Journal of Constructional Steel Research 108 (2015) 2330

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

Headed stud shear connector for thin ultrahigh-performance concrete

bridge deck
Jee-Sang Kim a, Jongwon Kwark b, Changbin Joh b, Sung-Won Yoo c, Kyoung-Chan Lee d,
a

Seokyeong University, 16-1 Jungneung-Dong, Sungbuk-gu, Seoul 136-704, Korea

Korea Institute of Civil Engineering and Building Technology, 283 Goyangdae-ro, Ilsangseo-gu, Goyang-si, Gyeonggi-do 411-712, Korea
Woosuk University 66 Daehak-ro, Jincheon-eup, Chungcheongbuk-do 365-803, Korea
d
Korea Railroad Research Institute, 176 Cheoldobangmulgwan-ro, Uiwang-si, Gyeonggi-do 437-757, Korea
b
c

a r t i c l e

i n f o

Article history:
Received 4 November 2014
Accepted 4 February 2015
Available online xxxx
Keywords:
Headed stud
Shear connector
Ultrahigh-performance concrete
Composite beam
Bridge slab deck

a b s t r a c t
Ultrahigh-performance concrete (UHPC) provides much higher compressive and tensile strength than
conventional concrete. UHPC is advantageous for use as a bridge slab deck owing to its higher strength, stiffness,
and durability. One drawback, however, is the fact that the joint region connecting the deck and girder generally
has a thicker cross-section to ensure proper shear connection, which hinders making the overall UHPC deck
thinner and lighter. In addition, the shear strength of stud shear connectors embedded in UHPC slab has not
been veried to be the same as that in a conventional concrete deck. This study investigates a stud shear
connector embedded in a UHPC deck through 15 push-out tests. The ultimate strength of the stud and relative
slips are measured. The test parameters were chosen to prove the feasibility of a thinner slab. The stud aspect
ratio, overall height-to-diameter, and cover thickness on top of the stud head requirement are also examined
to verify the existing geometrical constraints specied in the AASHTO LRFD and Eurocode-4 design codes for
UHPC decks. It was shown that the aspect ratio can be reduced from 4 to 3.1 without loss of shear strength of
the stud, and the cover could be reduced from 50 mm to 25 mm without causing a splitting crack at the UHPC
slab. However, the required ductility demand, 6 mm, was not realized in all cases. Therefore, the stud shear
connectors in a UHPC deck should be designed according to the elastic criterion.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Ultrahigh-performance concrete (UHPC) is an advanced composite
material consisting of a high-strength matrix and bers. It offers significantly superior compressive (N150 MPa) and tensile strength (N5 MPa)
compared to conventional concrete, as well as higher modulus of
elasticity (N40 GPa) [1]. It is typically made from a mixture of Portland
cement, silica fume, ller, ne aggregate, high-range water reducer,
water, and steel bers.
UHPC is being increasingly used worldwide in various components
of civil infrastructure. In particular, many studies have investigated its
application to bridge components such as girders, decks, and connection
joints owing to its higher strength, stiffness, and durability. Many
studies have investigated the use of UHPC as a deck slab component.
Saleem et al. [2,3] developed a low-prole UHPC deck system as an
alternative to an open-grid steel deck. Coreslab Structures Inc. developed a wafe-shaped UHPC panel that was installed in a bridge in Little
Cedar Creek, Wapello County, Iowa, US [4], and Aaleti and Sritharan

Corresponding author. Tel.: +82 31 460 5391; fax: +82 31 460 5364.
E-mail addresses: zskim@skuniv.ac.kr (J.-S. Kim), origilon@kict.re.kr (J. Kwark),
cjoh@kict.re.kr (C. Joh), imysw@woosuk.ac.kr (S.-W. Yoo), eclip77@gmail.com (K.-C. Lee).

http://dx.doi.org/10.1016/j.jcsr.2015.02.001
0143-974X/ 2015 Elsevier Ltd. All rights reserved.

[57] investigated the structural behavior and proposed a design

guide for this panel system, including connections.
Efforts have also been made to develop a hybrid beam that
comprises an FRP girder strengthened with a layer of UHPC slab on
top. Chen and El-Hacha [810] used 9.5-mm-diameter GFRP studs to
join the hollow-box FRP girder and a 53-mm-thick UHPC layer on top.
Nguyen et al. [11,12] developed a hybrid composite beam comprising
an FRP I-girder topped with a precast UHPC slab, which uses M16
bolts as shear connectors with an epoxy bonding. The UHPC slab was
50 mm thick, and the bolt was embedded to a depth of 35 mm, resulting
in only 15 mm of cover on top of the bolt head and stud height-todiameter aspect ratio of 2.2. This cover thickness and aspect ratio do
not satisfy the values of 50 mm and 4, respectively, specied in existing
design codes.
A UHPC bridge deck can feasibly have a thinner cross-section than a
conventional concrete deck, as shown in previous studies [512]. However, the joint region connecting the deck and the steel girder should
have thickness comparable to that in the conventional case to ensure
that shear connectors can be properly installed and embedded in the
deck in order to conform to existing design codes. For example, two previously developed UHPC deck systems have joint connections with
thicknesses of 127 mm (5 in.) [2,3] and 203 mm (8 in.) [4,5], which
are no less than that of a conventional concrete deck. Because the

24

J.-S. Kim et al. / Journal of Constructional Steel Research 108 (2015) 2330

thinnest sections of the UHPC deck are 32 mm (1.25 in.) [2,3] and
63.5 mm (2.5 in.) [47], a shear connection requires a signicantly
thick UHPC deck; this goes against the design objective of reducing
the self-weight and lowering the prole of the deck. This study investigates the structural behavior of stud shear connectors embedded in
UHPC decks of various thicknesses and conrms the validity of existing
design codes for this application.
Since 1960, composite structures have been widely used owing to
their structural efciency. They typically consist of a steel girder and
concrete deck that transfers shear force through suitable shear connectors such as angles, channel sections, headed studs, and perforated ribs.
Headed studs are used most commonly owing to their simple and quick
installation using a stud-welding gun and superior ductility than other
types of shear connectors.
The static strength of stud shear connectors was originally evaluated
based on Ollgaard et al.s [13] early experimental work. They showed
that the static strength of a stud shear connector is controlled by two
different failure mechanisms: surrounding concrete crushing failure, re0

lated to concretes compressive strength, f c , and shearing failure of the

shank of the stud, related to the studs ultimate tensile strength, Fu. The
smaller value between the two different mechanisms controls the
design shear strengths of a stud shear connector. The AASHTO LRFD provision 6.10.10.4.3 [14] denes the design static strength of a stud shear
connector, Qr, as
Q r sc Q n sc 0:5Asc

q
0
f c Ec

sc F u Asc

where the resistance factor, sc, is taken as 0.85. Eurocode-4 [15] denes
the design static shear strength, PRd, as
2

P Rd

0:29d

q
0
f c Ec

0:8 F u Asc
v

where the partial factor, v, is taken as 1.25, and an aspect ratio factor, ,
which depends upon the stud height-to-diameter ratio, hsc/d, is taken as
0.2(hsc/d + 1) for 3 hsc/d 4 and 1 for hsc/d 4.
Different design codes have different resistance or partial factors. However, they are similar in that the left-hand side terms of Eqs. (1) and (2)
refer to concrete crushing failure in terms of the surrounding concrete
strength (f'c) and modulus of elasticity (Ec) but not the mechanical property of the embedded stud. Furthermore, the right-hand side terms of
Eqs. (1) and (2) refer to stud shank failure in terms of the ultimate tensile strength (Fu) of the stud but not the mechanical property of the surrounding concrete. The concrete crushing failure controls when the

Table 1
Push-out test specimens.
Specimen
group

Deck
thickness
(mm)

Stud shear connector

Height
(mm)

Diameter
(mm)

Aspect ratio
(height/diameter)

Normal
UHPC-I
UHPC-II
UHPC-III
UHPC-IV

150
150
100
100
75

100
100
65
50
50

22
22
16
16
16

4.5
4.5
4.1
3.1
3.1

Cover
thickness
(mm)

EA

50
50
35
50
25

3
3
3
3
3

compressive strength of the concrete is low or moderate, and the stud

shank failure does when the strength is high. The threshold between
the two failure modes usually lies at a concrete compressive strength
of 3040 MPa.
Considering that the compressive strength of UHPC exceeds
150 MPa, the stud shank failure mode obviously always controls the
static strength of the stud shear connector. Ollgaard et al. [13] reported
that the concrete strength of their specimens was 1835 MPa. Therefore, the validity of existing design codes for stud shear connectors
should be conrmed for UHPC applications because it provides much
higher concrete strength than before.
Geometrical constraints are another important issue with regard to
UHPC decks in that they must be as thin as possible to reduce their
weight and construction costs. The constraints of existing design codes
may result in a UHPC deck with a thicker cross-section at the joint region between the deck and the girder. The thickness of wafe deck
panels [47] is 63.5 mm at the thinnest region between ribs but
200 mm at the joint region. Saleem [2,3] developed a low-prole deck
system that is as thin as 31 mm between ribs but is 125-mm-thick at
the joints. This study investigates a joint region with a thickness of
only 75 mm to overcome the stocky joint region resulting when applying a current design code to the shear connectors embedded in a UHPC
deck.
The rst geometrical constraint is the aspect ratio between the overall stud height and the shank diameter. The AASHTO LRFD [14] and
Eurocode-4 [15] design codes require an aspect ratio of at least four
and three, respectively. The second constraint is the concrete cover
thickness over the stud head to prevent a longitudinal splitting crack
on top of the shear connector. The AASHTO LRFD provision 6.10.10.1.4
[14] regulates that the clear depth of the concrete cover over the top
of a shear connector should not be less than 50 mm and should penetrate at least 50 mm into the concrete deck. For example, when using
the most common diameter of 17 mm for a stud for a bridge deck and

Fig. 1. Push-out specimen dimensions.

J.-S. Kim et al. / Journal of Constructional Steel Research 108 (2015) 2330

25

2. Experimental program

Table 2
UHPC mixture.
w/b ratio

Cement

Silica
fume

Filler

Fine
aggregate

Water
reducer

Steel
ber

0.07

1.0

0.25

0.3

1.1

0.016

16.5 mm 1%
19.5 mm 1%

when following the AASHTO LRFD design code [13], the deck thickness
should be at least four times the diameter plus 50 mm cover, for a thickness of at least 118 mm. This is one reason why the composite joint is
chunky.
The Eurocode-4 provision 6.6.5.1 [15] regulates that the surface of a
connector should not extend less than 30 mm clear above the bottom
reinforcement, and provision 6.6.5.2 regulates that the cover should
not be less than that required for reinforcement adjacent to the same
surface of concrete. As the UHPC deck does not always enclose reinforcement, applying the cover thickness given in the Eurocode-4 design
code [15] is not possible.
The UHPC material provides much higher strength and durability
[1], and therefore, the deck thickness can be much less than when
using conventional concrete. However, this may not be the case at the
deck girder joint owing to the geometrical constraints for the embedded
shear connectors to ensure the transfer of the longitudinal shear force.
This study investigates the static strength and behavioral validity of
stud shear connectors for a thin UHPC solid slab deck.
Several factors limit the use of stud shear connectors in a thin UHPC
deck slab. The rst is concern over whether the stud shear connector
embedded in UHPC provides equivalent static strength compared to
that in conventional concrete. The second is concern related to the geometrical properties; the installation of stud shear connectors is subject
to geometrical constraints such as the height-to-diameter ratio and
cover thickness over the stud head, and existing design codes may
not, in fact, allow the use of stud shear connectors for thin deck slabs.
The last is concern over the ductility provided by stud shear connectors because the strength of the surrounding concrete is much higher
than that of conventional concrete such that the structural behavior
may differ from stud shear connectors embedded in conventional
concrete.

Shear connectors at a exural composite member resist the relative

slip occurring at the interface between the girder and the slab deck. The
best way to measure the static strength of the shear connectors is a exural beam test under a distributed load. However, to reduce the cost and
time, a direct push-out test is generally used instead. The experiment in
this study follows a standard test procedure given in the Eurocode-4-11 design code [16].
Five groups of specimensNormal and UHPC-I to UHPC-IVare prepared, as listed in Table 1, and three specimensA, B, and Care prepared for each group. The key variables of the test program are the
deck thickness and resulting stud aspect ratio and cover. The Normal
case has specimens with a conventional reinforced concrete slab for calibrating the test setup and for the purpose of comparison. The UHPC-I
case has identical dimensions to the Normal case and differs only in
that the deck is made of UHPC instead of conventional concrete. Both
the Normal and the UHPC-I specimens have the same thickness as the
conventional concrete deck, and the stud shear connector satises the
given geometrical constraints: aspect ratio of at least four and cover of
at least 50 mm. UHPC-II and UHPC-III specimens have 100-mm-thick
decks. The cover on top of the stud in UHPC-II specimens is 35 mm
thick, which is less than the requirement. UHPC-III specimens satisfy
the cover requirement; however, the aspect ratio is only 3.1, which is
less than the required value of four. UHPC-IV specimens have the thinnest slab, which is only 75 mm thick, and its cover thickness and aspect
ratio of 25 mm and 3.1, respectively, do not satisfy the requirements.
The specimens are prepared for a two-face push-out test. Four studs
are welded at each face, as shown in Fig. 1. These headed studs meet the
Type B requirements specied in AWS.D 1.1 [17], namely, minimum
yield strength of 350 MPa and minimum tensile strength of 450 MPa,
and they are welded on the ange using a conventional stud-welding
gun. In this study, studs with two different diameters are used:
22 mm for Normal and UHPC-I groups and 16 mm for the other groups,
as shown in Table 1. The stud diameters are chosen according to the
thickness of the deck to meet the requirement of aspect ratio of four;
Normal and UHPC-I groups have 150-mm-thick slabs and the other
groups have thinner slabs. The tensile and shear strength properties of
headed studs were tested from a direct tension and double-shear guillotine tests. The direct tension test xture used for this purpose was similar to that suggested in AWS D1.1-2000 [17]. A double-shear guillotine

Back
BL

BR

120mm
FL
FR

100mm

Front

Fig. 2. UHPC casting in horizontal position.

Transverse
Displacement

Fig. 3. Displacement measurement plan.

26

J.-S. Kim et al. / Journal of Constructional Steel Research 108 (2015) 2330

500

Stress [MPa]

400
300
200

16 mm

100

22mm
0
0

0.02

0.04

0.06

0.08

0.1

Strain
Fig. 5. Direct tension test result on bare stud specimens.

Fig. 4. Push-out test setup and lateral supporting bars.

test was conducted on the middle third of the stud shank to determine
the steel shear strength; this test was conducted using an apparatus
similar to that used by Anderson and Meinheit [18].
Table 2 provides details of the mixture of the UHPC. Steel bers of
two different lengths, 16.5 and 19.5 mm, are mixed together with 1%
volume each. The UHPC is designed for a characteristic compressive
strength of 180 MPa, the measured minimum strength was 200 MPa;
measured minimum tensile strength was 18 MPa, and the measured
modulus of elasticity was 4.5 105 MPa. The measured compressive
strength of the Normal specimen group was 35 MPa.
Steel sections used for simulating a steel girder have width, depth,
web thickness, and ange thickness of 300, 300, 10, and 15 mm, respectively. UHPC cannot easily be cast in the vertical direction, and the casting direction may affect the test result. Therefore, the steel section is cut
at the middle of the web in the longitudinal direction and the cast on top
of a ange to simulate the eld application casting direction, as shown
in Fig. 2. The specimens were steam-cured and the initial curing temperature was 40 C, and it was increased by 10 C every hour up to 90 C.
Steam-curing was done for three days and the temperature was gradually decreased at the end of the curing process. After curing, the pushout specimens were prepared with two separate faces bolted together
at the cut section of the web using an M24 high-tension bolt.
The prepared specimens were loaded using a 2000 kN universal test
machine. Assuming stud shank failure from Eq. (1), the failure load was
expected to be 1368 kN considering eight studs for each specimen. According to the Eurocode-4-1-1 design code [16], cyclic loads were applied to stabilize the specimen and break the bond between the steel
section and the deck. The cyclic load was 5%40% of the expected failure
load with loading speed of 0.82 kN/s. After cyclic loading, the specimens

Table 3
Stud tension and Guillotine test result.

were loaded constantly by increasing the displacement control at a

speed of 0.005 mm/s until failure.
The relative slips between the steel section and the slab deck are
measured using four LVDTs located 120 mm apart from the top of
each slab, as shown in Fig. 3. To ensure avoiding the separation of the
slab from the steel section, lateral supporting bars are installed at the
top and bottom of the specimen, and any possible separation is monitored using two LVDTs located outward of each slabs, as shown in Fig. 4.

3. Test results and discussions

3.1. Stud tensile and double shear test on bare stud specimens
Table 3 summarizes the measured tension and shear test results, in
which the resulting values for the shear test were divided by two to obtain resisting force for one shear surface. The tensile yield and ultimate
strength of each of these steels exceeds the AWS D1.1-2000 [17]
requirements of 350 and 450 MPa, respectively. Fig. 5 shows the
strainstress curve obtained from the tension test. Yielding behavior
was clearly observed for the tensile test. On the other hand, the guillotine test results do not show a clear yield or proportional limit, as
shown in Fig. 6. The typical failure of a stud loaded in double shear is
shown in Fig. 7.
The ratio of the measured shear strength to the measured tensile
strength is 0.80 and 0.82, and the ratio to the nominal tensile strength
(450 MPa) is 0.87 and 0.85, for the 16 mm and 22 mm studs, respectively. This value is larger than that of 0.65 obtained by Anderson and
Meinheit [18]. The higher ratio appears attributable to the fact that the
xture holding stud for double shear does not fully restrain the stud
specimen, which eventually involves additional bending other than direct shearing load. In reality, a shear stud experiences a combined action
of shearing and bending. It is difcult to estimate the exact contribution
of the two different mechanical behaviors. To evaluate shearing resistance only, the stud head should be fully restrained in the vertical as
well as the horizontal directions to eliminate the bending action.

3.2. Ultimate strength and initial stiffness of stud embedded in UHPC deck

Stud
diameter
[mm]

Tension test

Guillotine test
(per interface)

Yield
force
[kN]

Yield
stress
[MPa]

Ultimate
force [kN]

Ultimate
stress [MPa]

Ultimate
force [kN]

Ultimate
stress
[MPa]

16
22

77
141

384
372

97
177

484
466

78
145

390
380

The most important data obtained from the push-out test are the applied ultimate load at failure. The resulting ultimate failure load (Pmax)
and relative slips are analyzed by the procedure given in the
Eurocode-4-1-1 design code [16], which denes the characteristic resistance (PRk) as the minimum failure load is reduced by 10%. The slip capacity of a specimen (u) is taken as the maximum slip measured at the
characteristic load level. The characteristic slip capacity (uk) is taken as

J.-S. Kim et al. / Journal of Constructional Steel Research 108 (2015) 2330

27

500

16mm
22mm

Stress [MPa]

400
300
200
100
0

10

12

Displacement [mm]
Fig. 6. Guillotine double shear test result for bare stud specimens.

the minimum test value of u was reduced by 10%. Table 4 shows

detailed test results for each specimen.
Because the concrete failure mode never controls the shear connector embedded in UHPC, the ultimate strength of stud shear connectors is
obtained as AscFu according to the AASHTO LRFD design code [14] if the
effect of the resistance factor is excluded. Fig. 8 shows the curve of normalized applied forces to AscFu and relative slip. Considering the ultimate tensile strength, Fu, of the stud material (450 MPa) and the
shank diameter of the stud (22 mm) for UHPC-I, the tensile strength
of a UHPC-I specimen is expected to be at least 171 kN. For UHPC-II, III, and -IV specimens, the minimum tensile strength is expected to be
90 kN considering a stud diameter of 16 mm. The Eurocode-4 design
codes [15], given by Eq. (2), evaluate the ultimate strength to be 20%
smaller than that evaluated by the AASHTO LRFD design code [13]. In
Eq. (2), when excluding the effect of the partial factor, the expected tensile strengths are 137 kN for Normal and UHPC-I and 72 kN for others.
The ratios between the measured stud characteristic shear strength
(PRk) from the push-out tests and measured ultimate tensile strength
(Pu_test) from the direct tension test and those expected from the
AASHTO LRFD [14] and Eurocode-4 [15] design codes are listed in
Table 5. The shear strength of the stud in the Normal specimen group
is almost identical to that from the design code Eq. (2) of Eurocode-4
[15]. From this observation, it can be said that the test setup used in
the experimental program is reasonable.
First, it should be noted that the measured ultimate strength of a
stud shear connector does not show signicant difference in all cases
but the Normal one. This suggests that the slab thickness does not affect
the strength of the stud shear connector. In particular, despite the
Eurocode-4 design code [15], Eq. (2) species reduced static strength
of the stud for a small aspect ratio; however, this is not the case for a
UHPC slab deck. The UHPC-II and -IV cases, which do not satisfy the
cover thickness requirement specied in the AASHTO LRFD design
code [14], did not show any splitting crack.

The measured static strengths of the stud shear connector embedded in the UHPC are 2%13% higher than the nominal tensile strength
of the stud itself, and they correspond to the AASHTO LRFD design equation [14]. They are evaluated 27%42% more conservatively if the
Eurocode-4 design equation [15] is applied. Therefore, the shear
strength of stud shear connector in UHPC is adequate to be evaluated
in accordance with the AASHTO LRFD [14] rather than the Eurocode-4
[15] design code.
The initial stiffness of stud shear connectors is assumed innite according to the strength design concept. In reality, they show some initial
slip in the early loading stage owing to surrounding concrete cracking
and stud deforming. The initial stiffness is calculated from the relative
slip between 10% and 40% of the ultimate load, as shown in Table 6.
The average stiffness of Normal specimens is 336 kN/mm for single
studs, and UHPC-I specimens show the highest stiffness of 762 kN/mm.
UHPC-III specimens have a stiffness of 736 kN/mm, which is comparable
to that of UHPC-I specimens. UHPC-II and UHPC-IV specimens have
slightly smaller stiffnesses (598 and 538 kN/mm, respectively) than
UHPC-I and UHPC-III.
Oehlers and Coughlan [19] proposed an equation for estimating initial shear stiffness from 116 push-out tests:

0 
K si P max =d 0:160:0017 f c

where the initial stud stiffness (Ksi) is obtained from the shear stud
strength (Pmax), diameter of stud (d), and concrete compressive
strength (fc).

Table 4
Push-out test results for single stud.
Specimens
Normal

UHPC-I

UHPC-II

UHPC-III

UHPC-IV
Fig. 7. Typical failure of stud after guillotine double shear test.

A
B
C
A
B
C
A
B
C
A
B
C
A
B
C

Pmax [kN]

PRk [kN]

u [mm]

uk [mm]

158
148
145
198
193
212
123
120
114
105
103
111
109
109
117

130

15.68
13.56
10.89
7.66
5.73
7.18
4.98
4.02
4.21
4.84
5.93
5.64
5.42
5.04
5.18

9.80

174

103

92

98

5.16

3.62

4.36

4.54

J.-S. Kim et al. / Journal of Constructional Steel Research 108 (2015) 2330

Applied Force/Nominal Ultimate Force [P/AscFu]

28

1.4
1.2
1
0.8
UHPC-I-B

0.6

UHPC-II-C

0.4

UHPC-III-B
UHPC-IV-B

0.2
0

Relative Slip [mm]

Fig. 8. Curve of normalized applied force with respect to relative slip.

Applying Eq. (3) to the Normal specimen group provides an initial

stiffness of 68 kN/mm, which is signicantly less than the test results.
However, Shim et al. [20] presented initial stiffness results of large diameter (25 mm and 30 mm) studs from push-out tests, which varied
from about 200 to 400 kN/mm. It can be said that test results for the
Normal specimen group match those of Shim et al. [20]. The results in
this study indicate that a stud embedded in the UHPC provides at least
60% higher stiffness than that in conventional concrete. Based on the results for UHPC-II and UHPC-IV, the cover thickness may affect the initial
stud stiffness: a thicker cover generates higher stud stiffness.

3.3. Aspect ratio of stud

The aspect ratio is another important issue to be considered when
applying stud shear connectors to a thin deck slab. The AASHTO LRFD
[14] and Eurocode-4 [15] design codes require an aspect ratio of at
least four, although the latter allows an aspect ratio of three only if the
strength is reduced, as given in Eq. (2). A UHPC slab deck is advantageous in that the deck slab can be made as thin as possible. Applying a
stud shear connector to a thin slab naturally reduces the stud height,
and the stud diameter should also be smaller to satisfy the aspect ratio
requirement, which requires a greater number of studs and therefore
increased construction time. Considering that the strength of the surrounding concrete in the case of a UHPC deck is much stronger than
that in conventional applications, this study investigates the validity of
a smaller aspect ratio. Aspect ratios of 4.5, 4.1, 3.1, and 3.1 were tested
for each specimen groupUHPC-I to -IV.
The test results do not show any signicant difference with changes
in the aspect ratio. The stud shear connector shows obvious shearing behavior and not bending because of the higher stiffness and strength of
the surrounding concrete. Xu and Sugiura [21] showed that lower concrete strength might lead to relatively more obvious bending

deformation with shear deformation in the push direction. In Fig. 9,

the fractured cross-section of the stud shows a clean-cut immediately
above the welding area, which means that the shearing behavior is primarily responsible for stud fracture. Therefore, an aspect ratio as low as
3.1, as investigated in this test program, is allowable for the stud shear
connectors embedded in UHPC without a loss of strength.

3.4. Cover thickness on top of stud head

AASHTO LRFD [14] requires a minimum cover thickness of at least
50 mm (2 in.) over the stud head. This makes it difcult to realize a
thin UHPC slab deck, and it could be an overly conservative solution
considering the mechanical properties of UHPC. In this test program,
the cover of UHPC-II and -IV specimens is 35 and 25 mm, respectively.
The resulting static strength of UHPC-II and -IV specimens is 12% and
7% higher than that of UHPC-III specimens. Although these specimens
show a ductility issue, the same is also observed in UHPC-III specimens,
whose cover is 50 mm. This ductility issue is therefore considered to be
independent of the cover thickness. The test results show that specimens with shallow covers do not suffer any strength reduction, cracking, and spalling of the cover as seen at the surface of the slab deck.
There is some reduction in the initial stiffness owing to the shallow
cover thickness; however, it remains much larger than that of a conventional concrete deck. Therefore, the regulation of the minimum

Table 6
Stiffness of a stud shear connector.
Specimens
Normal

UHPC-I

Table 5
Test results compared to shear strength from the direct tension test and design codes.
Specimens

PRk/Pu_test

PRk/AASHTO [AscFu]

PRk/Eurocode [0.8AscFu]

Normal
UHPC-I
UHPC-II
UHPC-III
UHPC-IV

0.76
1.02
1.10
0.98
1.03

0.85
1.06
1.18
1.06
1.11

0.99
1.32
1.48
1.32
1.39

UHPC-II

UHPC-III

UHPC-IV

A
B
C
A
B
C
A
B
C
A
B
C
A
B
C

Stiffness [kN/mm]

Average stiffness [kN/mm]

309
379
322
754
714
816
532
571
689
1088
535
585
340
487
788

336

762

598

736

538

J.-S. Kim et al. / Journal of Constructional Steel Research 108 (2015) 2330

29

constant longitudinal pitch; instead, the elastic theory and the resulting
variable longitudinal pitches should be applied.
4. Conclusions
This study investigates the structural performance and validity of
stud shear connectors for a thin UHPC slab deck. The following conclusions can be derived from the test program:

Fig. 9. Fractured section of stud shear connector after failure.

thickness of the cover over the stud head can be relaxed to at least
25 mm for a UHPC slab deck.

3.5. Ductility of stud

Two strategies have been used for the static design of stud shear connectors: elastic and strength designs. The elastic design results in a variable pitch, which is narrower for the high shear zone usually at the end
of beams and wider for the low shear zone usually at the center of the
beam. On the other hand, the strength design concept assumes that all
studs maintain their ultimate strength until failure at the ultimate
state of the entire structure, which results in a constant pitch regardless
of the longitudinal location of the beam. Nowadays, most design codes
are based on the strength design concept. However, the ductility demand should always be guaranteed to ensure the strength design concept. Although the ductility demand can be variable for each structure,
the Eurocode-4 design code [16] requires the characteristic relative
slip (uk) to be at least 6 mm from the push-out test as a criterion for
the ductility demand.
The test results of this study showed that most test specimens, except for UHPC-I-A and -C, had unsatisfactory ductility, as listed in
Table 4. Hegger et al. [22] also reported the same conclusions, in
which the characteristic relative slip was 5.7 mm for a shear connector
embedded in high-strength concrete. Therefore, another measure is required to enhance the ductility of a stud shear connector embedded in a
UHPC. Otherwise, exact composite analysis can provide a more precise
ductility demand that may be less than 6 mm for a specic structure.
In such a case, a stud shear connector can be used for a UHPC slab
deck with a constant pitch.
In order to overcome the ductility issue, stud shear connectors in
UHPC can be designed based on the elastic theory instead of the plastic.
The elastic theory, which results in variable stud pitches, provides narrow spacing near the support for resisting large shear forces and
wider at the mid-span for smaller shear forces. In contrast, the plastic
theory, which results in a constant pitch across the entire span, is
based on the ductile behavior of the stud. The experiment program in
this study showed that shear studs in UHPC do not meet the ductility
demand, which does not allow the application of the plastic theory.
Therefore, unless the ductility problem is resolved, stud shear connectors in a UHPC should not be designed using the plastic theory with

1) The static strength of stud shear connectors embedded in a UHPC is

always controlled by steel failure. This indicates that strength is affected only by the stud diameter and the ultimate strength of the
stud material and not by the surrounding concrete strength if
existing design codes are applied.
2) The test program proves that the actual static strength of shear connectors embedded in a UHPC is greater than that obtained by the
AASHTO LRFD design calculation [14] by a margin of 2%13%. Therefore, the AASHTO LFRD design code [14] can be used to evaluate the
ultimate strength of a stud shear connector in a UHPC. A comparison
with the Eurocode-4 design code [15] provides a margin of 27%42%,
which may lead to a relatively conservative result.
3) The aspect ratio of the stud height-to-diameter is limited to at least 4
in existing design codes. The test program in this study proves that
the aspect ratio can be as low as 3.1 because it does not have much
impact on the structural behavior or performance of stud shear connectors.
4) The cover thickness over the stud head is limited to 50 mm in the
AASHTO LRFD design code [14]. The test program proves that a
cover thickness even as low as 25 mm does not lead to any cracking
or spalling in the UHPC slab deck or reduction in the static strength
of a stud shear connector.
5) According to the Eurocode-4 design code [16], the stud shear connectors should provide a characteristic relative slip of at least
6 mm to guarantee the ductile behavior of the stud. The test result
shows that the stud shear connectors embedded in a UHPC provide
characteristic relative slips of 3.85.3 mm that do not satisfy the required ductility demand of 6 mm. Therefore, another measure
should be used to resolve this ductility issue in order to use stud
shear connectors for a thin UHPC slab. Otherwise, an elastic design
should be applied for stud shear connectors for a UHPC deck,
resulting in a variable stud pitch rather than a constant pitch.
6) A UHPC slab deck for composite construction can be as thin as
75 mm using stud shear connectors with a diameter of 16 mm and
height of 50 mm even at the deck girder joint region.
Acknowledgements
This research was supported by a grant (13SCIPA02) from the Smart
Civil Infrastructure Research Program funded by the Ministry of Land,
Infrastructure and Transport (MOLIT) of the Korean Government and
Korea Agency for Infrastructure Technology Advancement (KAIA).
References
[1] Russell HG, Graybeal BA. Ultra-high Performance Concrete: A State-of-the-Art Report for the Bridge Community. Publication No. FHWA-HRT-13-060. Federal Highway Administration; 2013.
[2] Saleem MA. Alternative to Steel Grid Bridge Decks. PhD Dissertation Florida International University; 2011.
[3] Saleem MA, Mirmiran A, Xia J, Mackie K. Ultra-high-performance concrete bridge
deck reinforced with high-strength steel. ACI Struct J 2011;108(5):6019.
[4] Heimann J. The implementation of full depth UHPC wafe bridge deck panels. Publication No. FHWA-HIF-13-031. Federal Highway Administration Highways for LIFE;
2013.
[5] Aaleti S, Petersen B, Sritharan S. Design guide for precast UHPC wafe bridge deck
panel system, including connections. Publication No. FHWA-HIF-13-032. Federal
Highway Administration Highways for LIFE; 2013.
[6] Aaleti S, et al. Structural behavior of wafe bridge deck panels and connections of
precast ultra-high-performance concreteexperimental evaluation, transportation
research record. J Transp Res Board 2011;2251:8292.

30

J.-S. Kim et al. / Journal of Constructional Steel Research 108 (2015) 2330

[7] Aaleti S, Sritharan S. Design of ultrahigh-performance concrete wafe deck for accelerated bridge construction. Transport Res Rec: J Transp Res Board 2014;2406:1222.
[8] Chen D, El-Hacha R. Behaviour of hybrid FRPUHPC beams in exure under fatigue
loading. Compos Struct 2011;94:25366.
[9] El-Hacha R, Chen D. Behaviour of hybrid FRP-UHPC beams subjected to static exural loading. Compos Part B 2012;43:58293.
[10] Chen D, El-Hacha R. Damage tolerance and residual strength of hybrid FRP-UHPC
beam. Eng Struct 2013;49:27583.
[11] Nguyen H, Mutsuyoshi H, Zatar W. Flexural behavior of hybrid composite beams.
Transport Res Rec: J Transp Res Board 2013;2332:5363.
[12] Nguyen H, Zatar W, Mutsuyoshi H. Hybrid ber-reinforced polymer girders topped
with segmental precast concrete slabs for accelerated bridge construction. Transport
Res Rec: J Transp Res Board 2014;2407:8393.
[13] Ollgaard JW, Slutter RG, Fisher JW. Shear strength of stud connectors in lightweight
and normal-weight concrete. AISC Eng J 1971;8(2):5564.
[14] AASHTO. AASHTO LRFD Bridge Design Specications. Washington DC: American Association of State Highway and Transportation Ofcials; 2012.
[15] CEN. 1994-2 Eurocode 4: Design of composite steel and concrete structures. Part 2:
General Rules and Rules for Bridges; 2005.

[16] CEN. 1994-4-4 Eurocode 4: Design of composite steel and concrete structures. Part
1-1: General rules and rules for buildings; 2004.
[17] AWS. Structural welding code-steel, AWS D1.1:2000. 17th ed. Miami, FL: American
Welding Society; 2000.
[18] Anderson NS, Meinheit DF. Design criteria for headed stud group in shear: Part
1Steel capacity and back edge effects. PCI J 2000;25:4675.
[19] Oehlers DJ, Coughlan CG. The shear stiffness of stud shear connections in composite
beams. J Constr Steel Res 1986;6(4):27384.
[20] Shim C-S, Lee P-G, Yoon T-Y. Static behavior of large stud shear connectors. Eng
Struct 2004;26(12):185360.
[21] Xu C, Sugiura K. FEM analysis of failure development of group studs shear connector
under effects of concrete strength and stud dimension. Eng Fail Anal 2013;35:
34354.
[22] Hegger J, Feldmann M, Rauscher S, Hechler O. Load-deformation behavior of shear
connectors in high strength concrete subjected to static and fatigue loading. IABSE
Symposium Report Budapest 2006: Responding to Tomorrows Challenge in Structural Engineering. IABSE; 2006. p. 1724.

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.