg) (kg) (kg) reducing 2. Material properties of elastic concrete
(kg) (kg)

1 0% 0 295 1087 839 165 2.174 Before push-out tests, the material properties of the different rubber
2 5% 50 400 703 1004 169 2.391 mixed elastic concrete were studied experimentally. Note that the
3 10% 100 590 1230 412 168 6.522 mixture ratios of elastic concrete discussed below were the same
4 15% 150 590 1230 412 168 7.39
of those used in push-out specimens, introduced in Section 3 of this
5 5%(S) 50 550 703 1004 169 5.652
paper.

Fig. 1. Raw materials and concrete test cubes.

 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126 117

Table 2 insure the high fluidity of concrete mixing. The mix proportions of
Results of the compressive strength and elastic modulus tests. crumb rubber followed the principle of volume percentage method
Group Rubber 7d 28 120 Elastic and the rubber content was divided into four groups: 0% (the same
content (MPa) d/(MPa) d/(MPa) Modula/(GPa) as ordinary concrete), 5% (50 kg/m3), 10% (100 kg/m3) and 15%
1 0% 26.41 43.30 45.07 33.72 (150 kg/m3) [23]. These four groups were designed to be the same
2 5% 25.75 36.27 38.67 27.90 concrete grade level, C30. The compressive strength of C30 is between
3 10% 38.00 43.70 46.30 21.83 35–45 MPa. In addition, a higher grade elastic concrete (C40) with 5%
4 15% 28.85 35.13 38.46 14.45
rubber content was designed to study the influence of different com-
5 5%(S) 35.98 48.40 50.83 29.12
pressive strength on the performance of elastic concrete, and here this
group was called 5% (S).
The optimal mix proportions of the raw material were achieved after
2.1. Raw materials 37 group testing, and the results are listed in Table 1. Eighteen standard
concrete test cubes (150 × 150 × 150 mm3) and nine prismatic concrete
The raw materials used for test samples were fine aggregate, coarse samples (100 × 100 × 300 mm3) were made together with the 18 spec-
aggregate, water and crumb rubber within 1–2 mm in diameters. In imens for push-out test according to Table 1. The specimens were cured
addition, the high range water-reducing admixture was adopted to in the standard curing room (temperature is 20 ± 3 °C and relative

(a) (b)

(c) (d) (e)

Fig. 2. Typical damage shapes of concrete cubes. (a) Damaged concrete cube with 0% rubber content; (b) Damaged concrete cube with 5% rubber content; (c) Damaged concrete cube with
10% rubber content; (d) Damaged concrete cube with 15% rubber content; (e) Damaged concrete cube with 5% (S) rubber content.

Fig. 3. Compressive stress–strain curve testing.

118 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126

Table 3
The ultimate stress and strain of concrete with different rubber contents.

Rubber Ultimate load Ultimate Peak Ultimate

content (kN) strength (MPa) strain strain

0% 34.063 34.063 0.002725 0.003516

5% 26.833 26.833 0.002996 0.004223
10% 34.091 34.091 0.004323 0.006943
15% 29.412 29.412 0.004472 0.007104
5%(S) 37.762 37.762 0.003056 0.004024

humidity is above 90%) for the first 28 days and then under the condi-
tion of room temperature. The raw materials and concrete test cubes
are shown in Fig. 1.

2.2. Compressive strengths and elastic modulus of elastic concrete

According to the test methods of building material properties

(GB/T50081-2002) [24] and (T0555-2005) [25], the compressive Fig. 5. Stress–strain curves of concrete with different rubber contents.
strength and elastic modulus of concrete were obtained. The results
are listed in Table 2. Here, d refers to days and the push-out tests
were conducted on 120 d. The damaged concrete cubes after the
There was one fatal penetrating crack on the surface when the crumb
compressive strength tests are shown in Fig. 2.
rubber content was 0%, and the top of the sample was badly damaged.
From Table 2, it can be concluded that with the increasing rubber
When it had 5% rubber mixed in the concrete, there were multiple
content, the elastic modulus of concrete declines. The largest drop am-
cracks rather than just one fatal penetrating crack, as shown in
plitude of elastic modulus is 57% for elastic concrete with 15% rubber
Fig. 4(b). When the rubber content reached 10%, the concrete sample
content. Elastic modulus is a reaction of a material's deformability.
maintained integrity with four relatively long cracks, uniformly distrib-
Hence, with the increase of rubber content, it becomes easier to deform
uted on the surface. The condition of the concrete sample was even bet-
for the concrete in the elastic stage.
ter when the rubber content was 15%.

2.3. Compressive stress–strain curves of elastic concrete

A testing machine with a capacity of 5000 kN was adopted to apply
load on the prismatic concrete samples in stress–strain curve testing. On
the two surfaces of the concrete sample, the measuring points of longi-
tudinal strain were arranged at the center of the trisection. The experi-
mental details are shown in Fig. 3.
During the loading process, the tiny crack first appeared at the top of
all specimens, and the crack propagated downwards gradually. The
more rubbers mixed in concrete samples, the slower the tiny crack ap-
peared and developed. When reaching bearing capacity, fragments
from the ordinary concrete instantaneously ruptured, and the rupture
was accompanied by a loud noise. Then the load quickly dropped. How-
ever, there were less ruptured fragments for the elastic concrete, and
the process of damage and unloading took more time. The details can
be found in Table 3, and here the ultimate strain was defined as the
strain value when the ultimate load dropped 15%.
As shown in Fig. 4, the typical damaged situation of prismatic
concrete samples with different rubber contents very significantly.
2.4. Discussion
Compared with the ordinary concrete, the increasing ranges of peak
strain in the elastic concrete with 5%, 10% and 15% rubber content are
10%, 60% and 65% respectively, and ultimate strain are 26%, 97% and
102%. Moreover, the descent stage of the stress–strain curve becomes
more gradual. Thus, the elastic concrete has superior plasticity and duc-
tility properties. The details can be found in Fig. 5.
Comparing the two groups of 5% rubber content elastic concrete
with different compressive strength levels, the ultimate strain of the
5% group was slightly larger than that of the 5% (S) one, shown in
Table 3 and Fig. 6.
According to the material properties test, with the increasing crumb
rubber content, the deformability and ductility of concrete improved
significantly. Elastic concrete can absorb energy more effectively when
the rubber proportion reaches 10% or more, and no fatal penetrating
crack appears when the damage happens. The reason for this is crumb
rubber can be regarded as a tiny plastic unit in the elastic concrete.

(a)0% (b)5% (c)10% (d)15% (e)5%(S)

Fig. 4. Typical damaged situations of prismatic concrete samples. (a)0%. (b)5%. (c)10%. (d)15%. (e)5%(S).
 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126 119

3. Push-out test

Push-out tests were divided into six groups, and each group had
three specimens. Table 4 shows the parameters of push-out specimens
and the size is shown in Fig. 7. The detailed specifications of each
component are listed below.

a) Steel beam: The rolled H-section steel beams were used with a size
of 200 mm × 200 mm × 8 mm × 12 mm. The material type was
Q235B, with yield strength of 235 MPa and ultimate strength of
400 MPa.
b) Concrete slab: The size of the concrete slab was 460 mm ×
400 mm × 160 mm. As described in Section 2, the slabs of concrete
were divided into four categories of crumb rubber content: 0%, 5%,
10% and 15%. There was also a group of 5% (S) with higher compres-
sive strength.
c) The type of studs was Grade 4.6, with ultimate tensile strength of
400 MPa and yield strength of 240 MPa. Two diameters, 16 mm
Fig. 6. Stress–strain curves of elastic concrete with different compressive strength. (M16) and 19 mm (M19), were used and the heights of the stud
were 90 mm and 110 mm, respectively.
d) The diameter of reinforced bar was 10 mm, and its yield strength was
335 MPa.
Table 4
The parameters of push-out specimens.

Specimens Concrete Rubber Size of studs Reinforcement

strength grade content
3.1. Test set-up
PS 1–3 C30 0% Diameter is 16 mm, HorizontalΦ10@100
PS 4–6 C30 5% 90 mm long Φ10@95
PS 7–9 C30 10% (1.5%) The push-out specimens were conducted in a hydraulic servo-
PS 10–12 C30 15% testing machine with a capacity of 1000 kN, shown in Fig. 8. The loading
PS 13–15 C40 5%(S) Vertical speed can be controlled efficiently and the pre-defined load value can be
PS 16–18 C30 5% Diameter is 19 mm, Φ10@110
created. The load procedure was divided into two steps, pre-loading and
110 mm long (1.45%)
loading. The static-dynamic strain indicator was used and its maximum
frequency was 200 Hz. The longitudinal slip between the concrete slab
and steel beam and the lift-up value of the concrete slab were measured
by electronic displacement meter with a precision of 1/1000 mm.
When external force is applied, the rubber can absorb energy due to its Fig. 8(b) shows the arrangement of the displacement measurement
strong deformability, and it can also disturb the propagation and break- on one side. The strain of studs at the roots of studs and a half of the
through of cracks. In this case, the integrity of the elastic concrete can be stud length were measured. Fig. 9 shows the process of protecting the
ensured and the fast fracture damage is postponed. strain gauges on studs.

Fig. 7. Push-out test specimen. (a) Front view. (b) Side view.
120 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126

Fig. 8. Loading device and push-out test specimen. (a) Front view. (b) Side view.

3.2. Test results
3.2.1. Modes of failure
Based on the results of eighteen push-out tests, all the samples were
stud shank failure, and only one specimen had damage at the bottom of
the concrete slab. The test results are summarized in Table 5. Fig. 10
presents the failure modes of concrete and shear studs.
Fig. 11 shows the failure of shear studs on the surfaces of steel and
concrete. When normal damage occurred, there was a certain slip on
the root of the studs, shown in Fig. 11(a), (d). The stud had plastic defor-
mation and the damage was ductile. The failure of PS-10 and PS-14 took
place in the weld zone, which didn't fall under normal damage conditions
and should be ignored when drawing the conclusion. There was signifi-
cant rust and welding porosity at the weld zone in PS-10, shown in
Fig. 11(b), (e). The incompletely filled weld at the bottom of the stud in
PS-14 is shown in Fig. 11(c), (f). Welding defects cut down the bearing
capacity directly. Hence, the welding quality should be guaranteed.
The strength of concrete with the reinforced bar is high enough to
bear the ultimate load in the push-out test, and the final failure mode
is mostly shank failure. In the test, the load was applied on the upper
surface of the steel beam and transferred to concrete slab through
studs. The shear behavior of concrete can be reflected by the failure
modes of concrete under the stud shank. Fig. 12 describes the crack pat-
terns and local damaged areas of different rubber mixed concrete. In or-
dinary concrete, the cracks are intensive and long. With the increasing
rubber content, the cracks became fewer and smaller, and the local
damaged areas became smaller. Thus, the specimens with elastic
concrete have superior local cracking resistance to ordinary concrete.
3.2.2. Bearing capacity and ultimate slip
Fig. 13 shows load-slip curves of 18 push-out specimens and each
group has relatively consistent results. The load-slip curve consists of
two parts, ascending and descending part. The ascending section
can be separated into elastic and plastic parts. In the elastic part, the

Fig. 9. Strain gauges of studs: (a) Pasting strain gauges (using Glue502); (b) (c) Protecting strain gauges (using Glue703 (b) and Glue 914 and bandage (c)).
 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126 121

Table 5 when rubber content increased to 15%. Moreover, when the load
Results of push-out tests. reached 0.7 times the ultimate load, the stiffness of studs had a step-
Specimens Ultimate Ultimate Failure modes down trend, shown in Fig. 14(b). Hence, the shear stud in elastic
load (kN) slip (mm) concrete with 10% rubber content had relatively high bearing capacity,
PS-1 156.000 6.145 Shank failure on one side better deformation capacity and better ductility.
PS-2 158.694 5.754 Shank failure on one side Based on the test results, a linear formula of ultimate slip is
and concrete slab damaged proposed, describing the slip relationship between ordinary concrete
PS-3 163.327 6.432 Shank failure on both sides
and elastic concrete:
PS-4 163.839 7.759 Shank failure on both sides
PS-5 153.515 7.799 Shank failure on one side
PS-6 153.363 4.950 Shank failure on both sides S ¼ ð4:2 m þ 1:031Þ S0 ð1Þ
PS-7 149.727 8.0582 Shank failure on both sides
PS-8 133.750 9.975 Shank failure on one side
PS-9 150.370 9.316 Shank failure on both sides
where S0 is the ultimate slip of ordinary concrete and m is the rubber
PS-10 70.790 22.123 Shank failure on both sides, content.
the right side damage at
the weld zone 4.1.2. Effect of elastic concrete compressive strengths
PS-11 139.186 11.16 Shank failure on one side
Fig. 15 shows the load-slip behavior of studs in elastic concrete with
PS-12 119.285 10.192 Shank failure on both sides
PS-13 159.490 5.147 Shank failure on both sides different compressive strengths and the deformed shapes of studs after
PS-14 119.604 1.616 Shank failure at the weld tests. With the increasing concrete strength, the ultimate strength of
zone on one side studs showed a minor increase, while the ultimate slip declined 42%.
PS-15 156.516 5.788 Shank failure on one side The stiffness of studs in C30 and C40 was almost identical in the elastic
PS-16 220.037 7.869 Shank failure on both sides
part, but the stiffness declined faster in C30 than C40 in the plastic part.
PS-17 193.300 7.986 Shank failure on one side
PS-18 205.325 9.224 Shank failure on one side Thus, the lower the compressive strength of the surrounding elastic
concrete, the better ductility the shear stud had.

4.1.3. Effect of stud diameters

load-slip curves showed an almost linear relationship. In the plastic
part, the slip increased rapidly and the stiffness reduced continuously.
Table 6 summarizes the average shear bearing capacity and ultimate
slip of studs for each group. Here the ultimate slip was defined as the
slip value when loading reached its peak.
4. Discussion
Load-slip behavior of studs with different diameters in elastic con-
crete is shown in Fig. 16. With the increasing stud diameter, the shear
bearing capacity of shear stud increased. The bearing capacity of speci-
men with the 19 mm stud was 1.31 times that of the 16 mm stud, and
the slip was 1.07 times. The stiffness of the two diameters of studs
was almost equivalent. Consequently, the diameter had little influence
on ductility and stiffness of the stud in elastic concrete.

4.1. Bearing capacity, stiffness and ductility of the shear stud in elastic
concrete
4.1.1. Effect of rubber contents
Fig. 14(a) shows load-slip behavior of studs in elastic concrete with
different rubber contents and the deformed shape of studs after tests.
The results showed that the static bearing capacities of the studs em-
bedded in elastic concrete with 5%, 10% and 15% rubber content de-
clined 1.53%, 5.89% and 19.69% respectively, while the ultimate slip of
studs increased 27.50%,49.26% and 66.12%, respectively. Consequently,
the ductility of the stud improved gradually with the increasing rubber
content. The stiffness of these first three groups was nearly identical in
elastic part as shown in Fig. 14(a), but the stiffness obviously declined
4.2. Stress mechanism of studs in push-out test
According to the measured data from 36 studs, the stud experienced
four processes during the load procedure. Fig. 17(a) shows the appear-
ance of the original state of the studs and the position of the measured
points. Here, “+” refers to the tensile force and “−” refers to the com-
pressive force. The yellow section shows the gap between stud and con-
crete, and the red section characterizes the segment of concrete under
compression. The measured strain of the studs in (b)–(e) is taken
from PS-2 when the load is 3 kN, 90 kN, 110 kN and 130 kN, respectively.
The unit of the strain is 10−6.
In the first process, the four different measured points all experi-
enced tensile strain. In the initial stage of loading, there was a minor

Fig. 10. Modes of failure: (a) Failure of concrete slab; (b)–(d) Failure of shear stud.
122 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126

Fig. 11. Failure of studs on the surfaces of steel and concrete: (a)–(c) Failure of studs on the steel surface; (d)–(f) Failure of studs on the concrete surface. Note: Normal damage situation
(PS-8); Welding defect (PS-10 & PS-14).

gap between the stud and the concrete in which stud was embedded in.
Consequently, the stud deformed with the steel beam to a minor extent
under load between 0 and 5 kN, shown in Fig. 17(b).
In the second stage, the four strain gauges changed into “−++−”,
as shown in Fig. 17(c). Welded on the steel beam and embedded in con-
crete, the shear stud was equivalent to a beam fixed on both ends, and
the point of inflection appeared in the middle of the studs.
With the increasing load, the strain data changed into “+++−”in
the third process, as shown in Fig. 17(d). The root of the stud bore the
tensile force from steel beam. Hence, plastic strain firstly appeared on
the root of the stud.
In the last stage, the strain data all changed into “+” again, as shown
in Fig. 17(e), demonstrating that the shear stud was almost destroyed.
The load pulled the middle of the stud, while the head of the shear
stud was still anchored in the concrete. Finally, ductile failure occurred
at the root of shear stud.
The specific strain curves of PS-2 are shown in Fig. 18(a). According
to the measured strain, the third process started from 106 kN for PS-2
(specimen with ordinary concrete). The third process of stud in elastic
concrete with 10% rubber content was from 102 kN, shown in
Fig. 18(b). In this process, the strain of the root of the shear stud was
positive, meaning this cross section was under tensile stress. Collecting
all the other measured strain results, we found that the beginning of the
third process of each specimen had approximately two-thirds of the
bearing capacity. There is little difference between elastic concrete and
ordinary concrete at this moment. From this point, the load-slip curve
tends to gradual and the shear stiffness starts to fall noticeably.
While at the end of the third process, the corresponding load varies
greatly, 115 kN and 124 kN respectively, for ordinary concrete and elas-
tic concrete. The middle section of the stud bore the bending moment
and tension. Ultimately, it was the bending rather than the tension
that led to the slip of the stud. The bending stress made one strain

Fig. 12. Crack patterns and local damaged areas of different rubber mixed concrete. (a)0%. (b)5%. (c)10%. (d)15%.
 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126 123

Fig. 13. Load-slip curves of 18 push-out test specimens.

negative and the other one positive in the middle section of the stud, 4.3. Comparison between design codes and test results on ultimate strength
while the tension made both of values positive. Thus, the difference be- of studs
tween 4# and 3# strain value characterizes the size of the bending mo-
ment. From Fig. 18 we can see that the strain values of studs in 10% The experimentally obtained bearing capacities of studs were com-
elastic concrete are nearly twice as large as those in ordinary concrete. pared with those calculated from equations provided by Eurocode-4,
Hence, the deformed shape of the middle section of studs embedded AASHTO LRFD, and GB50017-2003 [17–19]. All the nominal shear resis-
in elastic concrete is larger. Moreover, with the nearly equal bearing ca- tances of studs in these three codes are determined separately by the
pacity, during the whole process the strain value of the stud in elastic concrete or the studs.
concrete is larger. Hence, it leads to a larger slip of studs in elastic con- In Eurocode-4, the bearing capacity of shear studs is calculated as the
crete than ordinary concrete. minimum of Eq. (2), which presents “failure of the concrete” and “fail-
ure of the stud”, respectively, from left to right:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  
2 2
P Rd ¼ 0:29αd Ecm f ck =γ v or P Rd ¼ 0:8f u πd =4 =γ v ð2Þ

Table 6
Ultimate shear strength and ultimate slip of push-out tests. where d is the diameter of the studs; fu is the ultimate strength of studs;
fck is the compressive strength of concrete; Ecm is the elastic modulus of
Specimens Rubber Concrete Stud's Average Average
content strength diameter ultimate load ultimate slip
concrete;γv is a partial safety factor (=1.25). The units are N and mm.
(%) (mm) (kN) (mm) In AASHTO LRFD, the nominal shear resistance of a single stud em-
bedded in a concrete deck shall be taken as Eq. (3).
PS 1–3 0 C30 16 159.34 6.11
PS 4–6 5 C30 16 156.91 7.79 qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
PS 7–9 10 C30 16 149.95 9.12 Q r ¼ ϕ0:5Asc Ecm f ck ≤ ϕAsc f u ð3Þ
PS 10–12 15 C30 16 127.97 10.15
PS 13–15 5 C40 16 158.00 5.47
PS 16–18 5 C30 19 206.22 8.36 where Asc is the cross-sectional area of the stud (mm2) and ϕ is a
resistance factor for the shear connector (=0.85).
124 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126

Fig. 16. Load-slip behavior of studs in elastic concrete with different stud diameters.

where Pu is the bearing capacity of the stud (kN); As is the cross-

sectional area of the stud (mm2); fc and Ec are the compressive strength
and elasticity modulus of concrete (MPa); f is the ultimate tensile
strength of the stud; γ is the ratio of the minimum tensile strength to
the yield strength of the stud.
The formulas from the AASHTO LRFD and the GB50017-2003 ex-
press the same meaning as Eurocode-4 with respect to the nominal
strength of studs. The left of the inequality is for “failure of the concrete”
and the right side signifies “failure of the stud”, meaning the control fac-
tor is concrete or stud, respectively. If the concrete is strong enough, the
result of the left side of the inequality can be larger than the right side. In
this case, we should use the result from the right side of the formula as
nominal strength, meaning the control factor is the stud not the con-
crete. In this paper, when the rubber content was 15%, the control factor
was the concrete in AASHTO LRFD and Eurocode-4, as the elastic modu-
Fig. 14. Load-slip behavior of studs in different rubber mixed concrete. (a) Shear resistance lus and compressive strength are both very low. The studs in the other
and deformed shapes of studs; (b) Shear stiffness of studs. three crumb rubber mixed concrete slabs should use the stud control.
Fig. 19 shows the comparison between test results and three nation-
In the GB50017-2003, the bearing capacity of the shear stud is al codes. The push-out test results were higher than the nominal shear
defined as follows: strengths in the three current national codes. The nominal shear
strength of stud in AASHTO LRFD was the closest to the test results,
qffiffiffiffiffiffiffiffiffi and the Eurocode-4 and GB50017-2003 gave relatively conservative
P u ¼ 0:43As Ec f c ≤ 0:7As γ f ð4Þ values.

5. Conclusions

Elastic concrete was first introduced into steel-concrete composite

beams to investigate its effect on the static behavior of shear studs in
the current paper. After material properties testing of different rubber
mixed concrete, eighteen standard push-out tests were conducted.
Different crumb rubber contents, elastic concrete compressive strengths
and stud diameters were taken into consideration. The following
conclusions have been drawn:

(1) The more crumb rubber mixed in, the more integrity of the block
will have and cracks will be smaller and fewer in material prop-
erties tests. More importantly, the deformation capability and
ductility improve dramatically for elastic concrete, especially
when the crumb rubber content is equal to or more than 10%.
Moreover, with the same rubber content, the ultimate strain of
C30 (5%) is slightly larger than C40 (5% (S)).
(2) Stud shank failure is the main failure mode in the push-out tests,
and the damage is ductile. According to the crack patterns and
Fig. 15. Load-slip behavior of studs in elastic concrete with different compressive concrete local damage area under the stud shank, elastic concrete
strengths. is proven to have superior local cracking resistance to ordinary
 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126 125

Fig. 17. Four processes of studs' deformed shape during load procedure. (a) The original state; (b) The first process; (c) The second process; (d) The thrid process; (e) The fourth process.

ε ε

Fig. 18. The measured strain of the studs in ordinary concrete and elastic concrete. (a) Ordinary concrete. (b) Elastic concrete with 10% rubber content.
126 Q. Han et al. / Journal of Constructional Steel Research 113 (2015) 115–126
Fig. 19. Comparison between the design codes and test results.
concrete. Only two specimens damage in weld zone. Welding
defects cutting down the bearing capacity tremendously.
Hence, avoiding the welding defect is of great importance in a
real project.
(3) Load-slip curve of the shear stud contains an elastic part, plastic
part and descending part. In the elastic part, the load-slip curve
shows an almost linear relationship. In the plastic part, the slip
increases rapidly, while the stiffness reduces continuously. Not
all of the maximum slips of studs in the tests have exceeded
6 mm, especially those in C40 concrete. 6 mm is the Eurocode-
4 requirement for ideal plastic behavior value of the shear con-
nection. Hence, it is significant that the compressive stress of
concrete in composite beams should not be too large so as to en-
sure the ductile behavior of studs.
(4) The deformation capacity and ductility of studs improve signifi-
cantly with the increasing rubber content, but the stiffness de-
clines rapidly when rubber content is 15%. Hence, the shear
stud in elastic concrete with 10% rubber content has relatively
high bearing capacity, better deformation and better ductility.
We also draw the conclusion that the lower the compressive
strength of the surrounding elastic concrete, the better ductility
the stud has. Moreover, the diameter has little influence on duc-
tility and stiffness of the stud in elastic concrete.
(5) Most times during the push-out tests, the shear stud is equivalent
to a beam fixed on concrete and steel beam, and it goes through
four processes during the load procedure. Plastic strain first ap-
pears on the root of the stud and then ductile failure occurs.
(6) The AASHTO LRFD obtains the closest results to the test ones, while
the Eurocode-4 and GB50017-2003 are relatively conservative.
Acknowledgments
The authors would like to thank the National Natural Science Foun-
dation of China (51178307 & 51408408) and the Tianjin Natural Science
Foundation (11JCZDJC24000) for providing the funding to carry out this
experimental works. Special thanks to Bin Liu, Guang Yang, Xuezhe
Zhang, Sokun Yang and all the technicians at Tianjin University who
have contributed to the work in this project.
References
[1] J.G. Nie, Steel-concrete Composite Beams, China Science Press, Beijing, 2005.
[2] N.N. Eldin, A.B. Senouci, Rubber-tired particles as concrete aggregate, J. Mater. Civ.
Eng. 5 (4) (1993) 478–496.
[3] B.I. Topçu, The properties of rubberized concrete, Cem. Concr. Res. 25 (1995)
304–310.
[4] F.H. Olivares, G. Barluengaalet, Static and dynamic behavior of recycled tire rubber-
filled concrete, Cem. Concr. Res. 32 (2002) 1587–1596.
[5] L. Dennis, Capacities of headed stud shear connectors in composite steel beams with
precast hollowcore slabs, J. Constr. Steel Res. 63 (2007) 1160–1174.
[6] I.M. Viest, Investigation of stud shear connectors for composite concrete and steel T-
beams, J. Am. Concr. Inst. 27 (8) (1956) 875–891.
[7] J.C. Chapman, S. Balakrishman, Experiments on composite beam, Struct. Eng. 42 (11)
(1964) 369–383.
[8] R.P. Johnson, R.D. Greenwood, K. Van Dalen, Stud shear connectors in hogging mo-
ment regions of composite beams, Struct. Eng. 47 (9) (1969) 345–350.
[9] R.G. Slutter, J.W. Fisher, Fatigue strength of shear connectors, Lehigh Univ. Inst. Res.
147 (1966) 65–88.
[10] Ollgaard J, Slutter R G, Fisher J W. The Strength of stud shear connection in light-
weight and normal weight concrete. Engineering Journal, AISC1971; 8(2):55–64.
[11] J.G. Nie, Y.H. Wang, Research status on fatigue behavior of steel–concrete composite
beams, Eng. Mech. 29 (6) (2012) 1–11.
[12] L. An, K. Cederwall, Push-out tests on studs in high strength and normal strength
concrete, J. Constr. Steel Res. 36 (1) (1996) 15–29.
[13] C.S. Shim, P.G. Lee, T.Y. Yoon, Static behavior of large stud shear connectors, Eng.
Struct. 26 (12) (2004) 1853–1860.
[14] P.G. Lee, C.S. Shim, S.P. Chang, Static and fatigue behavior of large stud shear connec-
tors for steel-concrete composite bridges, J. Constr. Steel Res. 61 (9) (2005)
1270–1285.
[15] M.T. Mahmood, N.S. Poi, S.T. Cher, Push-off tests on pin-connected shear studs with
composite steel–concrete beams, Constr. Build. Mater. 23 (2009) 3024–3033.
[16] P. Marko, M. Zlatko, V. Milan, Bolted shear connectors vs. headed studs behavior in
push-out tests, J. Constr. Steel Res. 88 (2013) 134–149.
[17] W.C. Xue, M. Ding, H. Wang, Z.W. Luo, Experimental studies on behavior of stud
shear connectors under monotonic loads, J. Build. Struct. 30 (1) (2009) 95–100.
[18] D.Y. Xue, Y.Q. Liu, Z. Yu, J. He, Static behavior of multi-stud shear connectors for
steel–concrete composite bridge, J. Constr. Steel Res. 74 (2012) 1–7.
[19] BS5400, Steel, Concrete and Composite Bridges, Part 3: Code of Practice for Design of
Steel Bridges, British standard institution, London, 1978.
[20] ENV 1994–2, Eurocode-4: Design of Composite Steel and Concrete Structures, Part
2: Composite bridges, CEN-Eurocode Committee for Standardization, 1997.
[21] AASHTO LRFD, Bridge Design Specifications, 3rd ed. American association of state
highway and transportation officials, Washington, D.C, 2004.
[22] Ministry of construction of China, GB50017-2003, Code for Design of Steel Struc-
tures, China Planning Press, Beijing, 2003.
[23] L.H. Yang, H. Zhu, Strengths and flexural strain of CRC specimens at low tempera-
ture, Constr. Build. Mater. 250–253 (2010) 906–910.
[24] Chinese industrial standard of building material, GB/T50081-2002, Test methods of
mechanics property of ordinary concrete2002.
[25] Chinese industrial standard of building material, T0555-2005, Test Method of Pris-
matic Compressive Strength of Cement Concrete2005.