Structures 44 (2022) 988–1010

Contents lists available at ScienceDirect

Structures
journal homepage: www.elsevier.com/locate/structures

Behaviour and design of the ‘lockbolt’ demountable shear connector for

sustainable steel-concrete composite structures
Jun He a, b, Ahmed S.H. Suwaed c, George Vasdravellis d, *, Sihao Wang e
a
School of Civil Engineering, Changsha University of Science and Technology, China
b
Institute for Infrastructure and Environment, Heriot-Watt University, UK
c
Department of Reconstruction and Projects, University of Baghdad, Iraq
d
Institute for Infrastructure and Environment, Heriot-Watt University, UK
e
Department of Bridge Engineering, Tongji University, China

A R T I C L E I N F O A B S T R A C T

Keywords: In order to promote sustainable steel-concrete composite structures, special shear connectors that can facilitate
Demountable shear connector deconstruction are needed. A lockbolt demountable shear connector (LB-DSC), including a grout-filled steel tube
Lockbolt embedded in the concrete slab and fastened to a geometrically compatible partial-thread bolt, which is bolted on
Grout-filled tube
the steel section’s top flange of a composite beam, was proposed. The main drawback of previous similar
Pushout test
demountable bolts is the sudden slip of the bolt inside its hole. This bolt has a locked conical seat lug that is
Finite element model
Shear strength secured inside a predrilled compatible counter-sunk hole in the steel section’s flange to provide a non-slip bolt-
Relative slip flange connection. Deconstruction is achieved by demounting the tube from the top of the slab by unfastening
using a simple modified wrench. The mechanical behaviour of the proposed connector is assessed by four
pushout tests that were conducted per Eurocode 4 recommendations. The tests showed high shear resistance, and
high stiffness as compared to other DSCs, while the slip capacity results classified the LB-DSC as a ductile shear
connector according to Eurocode 4. A refined nonlinear finite element model (FEM) was validated through the
tests and reliably reproduced the experimental behaviour. Consequently, the calibrated FEM model was applied
to carry out extensive parametric analyses to investigate the strength and geometry effects of concrete slab,
infilled grout, tube, and bolt on the structural behaviour of the LB-DSC. On the basis of numerical and experi­
mental results, a design equation is derived to predict the shear resistance of the LB-DSC.

1. Introduction and concrete, high strength and stiffness, and accelerated prefabricated
construction [6–10]. The shear connection type between steel and
In order to achieve future sustainability targets, such as the “Net zero concrete components is inherently significant, because it enables the
carbon target” [1] and the Green Deal [2,3], more attention should be composite section to behave as one unit. In addition, the connection
paid to the construction sector, because the production of most must be able to effectively resist separation between the two compo­
commonly used construction materials (e.g. the cement and steel) lead nents [11–15]. Among various shear connectors in steel–concrete com­
to 15 % of the world human-induced CO2 emissions. Furthermore, it is posite structure, headed studs have become the most practical type for
anticipated that the material demand will be doubled in the next 20 achieving the composite action due to their standardized welding
years, indicating that a more responsible use of natural resources is technique, good mechanical performance, and extensive research data
highly needed [4]. Furthermore, considerable quantities of waste sent to available [16–21].
landfill due to construction and demolition activities. In order to However, from the viewpoint of sustainable development, composite
maintain a sustainable international environment, using demountable beams with welded headed studs are not recommended because of the
and reusable structural systems can be the solution [5]. difficulty of detaching the concrete slab from the supporting steel sec­
Steel-concrete structures are extensively used in buildings and tions at the end-of-life of a structure, or because of the excessive
bridges, thanks to the composite action of the individual benefits of steel downtime required to repair or replace a deteriorated composite bridge

* Corresponding author at: Heriot-Watt University, Riccarton, EH14 4AS, UK.

E-mail address: g.vasdravellis@hw.ac.uk (G. Vasdravellis).

https://doi.org/10.1016/j.istruc.2022.08.062
Received 23 March 2022; Received in revised form 21 June 2022; Accepted 12 August 2022
Available online 22 August 2022
2352-0124/© 2022 The Author(s). Published by Elsevier Ltd on behalf of Institution of Structural Engineers. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
J. He et al. Structures 44 (2022) 988–1010

Fig. 1. (a) Composite girder with LB-DSCs; (b) Assembly of LB-DSC; (c) Details of LB-DSC.

deck. Thus, recently some research has been conducted to develop novel and showed comparable behaviour to beams with headed studs.
shear connectors that allow for an easier separation of concrete slabs Recently, highly ductile demountable steel hollow sections as shear
from the steel sections and reuse of structural components. connectors for hollow-core precast slabs were developed [42,43] and
Pre-tensioned high strength friction-grip (HSFG) bolts were experi­ showed higher stiffness and slip capacity compared to welded studs.
mentally investigated as shear connectors [22–24]. Three types of post- Previous tests on bolts shear connector revealed an unfavourable
installed shear connectors to retrofit non-composite bridges [25,26] sudden slip under serviceability limit state (SLS) due to bolts sliding
were tested and showed significant increase in resistance and stiffness as inside their holes when friction resistance is exceeded, which may pre­
well as good fatigue performance. Bolt shear connectors were inspected vent their possible application in practice. Thus, epoxy resin injected
and showed low stiffness in comparison with headed studs [27,28]. The inside the bolt holes was used [44]; however, the slip capacity was still
behaviour of HSFG bolts in composite beams with geopolymer precast rather limited. Preloaded bolts based on steel–concrete friction resis­
concrete slabs was examined and identified significant shear resistance, tance were tested by pushout and beam tests and showed full-interaction
large slip capacity [29], and enough absorbed energy capacity when performance under SLS [45,46]. A locked-nut DSC was proposed, using a
subjected to low-cycle high-amplitude loading [30,31]. A novel cone-shaped nut that “locks” into counter-shaped corresponding holes in
connector consisted of a short bolt, long bolt, and coupler was proposed, the steel section flange to avoid the sudden slip [47]. Presently, a welded
however, a ductile behaviour can only be obtained when the diameter of demountable connector to eliminate the sudden slip was validated
the bolts was no less than 27 mm [32]. In addition, extensive research experimentally and numerically [48,49]. Inspired by the concept
has been conducted [33–36] on shear bolts machined from headed described in [47], the authors proposed a lockbolt demountable shear
studs. The results showed comparable resistance, higher slip capacity, connector (LB-DSC) to eliminate the sudden slip as well the tolerance
but about half the initial stiffness of the headed studs due to unavoidable issues in the holes drilled in the flange, and to allow for an easy sepa­
clearance between the bolt collar and its hole. Moreover, blind bolts ration of the slabs from the top of the composite floor.
were proposed [37] to retrofit composite beams and found that they had
higher shear resistance in comparison to headed studs, but with smaller 1.1. LBDSC description
stiffness. The blind bolts were also tested in full-scale composite beams
[38], under sustained loads [39], and under dynamic loading [40,41], Fig. 1a shows the details of the proposed LB-DSC. It includes a grout-

989
J. He et al. Structures 44 (2022) 988–1010

Fig. 2. Specimen preparation in the Lab: (a) Assembled shear connector without grout on steel section; (b) filled in grout, installed reinforcement cage and formwork;
(c) cast concrete.

filled headed steel tube with internal partial thread fastened over a 2. Experimental programme
compatible partial thread bolt (Fig. 1c). The bolt has a conical lug that
fits in compatible pre-drilled counter-sunk conical seat holes in the steel 2.1. Test specimens
flanges and, therefore, excludes the sudden slip and resolves any toler­
ance issues. The headed steel tube with 33–37 mm external diameter To investigate the shear performance of the LB-DSC, four standard
(depending on the bolt diameter and thickness of tube) is a machined pushout test specimens in accordance with Eurocode 4 [51] were
monolithic unit. The bottom thicker segment of the tube includes three fabricated, as shown in Fig. 2. The test specimen comprised of 600-mm
parts from up to down: A threaded part to connect the tube to the bolt; long HEB260 steel section attached to two concrete slabs (650 × 600 ×
then an enlarged unthreaded hole to hide bolt threads, preventing shear 150 mm), using two or four LB-DSCs in each slab. Holes with counter­
fracture through bolt’s thread; next, a transition chamfered hole, to sunk conical seat were drilled in each steel section, then partial thread
decrease stress concentration. To improve the shear resistance and bolts, as described in section 1.1, with compatible conical lug were
stiffness of the LB-DSC, flowable grout should be filled into the tube and locked inside these holes creating a stop mechanism against sudden slip
reached the bottom of the bolt though six longitudinal holes passed below SLS. Headed tubes with compatible partial internal thread were
through the previous three parts, as shown in Fig. 1(c). fasten over each bolt. Flowable grout of QuickCEM with characteristics
The construction procedure of the LB-DSC is depicted in Fig. 1b. Step of quick setting and hardening were poured inside each tube to accel­
1: counter-sunk holes were bored on the steel section’s flange, the erate test specimens’ fabrication process. Two layers of Ø10 mm
partial-thread bolts with conical seat lug were placed inside the holes deformed rebars were placed in each slab. The details of the pushout test
and fastened through the nut below the flange; Step 2: firmly hand- specimens are shown in Fig. 3. Specimens LB-DSC (1, 3 and 4) used the
tightening the tube to the bolt; Step 3: The tube was filled with flow­ flowable grout, while LB-DSC2 did not. In addition, the effects of the
able grout; Step 4: Steel reinforcements were arranged and concrete was compressive strength of concrete slab (50.4 or 63.3 MPa), number of
poured for the slab. connectors (2 or 4) in each slab and loading procedure: Monotonic (M)
The deconstruction can be realized by loosening the nut shown in (b) and Cyclic (C) loading were considered, as shown in Table 1.
and (c) of Fig. 1 from the bottom face of the steel–concrete composite
beam, or, preferably, by loosening the tube from the top face using an 2.2. Material properties
electric or mechanical compatible wrench. Thus, the concrete slab and
steel section can be easily detached from each other; and both can be The mechanic properties of grout mix and in-situ concrete of test
reused in other sites. specimens were obtained by concrete compression tests. Three con­
The structural behaviour of the LB-DSC was checked using a series of crete/grout cubes (100 mm as per BSI [52]) were prepared on the same
pushout tests using a bespoke horizontal testing arrangement to imitate date as the corresponding specimen using the same fresh concrete and
real beam conditions in [50]. The results indicated that the LBDSC has tested together with the corresponding pushout test. Table 1 summarizes
high shear resistance, stiffness, and slip capacity; however, the presence the average strengths.
of non-negligible uplift forces as a result of the specific horizontal test The material properties of the tube and bolt were acquired from
setup, implied that further investigation is needed to understand the standard tensile coupon tests [53] including the modulus of elasticity,
shear performance of the LBDSC and to propose accurate design the yield stress (0.2 % proof stress), the tensile strength and corre­
equations. sponding tensile strain. Table 2 lists the results as average values of three
Therefore, this paper presents a series of pushout tests that were identical coupons. Moreover, the engineering stress–strain curves for
carried out following the standard testing procedure of Eurocode 4 [51] tube and bolt are shown in Fig. 7.
to investigate the structural behaviour of the LB-DSC. A detailed finite
element method (FEM) model considering material and contact 2.3. Test setup
nonlinearity was created and calibrated against the experimental re­
sults. Then, the calibrated FEM model was applied to perform para­ A hydraulic actuator with force capacity of 2000 kN and displace­
metric analyses to study the effects of concrete strength of the slab, bolt ment capacity of 120 mm was applied to move the steel section in
diameter/strength, steel tube thickness/strength, and presence/strength relative to the attached concrete slabs using a stiff steel plate, as indi­
of infill grout on the shear resistance and stiffness of the LB-DSC. Finally, cated in Fig. 4. A layer of plaster gypsum of high strength was used to
design equations for the prediction of the shear resistance of the LB-DSC embed the bottom of concrete slabs on the rigid plate. Belts, shown in
are proposed to facilitate the application of the proposed connector in Fig. 4, were barely attached around the specimens to prevent the slabs
sustainable steel–concrete composite buildings. from falling sideway once the connectors are disconnected.

990
J. He et al. Structures 44 (2022) 988–1010

Fig. 3. Details of test specimens (unit: mm): elevation and side view of (a) LB-DSC1 to 3 and (b) LB-DSC4; (c) plan view of all LB-DSCs; (d) steel tube; (e) partly
threaded conical bolt (f) counter-sunk hole in steel beam flange.

991
J. He et al. Structures 44 (2022) 988–1010

(d) (e)
D=30

10.4
17.5

Flange

4.5
2.6

21
(f)
Fig. 3. (continued).

Table 1
Push out test specimens.
Specimen name Concrete slab strength (N/mm2) Grout strength (N/mm2) Bolt diameter (mm) Tube thickness (mm) Number of connectors Loading procedure

LB-DSC1 50.4 48 20 2.5 2 M

LB-DSC2 50.4 No grout 20 2.5 2 C+M
LB-DSC3 63.3 48 20 2.5 2 C+M
LB-DSC4 50.4 48 20 2.5 4 C+M

peak load. The load was applied using a 0.005 mm/s displacement
Table 2 control. The recorded shear resistance by the loadcell is the division of
Material properties of steel. ultimate load over the number of LB-DSC.
Coupon Modulus of Yield stress Tensile stress Strain at Fig. 4 shows the instrumentation used for the specimens. Two or four
elasticity (GPa) (MPa) (MPa) ultimate stress Linear Variable Differential Transformers (LVDTs), vertically mounted
Bolt 210 818 990 0.090 at the level of each connector, were used to measure the relative slip
Tube 205 333 645 0.188 between the steel section and the concrete slab, while the separation at
the steel–concrete interface was obtained using two LVDTs horizontally
mounted near the connectors. All the devices were connected to a data
2.4. Loading procedure and instrumentation logger and the data were recorded at 10 Hz rate of continues data-
recording.
Fig. 5 shows the loading procedure that was adopted from Eurocode
4 and includes 25 load cycles between 5 and 40 % of the estimated ul­
timate load, followed by an imposed quasi-static monotonic load until
failure of the specimen or until load drop by 20 % with respect to the

992
J. He et al. Structures 44 (2022) 988–1010

Fig. 4. Test setup.

3. Finite element model bolting, were simulated using tie constraints. Complete bond (i.e.
without considering slip or debonding) was assumed between re­
3.1. Model description inforcements and the adjacent concrete.
The bottom of the rigid plate was assumed to be fully fixed, as shown
Abaqus, the nonlinear FEM program [54], was used to simulate a in Fig. 6a. Planes ‘A’ and ‘B’ of steel section and concrete slab represent
quarter of pushout test specimen as shown in Fig. 6. The 3-dimensional the symmetry boundary conditions. A displacement rate of 0.005 mm/s
8-node reduced integration solid element (C3D8R) was used to model was enforced on the steel section upper part to simulate the experi­
the steel section, concrete slab, infilled grout, bolts, nuts, and steel tube. mental program.
The rebar mesh was simulated using (T3D2) 2-node truss element. The
overall mesh used to model tube and bolt was 2 mm to give more precise 3.2. Material modelling
predictions of the load-slip curves and failure modes.
Formulations of hard contact and penalty friction were employed to 3.2.1. Model of steel
simulate the following interfaces: concrete slab-tube, grout-bolt, grout- A bi-linear elastic–plastic stress–strain curve was used to represent
tube, and bolt lug-countersunk hole. The friction coefficients were set the stress–strain curve of the steel section (S355) and the steel re­
as 0.4 for steel–concrete interfaces and 0.25 for steel-steel interfaces, inforcements (S500), using nominal material characteristics. The true
respectively, which were determined through parametric sensitivity stress–strain curve of the high strength bolt and the steel tube were
analysis and verified by the comparison of load-slip curves from FE chosen based on the standard tensile tests, as displayed in Fig. 7a and b,
analysis with those from previous test results [49,50,55]. The thread-to- respectively.
thread connections in the bolt – tube bolting, as well in the bolt – nut
3.2.2. Model of concrete and grout
Fig. 8 shows uniaxial compressive and tensile stress–strain curve that
represent the nonlinear behaviour of concrete and grout. Three parts
represent the relation of stress (σc) and strain (εc) for concrete in
compression. The first part (Eqs. (1) and (2)) is when the stress is below
0.4 fc, according to fib-2010 [56], where fc is the cylinder concrete
compressive strength. The strain corresponding to the peak stress fc is
defined as peak strain εcp, and the value is equivalent to 0.0025.
σ c1 = Ec εc (1)

Ec = Ec0 ⋅αE ⋅(fc /10)1/3 (2)

where Ec is the modulus of elasticity of concrete (GPa); Ec0 = 21.5 GPa;

αE = 1.0 for quartzite aggregates.
The second part is the nonlinear parabolic portion (Eq. (3)) ranging
from 0.4 fc to fc, as follows:

Fig. 5. Loading procedure.

993
J. He et al. Structures 44 (2022) 988–1010

Fig. 6. Finite element model of pushout test: (a) Model of the specimen; (b) Mesh details of the LB-DSC.

( )2
Ec εc /fc − εc /εcp where σ t and ft are the tensile stress and ultimate strength of concrete,
σ c2 = fc ( ) (3)
1 + Ec εc /fc − 2εc /εcp respectively (MPa); w is the crack width (mm); wc is the crack width
when the stress is completely released, wc = 5.14GF/ft (mm); GF is the
The third part of the stress–strain curve is a descending linear branch
fracture energy needed for the creation of a unit area of stress-free crack,
(Eq. (4)). The stress at the ultimate strain εcu is assumed to be 0.85 fc.
(N/mm); c1 and c2 are the constants: c1 = 3 and c2 = 6.93.
( )
ε − εcp Abaqus model of concrete damage plasticity (CDP) was adopted to
σ c3 = fc 1 − 0.15 c (4) describe the degraded response of concrete, which assumes that tensile
εcu /εcp
cracking and compressive crushing are two main failures, their degraded
A linear stress–strain relationship was implemented for the un- degrees are expressed as dc (Eq. (6)) and dt (Eq. (7)), respectively.
cracked concrete section under tension. The modulus of elasticity in
σc
tension was presumed the same as that in compression. After the con­ dc = 1 − (6)
crete section cracks, a nonlinear relationship of stress-crack width (Eq. Ec εpl
c (1/bc − 1) + σ c

(5)), is assumed.
σ t l0
[ ( )3 ] dt = 1 − (7)
σt w w( ) Ec wpl (1/bt − 1) + σt l0
(5)
w
= 1 + c1 e− c2 wc − 1 + c31 e− c2
ft wc wc
where bc is the ratio of plastic strain εpl in pl in
c to inelastic strain εc ,bc = εc /εc ;

994
J. He et al. Structures 44 (2022) 988–1010

Fig. 7. Stress–strain relationship for (a) Bolt and (b) Tube.

Fig. 8. Material stress strain law used for concrete and grout (a) in compression and (b) in tension.

and bc is taken as 0.7, assuming that most of the inelastic compressive concrete slab as can be observed in Fig. 10a. In addition, the peak
strain maintains after unloading; l0 is assumed to be one-unit length; bt is applied load was 1755 kN in this specimen, which means that the
the ratio of the “plastic” crack width wpl to the crack width w, bt = wpl/w, compressive stress in each slab was higher than the rest of the pushout
and bt is set as 0.1, which means that the unloading returns almost back tests.
to the origin [57].
4.2. Load-slip behaviour
4. Experimental and FEM results
Fig. 11 describes the load-slip behaviour of the specimens. A typical
4.1. Failure modes
load-slip curve includes three main stages. The first stage is linear and
involves elastic deformations, starting from 0.0 to 0.3–0.5 mm of slip,
In the specimens where two connectors in each slab were installed, i.
and matches to loads from 60 to 80 kN (i.e., about 33 % of shear
e. LB-DSC1, LB-DSC2 and LB-DSC3, bolt shear failure happened at the
resistance). The 25 loading cycles show that the difference in residual
shank above the lug, where the stress reached the tensile strength at the
slip between two successive cycles approaches zero as the cycles ap­
ultimate state as shown from the FEM simulations in Fig. 9b and 10b. At
proaches 25 which indicate that the current connector is reliable and not
the same time, a small amount of concrete in front of the fractured bolt
susceptible to premature or progressive failures. As the load increases,
was crushed; the plastic strain in the concrete at this local area was more
bolt bending, shearing and bearing against the surrounding grout and
than the ultimate compressive strain of concrete, as shown by the FEM
concrete result in reduction of LB-DSC stiffness, which leads to a
model in Fig. 9c and 10c. The specimen with four LB-DSCs installed in
nonlinear load-slip relationship which defines the second stage. After
each slab (LB-DSC4) failed unexpectedly due to cracking of the slab
the maximum load, the slip increases rapidly, and the load capacity
mainly at the bottom part as shown in Fig. 10d, and with extreme
drops abruptly due to bolt shear failure just above the lug of the shank
yielding of bolt’s shank just above the lug but without fracture, as shown
which defines the third stage. For the case of LB-DSC4 specimen, the
in Fig. 10e. The initiation and propagation of concrete cracks before the
stiffness reduction is due to, in addition to bolt’s deformation, excessive
bolt fracture can be attributed to absence of gypsum plaster beneath the
cracks in the concrete slab. It is worth noting that the slip capacity of all

995
J. He et al. Structures 44 (2022) 988–1010

LB-DSC2-Steel

Loading

LB-DSC2-Concrete

(a)

(b) (c)
Fig. 9. Failure mode of Specimen LB-DSC2 without infill grout: (a) Bolt shear failure from test; (b) Mises stress of shear connectors at ultimate state from FEM; (c)
PEEQ of concrete slab at ultimate state from FEM.

the specimens was more than 6 mm, which fulfils the Eurocode 4 0.95, with a standard deviation of 0.08, implying that the numerical
requirement for ductile shear connectors. In addition, the validation of model was able to reliably calculate the separation of the LB-DSCs.
the FEM model was proved by noticing that FEM and experimental Recent pushout tests conducted by the authors in [50] assessed the
curves are in good agreement with each other, showing that the nu­ behaviour of LB-DSCs using a horizontal setup of pushout test. The tests
merical model can capture all the stages of the load-slip responses with in [50] recorded values higher than 6 mm of the separation displace­
satisfactory agreement. Note that the effect of the 25 initial cycles of ment at peak load, resulting in tension forces in bolts of the LB-DSCs up
loading on the stiffness of the LBDSC was not considered in the FEM to 37 % of the shear resistance. This high value of tension indicated that
simulations. tension-shear interaction may have compromised the shear resistance of
Fig. 12 shows the load-separation relationships of all test specimens. the connectors in the horizontal setup pushout tests. Therefore, in the
It should be noted that the measured separation for LB-DSC1 was acci­ present standard pushout tests, the separation was also analysed using
dentally lost when the applied load was in between 60 and 208 kN. The the FEM model and the results are shown in Fig. 13. The maximum
separation capacity (defined as the measured separation at 80 % of the tension force is 42 kN (except for LB-DSC4, for which it is 58 kN). Fig. 14
peak load) was recorded for all test specimens and the maximum was illustrates the relation between the tension force and the shear resis­
0.51 mm or 18 % of the corresponding slip, which is below the limit of tance. No matter whether the test specimen infilled with or without
50 % in Eurocode 4. grout, the tension force resisted by the “hat” of the tube for LB-DSC1-3 is
In addition, the mean value of the ratio of separation at ultimate load 13–15 % of the shear force. This ratio is less by about 50 % than that (23
from the FEM model to that from pushout tests (except LBDSC4, since it % to 37 %) from horizontal pushout tests [50], and therefore the tension
failed due to concrete cracks, leading to large separation at failure) is in the LB-DSC can be ignored. For specimen LB-DSC4 that failed due to

996
J. He et al. Structures 44 (2022) 988–1010

LB-DSC1-Steel LB-DSC3-Steel

LBDSC1-Concrete
LB-DSC3-Concrete

LB-DSC4-Concrete

(a)

(b) (c)

(d) (e)

Fig. 10. (a) Failure mode of specimens LB-DSC1, 3 and 4 from tests; (b) Mises stress and (c) PEEQ at ultimate state of LBDSC3 from FEM; (d) PEEQ and (e) Mises
stress at ultimate state of LBDSC4 from FEM.

997
J. He et al. Structures 44 (2022) 988–1010

Fig. 11. Load versus slip responses.

concrete cracks, the ratio is slightly larger. Fig. 15 provides a compari­ deviations of 0.03, 0.02 and 0.06, respectively, indicating the FEM can
son between the load-slip curves of specimen LBDSC1 (from the present reliably estimates the behaviour of the LB-DSC.
standard pushout tests) and specimen LN9 (from the horizontal pushout To evaluate the presence of infill grout on the structural perfor­
tests presented in [50]). The two specimens have the same geometry of mance, LB-DSC2 is compared to LB-DSC1 in Fig. 16a. The grout of 48
connector, while the strength of the concrete slab is slightly different: MPa compressive strength improved the shear resistance by 10 % and
LB-DSC1 has 50 MPa, while LN9 has 37 MPa. It can be observed that the the stiffness by 230 %; hence, a significant enhancement in LB-DSC
shear stiffness is almost the same, whilst the shear resistance of LB-DSC1 stiffness can be achieved with the presence of the infill grout and, to a
is higher than that of LN9 by 23 %. The slip capacity of LB-DSC1 is 10.9 lesser amount, in the shear resistance. Evaluation of the effect of con­
mm which is less than that of LN9 by 148 %. The higher slip capacity in crete compressive strength was done by comparing LB-DSC1 (50.4 MPa)
LN9 is attributed to a large rotation that occurred in the connectors of to LB-DSC3 (63.3 MPa) as shown in Fig. 16b. The shear resistance and
the horizontal setup tests due to the secondary moments induced from stiffness improved by 12 % and 11 %, respectively, thus, increasing
the eccentricity [50], and thus it is not representative of the slip capacity concrete strength of the slab can slightly improve the structural per­
of the proposed connector in a composite beam. Therefore, the standard formance. In terms of the effect of number of bolts, when it was
pushout tests presented in this paper are used to propose more reliable increased from 2 per slab in LB-DSC1 to 4 per slab in LB-DSC4 (Fig. 16c),
design equations to predict the shear resistance of the LB-DSC. the experimental shear resistance per bolt was reduced by 16 %, while
the shear stiffness per bolt was increased by 13 %; therefore, a group
4.3. Shear resistance, ductility, and stiffness of the LB-DSC effect similar to that for welded shear studs [20] and bolt connectors
[58] is also observed for the LB-DSCs. A possible explanation to the 16 %
Table 3 shows the results from both pushout tests and the numerical reduction in shear resistance is that the specimen LB-DSC4 failed un­
simulations including shear stiffness k, shear resistance Fu, corre­ expectedly due to cracking of the slab mainly at the bottom part as
sponding slip Su, slip capacity SP, residual slip SR after 25 load cycles, shown in Fig. 10d as explained before in Section 4.1.
and separation capacity Up. The shear stiffness is adopted as the secant
line at Fu/3 of load-slip curve, since the curve is almost linear until that 4.4. Comparison with recent DSCs
load.
The average ratios of numerical to experimental shear stiffness, shear Fig. 17 shows a comparison of the load-slip relationships of the
resistance and corresponding slip are 0.95, 1.00 and 0.97, with standard LBDSC, threaded studs (M9), bolts (M10) [59], blind bolts (B1-Lindapter

998
J. He et al. Structures 44 (2022) 988–1010

Fig. 12. Load versus separation responses.

Hollo-Bolt; B2: Ajax Bolt) [60], friction bolts with cast-in cylinders compressive strength, infilled grout compressive strength, and bolt
(P3.1), embedded mechanical coupler device with pre-tensioned bolts tensile strength on the LBDSC shear behaviour. The values of the default
(P15.1) or with injection bolts (P15.2) [44], embedded bolts (M3) [31], (base) parameters are: bolt diameter = 20 mm; bolt tensile strength =
embedded bolts with coupler (B9) [32], high-tension friction grip bolts 990 MPa; tensile strength of tube = 645 MPa, thickness of tube = 2.5
(T1-4) [61], and locking-nut LNDSC (test 10) [62]. Table 4 lists the mm; concrete compressive strength = 15 or 50 MPa, and compressive
characteristics of geometry and material of the aforementioned DSCs, in strength of grout = 45 MPa.
addition to the failure mode, shear resistance, slip capacity, and
normalized resistance, defined as the ratio R of tested shear resistance 5.1. Compressive strength of concrete slab
(Qu) to tensile resistance (fu(πD2/4) of the bolt for comparison of DSCs
with different diameter and tensile strength. To understand the influence of concrete compressive strength of the
The curves shown in Fig. 17 are further categorised according to the slab on the LB-DSC shear resistance and stiffness, the compressive
failure mode, i.e. bolt fracture (Fig. 17a) or concrete failure (Fig. 17b). strength of the slab was chosen as 15, 30, 40, 50 and 60 MPa, while the
The comparison reveals that both the shear resistance and the initial other parameters were kept constant. The corresponding load versus slip
stiffness of the LB-DSC are the highest among the bolted DCSs regardless curves shown in Fig. 18 show that the shear stiffness ascends almost
of the failure mode. In addition, the LB-DSC has competitive slip ca­ linearly, and the shear resistance ascends nonlinearly when the
pacity in comparison to friction bolts and the LNDSC, but pretension is compressive strength of the concrete slab ascends from 15 N/mm2 to 60
not required to be applied to the LB-DSC. The “locking” mechanism of N/mm2. The increments are 30 % and 24 % for shear resistance and
the bolt into the counter-sunk hole in the steel flange effectively pre­ stiffness respectively, indicating that increasing compressive strength of
vents the sudden slip of the bolt inside the hole due to the exceedance of concrete can enhance the performance of the LB-DSC. In addition, it is
the frictional resistance, resulting in higher initial stiffness of the LB-DSC shown that the increase in compressive strength is insignificant for
as compared to the majority of other bolted DSCs. concrete strength greater than 30 MPa, since the failure mode converted
from concrete cracking to bolt shearing-off.
5. Parametric study

5.2. Thickness of tube

The calibrated FEM model was adopted to conduct a parametric
study to investigate the effects of several parameters on the structural
The variation of thicknesses of steel tube was 2, 2.5 and 3 mm and it
performance of the LB-DSC. The investigated parameters include the
was considered in both specimens with low compressive strength of
bolt diameter, tube thickness, tube tensile strength, slab concrete
concrete slab (15 MPa) and with high strength (50 MPa), as shown in

999
J. He et al. Structures 44 (2022) 988–1010

Fig. 13. Tension force versus separation for test specimens.

Fig. 14. The relation of shear and tension force at ultimate state from FEM.

Fig. 19. Increases of 18 % and 5 % in the shear resistance were recorded Fig. 15. The comparison of load-slip curves between standard and horizontal
as the thickness changes from 2 to 3 mm for low and high strength push tests for LBDSC.
concrete slab, respectively. The shear stiffness almost keeps the same (5
% variation). Thus, using thicker steel tubes guarantees higher shear 5.3. Tensile strength of tube
resistance of the LB-DSC when it is embedded in low strength concrete
slab, whilst the effect is insignificant in high strength concrete slab. Different tube tensile strengths ranging from 400 to 700 MPa were
considered in models with slab concrete compressive strength of 15 MPa
and 50 MPa respectively. Fig. 20 shows the corresponding load-slip

1000
J. He et al. Structures 44 (2022) 988–1010

Table 3
Comparison of experimental and numerical results.
Specimen Shear resistance Slip at Fu Slip Residual Separation Stiffness at 1/3 Fu Failure mode
Fu (kN) Su(mm) capacity slip capacity k (kN/mm)
Sp (mm) SR (mm) Up (mm)
Test FEM FEM Test FEM FEM Test Test Test Test FEM FEM
/Test /Test /Test

LB-DSC1 263 265.1 1.01 9.42 9.10 0.97 10.9 – 0.10 141 139 0.99 Bolt shear off
LB-DSC 2 239 239.3 1.00 11.35 10.75 0.95 12.7 1.70 0.39 61 57 0.93 Bolt shear off
LB-DSC 3 295 287.6 0.97 10.43 9.44 0.91 12.6 0.92 0.62 157 152 0.97 Bolt shear off
LB-DSC 4 219 222.4 1.01 9.82 10.40 1.06 9.8 0.88 0.51 160 147 0.92 Concrete crush and
split
Average 254 1.00 0.97 11.5 153* 0.95
Standard 0.02 0.06 0.03
deviation
*
Specimens with grout-infilled tube.

Fig. 16. Comparison of load-slip curves for test specimens: (a) with or without grout; (b) with different strength of concrete slab; (c) with different
connector number.

1001
J. He et al. Structures 44 (2022) 988–1010

(a) (b)
Fig. 17. Comparison of load-slip curves for bolt DSCs: due to (a) Bolt fracture failure; (b) Concrete failure.

Table 4
Comparison of LBDSC with other bolt DSCs.
Specimens Bolt diameter D Bolt tensile strength fu Concrete slab strength fc Pretension Shear resistance Qu R = Q u/ Slip Failure
(mm) (MPa) (MPa) T (kN) (kN) fu(πD2/4) capacity mode

LB-DSC1 20 990 50.4 262.6 0.84 10.9 BF

LB-DSC4 20 990 50.4 219.4 0.71 9.8 CF
B9 22 985.6 48.9 219 0.58 3.37 BF
P3.1 20 948.7 59.4 100 136 0.46 9.4 BF
P15.1 20 948.7 44.3 176 142 0.48 9.0 BF
P15.2 20 948.7 44.3 0 131 0.44 5.6 BF
M3 20 925 52.2 171 0.59 11 BF
T1-4 20 1150 50 155 213 0.59 20.7 BF
M9 20 510 50.8 80 0.50 6.1 CF
M10 20 768.8 51.3 87 0.36 7.9 CF
B1 20 920 42 126 0.44 1.9 CF
B2 20 900 42 123 0.44 11.6 BF
Test10 16 889 43 88–106 181 1.01 16.2 BF

Note: BF: bolt fracture failure; CF: concrete failure.

curves. The shear resistance was improved by 16 % and 6 % for speci­ resistance by 12 % and 20 % in LB-DSCs with high and low strength
mens with low and high strength slab respectively, as the tensile concrete slabs, respectively. Since the bolt was confined by infilled
strength increases from 400 MPa to 700 MPa. The shear stiffness prac­ grout, both the stiffness and the resistance of the LB-DSC were improved,
tically did not change for all specimens (6 % variation). Therefore, using especially in specimens with bolt fracture failure. Fig. 21 shows that the
steel tube of high tensile strength lightly improves the shear resistance of effect of compressive strength of the infilled grout results in minor en­
the LB-DSC embedded in low compressive strength concrete slab, whilst hancements in shear resistance and stiffness of the LB-DSC. Therefore,
the effect is insignificant for connectors embedded in high strength the steel tube should be filled with grout, but the compressive strength of
slabs. the in-filled grout is not important as long as it is beyond 30 MPa.

5.4. Compressive strength of in-filled grout 5.5. Bolt’s diameter

Fig. 21 depicts the effect of the infilled grout in the tube, and its Fig. 22 shows the effect of varying bolt’s diameter as follows (16, 20,
compressive strength considering the values of 30, 45 and 60 MPa, on 22 and 25 mm) on the shear resistance and stiffness, for low and high
the performance of the LB-DSC. The infilled grout can considerably compressive strength concrete slabs. It should be noted that the inner
improve the stiffness by up to 240 % with respect to no grout-infilled diameter of the steel tube was changed to 26, 30, 32, and 35 mm,
connectors. Additionally, the infilled grout can increase the shear accordingly, to ensure that the bolt’s base is confined by the grout. It can

1002
J. He et al. Structures 44 (2022) 988–1010

Fig. 18. Effect of concrete slab strength: (a) load-slip curves; (b) Relation between shear strength, stiffness and concrete strength of the slab.

Fig. 19. Effect of steel tube thickness: load-slip curves for specimens with concrete slab of (a) 15 MPa and (b) 50 MPa; (c) Relation between shear strength, stiffness
and steel tube thickness.

1003
J. He et al. Structures 44 (2022) 988–1010

Fig. 20. Effect of tube tensile strength: load-slip curves for specimens with concrete slab of (a) 15 MPa and (b) 50 MPa; (c) Relation between shear strength, stiffness
and tube tensile strength.

1004
J. He et al. Structures 44 (2022) 988–1010

Fig. 21. Effect of infilled grout: load-slip curves for specimens with concrete slab of (a) 15 MPa and (b) 50 MPa; (c) Relation between shear strength, stiffness and
infilled grout strength.

be found that both the shear resistance and stiffness increase nonlinearly 6. Design rules for the shear resistance of the LBDSC
as the bolt diameter increases. The increments in shear resistance and
stiffness are about 50 % and 35 % respectively when the bolt diameter The shear resistance of headed studs in steel–concrete beams is
changes from 16 mm to 25 mm. Therefore, using bolts with large di­ dictated by either fracture of the stud or cracking of the concrete slab.
ameters can significantly enhance the shear resistance and stiffness of Therefore, various design codes provide formulas to estimate the shear
the LB-DSC. resistance based on the minimum of the two failure modes. The design
equations pertaining to welded headed studs prescribed in the Eurocode
4 [51], AISC360-10 [63], ACI 318-08 [64], GB50017 [65], AASHTO
5.6. Bolt’s tensile strength [66] are summarised in Table 5.
Currently, the design shear resistance of the LB-DSC is not included
Fig. 23 shows the influence of bolt tensile strength (800, 990, and in design codes. Based on the code-based shear resistance prediction
1100 MPa for Bolt 6.8, 8.8 and 10.9) on the shear resistance and stiffness expressions in Table 5 for headed studs, the following equation is pro­
of the LB-DSC, in low and high compressive strength concrete slabs. The posed to estimate the shear resistance of the LB-DSC:
shear resistance rises almost linearly with the increasing of the bolt √̅̅̅̅̅̅̅̅̅
( )
tensile strength, and the increasing ratio of the specimens with high ′
Pu = min α1 At Ec f c , α2 As fu (8)
compressive strength slab (31 %) is larger than that of the specimens
with low compressive strength slab (13 %) when the tensile strength of where, Pu is shear resistance; As is cross-sectional area of the bolt shank;
bolt of 800 MPa enlarged to 1100 MPa. However, the stiffness almost At is cross-sectional area of grout-infilled tube; fc′ is cylindrical
remains unaffected with only 3 % variation. Thus, increasing bolt tensile compressive strength of concrete; fu is tensile strength of the bolt, and Ec
strength can increases the shear resistance but does not influence the is Young’s modulus of the concrete; α1 and α2 are factors for considering
stiffness of the LB-DSC.

1005
J. He et al. Structures 44 (2022) 988–1010

Fig. 22. Effect of bolt diameter: load-slip curves for specimens with concrete slab of (a) 15 MPa and (b) 50 MPa; (c) Relation between shear strength, stiffness and
bolt diameter.

1006
J. He et al. Structures 44 (2022) 988–1010

Fig. 23. Effect of bolt tensile strength: load-slip curves for specimens with a concrete slab of (a) 15 MPa and (b) 50 MPa; (c) Relation between shear strength, stiffness
and bolt tensile strength.

concrete cracking failure and bolt fracture, respectively, and are

Table 5
depicted in Fig. 24, in which the mean value of α1 is almost constant for
Code-based calculation of the shear strength for headed stud connectors.
bolt diameter 16–25 mm, and α2 decreases as bolt diameter (d) in­
Code Expression creases. Thus, for concrete failure, the value of α1 is fitted as a constant
Eurocode 4
√̅̅̅̅̅̅̅̅̅
of 0.30. As for the factor α2, there are several recommendations in the
Pu = 0.29αd2 Ec f c /γv ⩽0.8As fu /γv ;
′

(h ) literature for bolts. Eurocode 3 [67] prescribes the values of α2 = 0.6, for
h h
α = 0.2 sc + 1 for3⩽ sc ⩽4; α = 1 for sc > 4 bolt classes 4.6, 5.6, 8.8, and α2 = 0.5, for bolt classes 4.8, 5.8, 6.8, 10.9,
d d d
AISC360-10
√̅̅̅̅̅̅̅̅̅
when shear plane passes through the threaded portion of the bolt. To
Pu = 0.5As Ec f c ⩽Rg Rp As fu
′

ACI 318–08
√̅̅̅̅ calculate the ultimate strength of post-installed shear connectors, α2 =
Pu = kcp k f c h1.5
ef ⩽φAs fu
′

√̅̅̅̅̅̅̅̅̅ 0.5 was suggested [25]. And α2 = 0.54 was recommend [58] for multiple
AASHTO M16 bolt connectors in prefabricated concrete slab considering group
Pu = ϕsc 0.5As Ec f c ⩽ϕsc As fu
′

effect of bolt connectors. To predict the shear resistance of the high-

√̅̅̅̅̅̅̅̅
GB50017 Pu = 0.43As Ec fc ⩽0.7γAs fu
strength friction-grip bolt (HSFGB), α2 was modified to 0.66 [29].
NOTE: Pu: shear resistance of headed studs; As: cross-sectional area of stud
Based on pushout out tests on assembly of shear connectors, α2 was
shank; fc′ and fc: specified and allowable cylinder compressive strength of con­
crete, respectively; fu: tensile strength of studs; Ec: Young’s modulus of concrete;
determined to be 0.8 [32]. However, from extensive numerical para­
d: diameter of the studs; hsc: height of the stud; Rg, Rp, φ, φsc, γ: partial safety metric analyses [27], it was found that α2 decreases with further in­
factors specified in codes. crease of bolt diameter and the relation of α2 and diameter ratio (dref/d)
can be expressed as power law function. Therefore, the factor α2 should
consider the effect of diameter in Eq. (8) for bolt failure, which can be

1007
J. He et al. Structures 44 (2022) 988–1010

Fig. 24. The determination of the ratio: (a) α1; (b) α2; and (c) fitted curves.

expressed as follows: 7. Conclusions

( )β
d
α2 = αb ref ⩽1 for d⩾16 mm (9) A lockbolt demountable shear connector (denoted as LB-DSC) is
d proposed in this paper for application in sustainable steel–concrete
structures. The LB-DSC was assessed experimentally through four
where dref is representing a reference bolt diameter, dref = 20 mm in the
Eurocode 4 pushout tests. A nonlinear numerical model using Abaqus
present study, αb is the shear resistance factor for the bolts with refer­
was also calibrated through comparison with experimental results and
ence diameter, while the exponent β is used to account for the sensitivity
then used to conduct a parametric study. On the basis of the experi­
of the diameter variation. Thus, for bolt fracture failure, the value of αb
mental and numerical results, the following conclusions can be
and β are fitted as 0.84 and 0.84, respectively.
obtained:
Fig. 25 compares the shear resistance of the LBDSC predicted by Eqs.
(8) and (9) (Pu_cal) to those obtained from FEM (Pu_FEM.) and pushout test
• The experimental and numerical results indicate that the LB-DSC
(Pu_test). The average ratio of Pu_cal / Pu_FEM is 0.995, with a standard
with the geometry presented in this paper has high shear resis­
deviation of 0.05, while average ratio of Pu_cal / Pu_test is 1.04, with a
tance (254 kN in average), high stiffness (higher than most DSCs),
standard deviation of 0.13, indicating that Eq. (8) can predict the shear
and slip capacity much higher than the 6 mm required by Eurocode 4
resistance of LB-DSC with acceptable accuracy. It should be noted that
(11.5 mm in average). Thus, the LB-DSC can provide an effective
the use of the proposed parameter (α2) is limited to high strength bolts
shear connection between a steel section and the concrete slab in a
with diameter d ≥ 16 mm, because a larger factor α2 would be obtained
steel–concrete composite beams and could be used in composite
for smaller bolt diameters. Moreover, bolts with diameter smaller than
beams with partial shear connection.
16 mm are not recommended in the proposed LB-DSCs for application in
• The lockbolt geometry effectively prevents sudden slip before SLS
large span steel–concrete structures.
resulting in high initial stiffness. This is clearly demonstrated in the

1008
J. He et al. Structures 44 (2022) 988–1010

especially when LB-DSC is embedded in high compressive strength

concrete slab.
• The shear resistance of the LB-DSC can be reliably predicted by the
proposed design equations with calibrated parameters according to
the numerical and experimental results, provided that high strength
bolt with a diameter of at least 16 mm is used.

Declaration of Competing Interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence
the work reported in this paper.

Acknowledgments

The authors wish to express their gratitude to the financial support

provided by Horizon 2020 - Marie Skłodowska - Curie Individual
Fellowship of European Commission (REUSE: 793787), the National
Nature Science Foundations of China (51978081, 5211101838), and the
Natural Science Foundation of Hunan Province, China (2021JJ30712,
2022JJ10049). The authors are grateful to the technicians at the Heavy
Structures Lab at Heriot-Watt University, Edinburgh, UK.

References

[1] Hardy A. State of the Nation 2020: Infrastructure and the 2050 net-zero target.
Institution of Civil Engineers; 2021 (n.d.). https://www.ice.org.uk/news-and-
insight/policy/son-2020-infrastructure-and-2050-net-zero-target (accessed April
27, 2021).
[2] Galvin R, Healy N. The Green New Deal in the United States: what it is and how to
pay for it. Energy Res Social Sci 2020;67:101529.
[3] Arsova S, Corpakis D, Genovese A, Ketikidis PH. The EU green deal: spreading or
concentrating prosperity? Resour Conserv Recycl 2021;171(1):105637.
[4] IEA (International Energy Association). Tracking Industrial Energy Efficiency and
CO2 Emissions. Paris: IEA; 2007.
[5] Brambilla G, Lavagna M, Vasdravellis G, Castiglioni CA. Environmental benefits
arising from demountable steel-concrete composite floor systems in buildings. Res
Conserv Recycl 2019;141(2):133–42.
[6] Brozzetti J. Design development of steel-concrete composite bridges in France.
J Constr Steel Res 2000;55(1–3):229–43.
[7] Galambos TV. Recent research and design developments in steel and composite
steel-concrete structures in USA. J Constr Steel Res 2000;55(1–3):289–303.
[8] Hanswille G, Sedlacek G. Steel and Composite Bridges in Germany: State of the Art.
Wuppertal: University of Wuppertal, Institute for Steel and Composite Structures;
Fig. 25. The comparison of predicted and (a) FEM analytical (b) tested 2007.
shear strength. [9] Kuhlmann, U., Braun, B., Feldmann, M. COMBRI Design Manual - Part II: State of-
the-Art and Conceptual Design of Steel and Composite Bridges, Germany; 2008.
[10] He J, Liu Y, Chen A, Yoda T. Experimental study on inelastic mechanical behaviour
comparisons with previous bolted DSCs. In addition, it renders the of composite girders under hogging moment. J Constr Steel Res 2010;66(1):37–52.
[11] Lam D, El-Lobody E. Behaviour of headed stud shear connectors in composite
connector less sensitive to construction tolerances.
beam. J Struct Eng 2005;131(1):96–107.
• The LBDSCs with or without infilled grout exhibited ductile shear [12] Colajanni P, La Mendola L, Monaco A. Stress transfer mechanism investigation in
failure including fracture of bolt shank and local crushing of hybrid steel trussed–concrete beams by push-out tests. J Constr Steel Res 2014;95:
concrete. 56–70.
[13] He J, Liu Y, Pei B. Experimental study of the steel-concrete connection in hybrid
• The infilled grout in the steel tube results in a significant increase in cable-stayed bridges. J Perform Constr Facil 2014;28(3):559–70.
the initial stiffness of the LBDSC, and it has an insignificant effect on [14] He J, Lin Z, Liu Y, Xu X, Xin H, Wang S. Shear stiffness of headed studs on
its shear resistance. In addition, the compressive strength of the grout structural behaviours of steel-concrete composite girders. Steel Comp Struct 2020;
36(5):553–68.
does not affect the shear resistance considerably. Therefore, grouting [15] Lin W, Yoda T, Taniguchi N, Kasano H, He J. Mechanical performance of steel-
the tube without special consideration of the grout strength is concrete composite beams subjected to a hogging moment. J Struct Eng ASCE
allowed. 2014;140(1):04013031.
[16] An L, Cederwall K. Push-out tests on studs in high strength and normal strength
• The standard pushout tests presented in this study revealed a dif­ concrete. J Constr Steel Res 1996;36(1):15–29.
ference with recent horizontal setup pushout tests on the same [17] Shim C, Lee P, Yoon T. Static behaviour of large stud shear connectors. Eng Struct
connector and highlighted the importance of test setup. The sepa­ 2004;26(12):1853–60.
[18] Lee P, Shim C, Chang P. Static and fatigue behaviour of large stud shear connectors
ration capacity in the present tests only accounts for 3–18 % of the for steel-concrete composite bridges. J Constr Steel Res 2005;61(9):1270–85.
corresponding slip, and the separation induced tension force is less [19] Pallarés L, Hajjar JF. Headed steel stud anchors in composite structures, Part I:
than 15 % of the applied shear force, which can be ignored. Shear. J Constr Steel Res 2010;66(2):198–212.
[20] Xue D, Liu Y, Yu Z, He J. Static behaviour of multi-stud shear connectors for
• Increasing slab concrete compressive strength and tube thickness can
steel–concrete composite bridge. J Constr Steel Res 2012;74:1–7.
enhance shear stiffness and resistance to a moderate degree, while [21] Lin Z, Liu Y, He J. Behaviour of stud connectors under combined shear and tension
increasing bolt diameter can significantly improve the shear resis­ loads. Eng Struct 2014;81:362–76.
tance and stiffness. On the other hand, only shear resistance can be [22] Dallam LN. Pushout tests with high strength bolt shear connectors. Rep. 68-7, Dept.
of Civil Engineering, Univ. of MissouriColumbia, Columbia, MO; 1968.
obviously improved as the increasing of bolt tensile strength,

1009
J. He et al. Structures 44 (2022) 988–1010

[23] Dallam LN, Harpster JL. Composite beams tests with high-strength bolt shear [45] Suwaed ASH, Karavasilis TL. Removable shear connector for steel-concrete
connectors. Rep. 68-3, Dept. of Civil Engineering, Univ. of Missouri-Columbia, composite bridges. Steel Comp Struct 2018;29(1):107–23.
Columbia, MO. 1968. [46] Suwaed ASH, Karavasilis TL. Demountable steel-concrete composite beam with
[24] Marshall WT, Nelson HM, Banerjee HK. An experimental study of the use of high- full-interaction and low degree of shear connection. J Constr Steel Res 2020;171:
strength friction-grip bolts as shear connectors in composite beams. Struct Eng 106152.
1971;49(4):171–8. [47] Suwaed ASH, Karavasilis TL. Novel demountable shear connector for accelerated
[25] Kwon G, Engelhardt MD, Klinger RE. Behavior of post-installed shear connectors disassembly, repair, or replacement of precast steel-concrete composite bridges.
under static and fatigue loading. J Constr Steel Res 2010;66:532–41. J Bridg Eng 2017;22(9):04017052.
[26] Kwon G, Engelhardt MD, Klingner RE. Experimental behavior of bridge beams [48] He J, Suwaed ASH, Vasdravellis G, Wang S. Shear performance of a novel
retrofitted with post installed shear connectors. J Bridge Eng 2011;16(4):536–45. demountable connector for reusable steel-concrete composite structures. Life-Cycle
[27] Pavlović, M. Resistance of Bolted Shear Connectors in Prefabricated Steel-Concrete Civil Engineering: Innovation, Theory and Practice – Proceedings of the 7th
Composite Decks, Ph.D. Thesis; University of Belgrade, Faculty of Civil International Symposium on Life-Cycle Civil Engineering, IALCCE 2020, 2020, pp.
Engineering, Belgrade, Serbia;. 2013. 775–781.
[28] Pavlovic M, Markovic Z, Veljkovic M, Buđevac D. Bolted shear connectors vs. [49] Suwaed ASH, He J, Vasdravellis G. Experimental and numerical evaluation of a
headed studs behaviour in push-out tests. J Constr Steel Res 2013;88:134–49. welded demountable shear connector through horizontal pushout tests. J Struct
[29] Liu X, Bradford M, Lee M. Behavior of high-strength friction-grip bolted shear Eng 2022;148(2):04021274.
connectors in sustainable composite beams. J Struct Eng ASCE 2015;141: [50] He J, Suwaed ASH, Vasdravellis G. Horizontal pushout tests and parametric
04014149. analyses of a locking-bolt demountable shear connector. Structures 2022;35:
[30] Ataei A, Bradford MA, Liu X. Experimental study of composite beams having a 667–83.
precast geopolymer concrete slab and deconstructable bolted shear connectors. [51] Eurocode 4. Design of composite steel and concrete structures. Part 1–1. General
Eng Struct 2016;2016(114):1–13. rules and rules for buildings. EN 1994-1-1. Brussels, Belgium: BSI; 2004.
[31] Ataei A, Zeynalian M, Yazdi Y. Cyclic behaviour of bolted shear connectors in steel- [52] BSI (British Standards Institution). Testing hardened concrete, Part 3: Compressive
concrete composite beams. Eng Struct 2019;198(1):109455. strength of test specimens. BS EN 12390-6. London: BSI; 2009.
[32] Yang F, Liu Y, Jiang Z, Xin H. Shear performance of a novel demountable steel- [53] ASTM. Standard test methods for tension testing of metallic materials. ASTM E8/
concrete bolted connector under static push-out tests. Eng Struct 2018;160: E8M. West Conshohocken, PA: ASTM; 2011.
133–46. [54] ABAQUS (2018) [Computer software]. Providence, RI, Dassault Systemès Simulia.
[33] Lam D, Saveri E. Shear capacity of demountable shear connectors. Proceedings of [55] He J, Suwaed ASH, Vasdravellis G, Wang S. Standard pushout tests and design rules
the 10th International Conference on Advances in Steel Concrete Composite and for a bolted-welded hybrid demountable shear connector. J Struct Eng 2022;148
Hybrid Structures. 2012. (8):04022097.
[34] Moynihan MC, Allwood JM. Viability and performance of demountable composite [56] fib (Fédération internationale du béton). 2010. First complete draft. Vol. 1. Comité
connectors. J Constr Steel Res 2014;88:47–56. Euro-International du Beton. In Model code 2010, secretariat permanent.
[35] Dai X, Lam D, Saveri E. Effect of concrete strength and stud collar size to shear Lausanne, Switzerland.
capacity of demountable shear connectors. J Struct Eng 2015;141(11):04015025. [57] Birtel, V., Mark, P. Parameterized finite element modelling of RC beam shear
[36] Rehman N, Lam D, Dai X, Ashour AF. Experimental study on demountable shear failure. The 19th Annual International ABAQUS Users’ Conference, ABAQUS Inc,
connectors in composite slabs with profiled decking. J Constr Steel Res 2016;122: Boston, USA, 2006. pp. 95–108.
178–89. [58] Yang T, Liu S, Qin B, Liu Y. Experimental study on multi-bolt shear connectors of
[37] Pathirana SW, Uy B, Mirza O, Zhu X. Strengthening of existing composite steel- prefabricated steel-concrete composite beams. J Constr Steel Res 2020;173:
concrete beams utilising bolted shear connectors and welded studs. J Constr Steel 106260.
Res 2015;114:417–30. [59] Rehman N. Behaviour of demountable shear connectors in composite structures.
[38] Pathirana SW, Uy B, Mirza O, Zhu X. Flexural behaviour of composite steel- UK: University of Bradford; 2017. Ph. D dissertation.
concrete utilizing blind bolt shear connectors. Eng Struct 2016;114:181–94. [60] Pathirana SW, Uy B, Mirza O, Zhu X. Bolted and welded connectors for the
[39] Ban H, Uy B, Pathirana SW, Henderson I, Mirza O, Zhu X. Time-dependent rehabilitation of composite beams. J Constr Steel Res 2016;125:61–73.
behaviour of composite beams with blind bolts under sustained loads. J Constr [61] Zhang Y, Chen B, Liu A, Pi Y, Zhang J, Wang Y, et al. Experimental study on shear
Steel Res 2015;112:196–207. behavior of high strength bolt connection in prefabricated steel-concrete composite
[40] Henderson IEJ, Zhu XQ, Uy B, Mirza O. Dynamic behaviour of steel-concrete beam. Compos B Eng 2019;159:481–9.
composite beams with different types of shear connectors. Part I: Experimental [62] Suwaed ASH. Development of novel demountable shear connectors for precast
study. Eng Struct 2015;103:298–307. steel-concrete composite bridges. Ph. D dissertation. UK: University of Warwick;
[41] Henderson IEJ, Zhu XQ, Uy B, Mirza O. Dynamic behaviour of steel-concrete 2017.
composite beams with different types of shear connectors. Part II: Modelling and [63] American Institute of Steel Construction (AISC). Specification for Structural Steel
comparison. Eng Struct 2015;103:308–17. Buildings. Standard No. ANSI/AISC 360–10. USA. 2010.
[42] Feidaki E, Vasdravellis G, He J, Wang S. Steel-yielding demountable shear [64] American Concrete Institute (ACI). Building code Requirements for Structural
connector for composite floors with precast hollow-core slab units. J Struct Eng Concrete. Standard No. ACI 318-08. USA. 2008.
2019;145(8):04019076. [65] GB50017. Code for design of steel structures. Beijing; 2003.
[43] He J, Vasdravellis G, Wang S. Circular perforated steel yielding demountable shear [66] AASHTO. AASHTO LRFD bridge design specifications, Washington, DC; 2014.
connector for sustainable precast composite floors. Steel Comp Struct 2021;39(6): [67] BSI (British Standards Institution). EN1993–1-8: Eurocode 3: Design of steel
701–21. structures. Part 1–8: Design of joints, European Committee for standardization
[44] Kozma A, Odenbreit C, Braun M, Veljkovic M, Nijgh M. Push-out tests on (CEN), Brussels Belgium; 2005.
demountable shear connectors of steel-concrete composite structures. Structures
2019;21:45–54.

1010