Push-out behavior of headed stud shear connectors in FRP-UHPC-steel double-skin tubular members (DSTMs)

Abstract

The FRP-concrete-steel double-skin tubular members (DSTMs) have attracted worldwide research due to their excellent mechanical performance. However, the limitation of these DSTMs lies in the low strength and crack resistance of normal concrete. To further improve their mechanical performance and durability, a solution by replacing the normal concrete in DSTMs with ultra-high performance concrete (UHPC) is proposed. This research focused on the shear behavior of headed stud shear connectors in DSTMs, a crucial component in their overall performance. Through rigorous testing of sixteen push-out specimens, the effects of parameters including the compressive strength of concrete/UHPC, the steel fiber volume ratio in UHPC, and the diameter and layout of headed studs on the interfacial shear behavior of DSTMs were investigated. The test results not only proved that the application of UHPC in DSTMs resulted in a different failure mode and a higher interfacial shear performance, but also revealed that the investigated parameters obviously effected the interfacial shear load-slip responses of DSTM push-out specimens. Subsequently, the shear capacities of push-out specimens predicted by four of the existing theoretical models were compared with the test results. The comparison results indicated that a more accurate model for predicting the interfacial shear capacities of headed stud shear connectors in FRP-UHPC-steel DSTM is needed in future research.

Keywords

FRP-concrete-steel double-skin tubular members (DSTMs)push-out test UHPC headed stud shear connector shear capacity

Introduction

Fiber reinforced polymer (FRP) is a light-weight, high-strength and durable material. Over the past two decades, apart from the successful applications in the retrofit or strengthening of existing structures (Hollaway and Teng 2008; Teng et al., 2012), the combined use of FRP with other construction materials to form composite structures is also a valuable research direction (Teng et al., 2007). FRP-concrete-steel double-skin tubular member (hereafter refer to as “DSTM” for simplicity) proposed by Teng et al. (2004, 2007) is one of the most famous novel FRP composite members, due to its high strength and ductility, excellent durability, lightweight and ease of construction. A DSTM consists of an outer FRP tube and an inner steel tube, with concrete filled in between. Not only the two tubes can be either circular or rectangular, but also they can adopt different shapes. Take the circular section as an example, the two tubes may be concentrically placed as columns (DSTCs, as shown in Figure 1(a)), or eccentrically placed as beams (DSTBs, as shown in Figure 1(b)). Currently, numerous studies mainly delved into the compressive behavior of DSTCs, while there were only a few studies focusing on the flexural behavior of DSTBs (e.g. Mo et al., 2022; Zhao et al., 2016). The flexural tests confirmed that DSTBs possess a ductile performance. However, these tests also revealed that shear connectors need to be set between concrete and steel tube in DSTBs to eliminate significant slip between them (e.g. Idris and Ozbakkaloglu 2014; Yu et al., 2006).

Figure 1.

Typical sections of circular double-skin tubular members.

Although the conventional FRP-concrete-steel DSTBs have been proven to possess high strength and ductility performance (Idris and Ozbakkaloglu 2014; Mo et al., 2022; Yu et al., 2006; Zhao et al., 2016), the low strength and crack resistance of normal concrete has been identified as a key factor that limited further enhancement of the flexural performance of DSTBs. Ultra-high performance concrete (UHPC), known for its impressive strength-to-weight ratio and exceptional durability (Hung et al., 2021; Kang et al., 2010; Kim et al., 2015; Magureanu et al., 2012; Russell and Graybeal 2013; Xiao et al., 2022; Xu and Wille 2015; Xue et al., 2020; Yoo et al., 2013; Zhou et al., 2018), was adopted to replace the normal concrete in DSTBs further to enhance the load-bearing and deformation capabilities. In such FRP-UHPC-steel DSTBs, the shear connectors on the steel-UHPC interfaces are indispensable, since they are essential for ensuring the UHPC-steel interfacial shear bond strength. Headed studs are the most common form of the shear connectors, due to their low cost and ease of installation (Collings, 2005; Johnson, 1994; Nie, 2011; Oehlers and Bradford 1999).

Extensive studies on the shear behaviors of headed studs in the steel-UHPC interfaces by the traditional push-out test method in Eurocode-4 (2004), as shown in Figure 2. The investigated parameters of these studies (e.g. Cao et al., 2017; Deng et al., 2024; Ding et al., 2021, 2023; Duan et al., 2022; Gao et al., 2020; Hu et al., 2020; Kim et al., 2013, 2015; Kruszewski et al., 2018; Lai et al., 2024; Li et al., 2021, 2023a, 2023b; Liu et al., 2019; Qi et al., 2019; Semendary et al., 2022; Sun et al., 2017; Wang et al., 2017, 2018, 2019; Wei et al., 2022; Wu et al., 2022; Xu et al., 2022a, 2022b; Zhao et al., 2021) include the aspect ratio, diameter, and layout of stud as well as the strength, thickness, and steel fiber content of UHPC. Some studies (e.g. Cao et al., 2017; Ding et al., 2021; Hu et al., 2020; Qi et al., 2019; Wang et al., 2018, 2019; Zhao et al., 2021) showed that the push-out specimens have a similar failure mode in which the headed studs fracture near the stud roots. In addition, following phenomenon and conclusions were also drawn from these studies: the load-slip curves of specimens consist of three stage (e.g. Duan et al., 2022; Li et al., 2023a; Wei et al., 2022); short studs could develop full strength in UHPC layer (e.g. Cao et al., 2017; Deng et al., 2024; Liu et al., 2019); reducing stud spacing below the a threshold could result in a decline in the shear capacity (e.g. Ding et al., 2021; Li et al., 2023b); the change of stud aspect ratio in a certain range could affect the shear strength of studs (e.g. Kim et al., 2013; Zhao et al., 2021), while more studies believed that the effect of aspect ratio on the shear strength of the headed studs was limited within the studied parameters (e.g. Gao et al., 2020; Kim et al., 2015; Lai et al., 2024; Li et al., 2023b; Qi et al., 2019; Wang et al., 2019); the thickness of UHPC layer showed limited effect on shear strength of the headed studs (e.g. Kim et al., 2013; Wang et al., 2019); the shear capacity of the specimens increased with the diameter of headed studs (e.g. Duan et al., 2022; Gao et al., 2020; Li et al., 2021); studs in UHPC exhibit greater shear strength and stiffness, but lower ductility compared to those in traditional concrete (e.g. Wu et al., 2022); the shear stiffness and shear capacity of grouped studs were lower than those of single studs (e.g. Xu et al., 2022a); aside from a minor amount of localized crushing at the stud root in the UHPC slabs, the overall integrity of the slabs remains intact, with no cracking observed (e.g. Ding et al., 2023; Sun et al., 2017). Kruszewski et al. (2018) noted that as the compressive strength of UHPC increases from 114 MPa to 193 MPa, there is a 13% enhancement in the shear capacity of the studs. Furthermore, the application of UHPC with the use of higher compressive strength results in a reduction of both yield slip and ultimate slip. In the research conducted by Wang et al. (2017), the thickness of the UHPC cover was just 10 mm, significantly lower than the 50 mm mandated by the AASHTO (2012) and also below the 15 mm specified by the GB-50017 (2003) for standard concrete. Throughout the testing, no cracking or spalling was noted on the outer surfaces of the UHPC plates, and the 10 mm cover thickness proved sufficient for the removable shear connectors in steel-UHPC composite beams.

Figure 2.

Traditional push-out specimen.

The research indicates that current push-out tests assessing the shear performance of headed stud shear connectors at steel-UHPC interfaces primarily focus on the diameter and aspect ratio of the studs, as well as the compressive strength of UHPC. However, there have been no reports on experimental studies that utilize the fiber volume ratio of UHPC and the tensile strength of UHPC as research parameters. Moreover, the impact of FRP confinement on the interfacial shear behavior between steel and UHPC was not considered. Against this background, the author adopted a modified push-out test method proposed by Zhao (2017) to examine the shear behavior of headed studs in FRP-UHPC-steel DSTMs, as shown in Figure 3. The failure modes and load-slip curves of push-out specimens were observed and recorded. The influences of the investigated parameters on the interfacial shear performance were investigated and analyzed. Ultimately, the experimental data is utilized to assess and analyze the accuracy and applicability of four representative theoretical models for predicting the shear capacity of headed studs. This study is a pioneering investigation of the shear behavior of headed studs in FRP-UHPC-steel DSTMs. The experimental results obtained from this study can provide a foundation and reference for subsequent experiments, numerical simulations, and other related studies.

Figure 3.

Modified push-out specimen.

Experimental program

Specimen details

All 16 push-out specimens were prepared in eight pairs, with each pair comprising two nominally identical specimens. The test variables included the compressive strength of the concrete, fiber volume ratio, headed stud diameter and layout of studs. The dimensions of all the test specimens were same, as shown in Figure 4. The height, internal diameter and thickness of GFRP tubes are 500 mm, 600 mm and 6 mm respectively, the height, external length of side and thickness of steel tubes are 500 mm, 400 mm and 12 mm respectively. The length of all the headed studs is 60 mm. Except for two specimens with single row headed studs, 16 headed studs were arranged into two rows with a vertical interval of 240 mm except for two specimens with single row headed studs. For the all specimens, there are eight headed studs in each row, and two headed studs on same side were 160 mm apart from each other in the transverse direction.

Figure 4.

Dimensions of specimens (unit: mm).

Table 1 is the specimen details, the name of specimens start with letters “FUC”, “SUC” or “NC” to represent the first batch of UHPC, the second batch of UHPC and normal concrete respectively; followed by a number “1” or “2” representing the number of row of headed studs; followed by a number “0”, “1” or “2” representing the fiber volume ratio of UHPC; followed by a number “13”, “16” or “19” representing the diameter of headed studs; and followed by the last number “1” or “2” differentiating two nominally identical specimens in each pair.

Table 1.

Specimen details.

Specimen	Concrete type	Number of stud row	Fiber volume ratio	Stud diameter (mm)
FUC-1-2-16-1, 2	First batch of UHPC	1	2%	16
FUC-2-2-13-1, 2	First batch of UHPC	2	2%	13
FUC-2-2-16-1, 2	First batch of UHPC	2	2%	16
FUC-2-2-19-1, 2	First batch of UHPC	2	2%	19
FUC-2-1-16-1, 2	First batch of UHPC	2	1%	16
FUC-2-0-16-1, 2	First batch of UHPC	2	0%	16
SUC-2-0-16-1, 2	Second batch of UHPC	2	0%	16
NC-2-0-16-1, 2	Normal concrete	2	——	16

Preparation of specimens

The push-out specimens were prepared in the following procedure: (1) cleaning the outer surface of steel tubes; (2) welding headed studs on the surface of the steel tubes by a technical stud welding machine (Figure 5(a)); (3) applying lubricating oil to the surface of the steel tubes; (4) fixing the inner steel tubes and outer FRP tubes to a base plate (Figure 5(b)); (5) casting UHPC between the two tubes (Figure 5(c)).

Figure 5.

Preparation procedure of the specimens.

Material properties

Two batches of UHPC and normal concrete (NC) used in the present study were provided by China Railway Bridge Science Research Institute, Ltd. The first batch of UHPC consists of premix powder (i.e. cement, silica fume, CaCO₃ powder, fine aggregate), water, super-plasticizer and steel fibers with three different volume ratios of 0%, 1% and 2% (hereafter referred to as UHPC-0%, UHPC-1% and UHPC-2%, respectively). The diameter and length of the steel fibers were 0.2 mm and 13 mm, respectively, and a tensile strength greater than 2500 MPa were adopted. The second batch of UHPC consists of premix powder (i.e. cement, silica fume, quartz powder, fine aggregate), water and super-plasticizer. The normal concrete consists of P.O 42.5 ordinary cement, ordinary river sand, graded crushed stone, water, and water reducer. Table 2 and Table 3 show the mix proportions of UHPC and NC respectively.

Table 2.

Mix proportion of UHPC.

Component	First batch of UHPC-0% (kg/m³)	First batch of UHPC-1% (kg/m³)	First batch of UHPC-2% (kg/m³)	Second batch of UHPC (kg/m³)
Cement	850	850	850	859
Silica fume	175	175	175	183
CaCO₃ powder	102	102	102	——
Quartz powder	——	——	——	116
Fine aggregate	1020	1020	1020	1050
Water	221	221	221	215
Super-plasticizer	21	21	21	17
Steel fiber	0	78.5	157	0

Table 3.

Mix proportion of NC.

Grade	Cement (kg/m³ )	Fine aggregate (kg/m³ )	Coarse aggregate (kg/m³ )	Water (kg/m³ )	Water reducer (kg/m³ )	W/b ratio
C50	408	752	1002	165	7.39	0.40

Twelve 100 mm cubic samples and three 150 mm cubic samples were cast simultaneously with the push-out specimens to test the compressive strength of UHPC and NC, respectively. The tensile strength of concrete were obtained from the direct tensile tests of fifteen dog-bone specimens. The elastic modulus of UHPC and NC were obtained from 12 100 × 100 × 300 mm prism samples and three 150 × 150 × 300 mm prism samples, respectively. All the specimens were cured at room temperature and the material property tests were conducted simultaneously with the push-out tests. The material properties of UHPC and NC were tested according to Chinese standards T/CBMF37-2018/T/CCPA7 (2018) and GB/T-50081 (2019), respectively. The experimental set-ups for testing the strength of concrete are showed in Figure 6. Table 4 shows the result of material property tests.

Figure 6.

Set-ups for testing properties of concrete.

Table 4.

Material properties of concrete.

Concrete type	Fiber volume ratio (%)	Compressive strength (MPa)	Tensile strength (MPa)	Elastic modulus (GPa)
First batch of UHPC-0%	0	133.3	8.0	43.5
First batch of UHPC-1%	1	141.7	8.8	43.9
First batch of UHPC-2%	2	151.7	9.7	45.1
Second batch of UHPC	0	133.7	6.5	43.7
NC	——	54.9	3.2	33.3

The material properties of the headed studs and steel tubes were tested according to ASTM E8 (2016) and GB/T-2975 (2018), respectively. Figure 7 shows the stress-strain curves of headed studs. The detailed material properties of steel components are summarized in Table 5.

Figure 7.

Tensile stress-strain curves of headed studs.

Table 5.

Mechanical properties of steel components.

Component	Elastic modulus (GPa)	Yield strength (MPa)	Tensile strength (MPa)	Total elongation
Steel tube	201	388	502	23.2%
Headed stud	195	——	465	13.5%

The material properties tests of GFRP tubes were conducted according to ASTM D3039/D3039M-17 (2014). The tensile strength, elastic modulus and rupture strain of the GFRP tubes are 787.5 MPa, 45 GPa and 1.75%, respectively.

Test set-up and instrumentation

All push-out specimens were conducted on a loading equipment with a 6000 kN capacity in the structural laboratory of China Railway Bridge Science Research Institute, Ltd. The testing set-ups are shown in Figure 8. Four linear variable displacement transducers (LVDTs) were installed on the inner surface of steel tubes to measure the relative displacement between the steel tubes and the base plate, as shown in Figure 9. Twelve strain gauges were distributed on the external surface of the GFRP tube. Figure 10 shows the layout of strain gauges. There are three rows of strain gauges with a vertical interval of 120 mm and four strain gauges were uniformly distributed in each row.

Figure 8.

Set-ups of push-out test.

Figure 9.

Arrangement of LVDTs.

Figure 10.

Arrangement of strain gauges on GFRP tube.

Prior to formal loading, a preload cycling between 5% and 40% of the predicted maximum load (i.e. the maximum load was predicted using the equations of Eurocode-4 (2004)) was applied for three times. The formal loading was divided into two stages, with the first loading stage applied in force control at a rate of 0.5 kN/s until the load reached 80% of the predicted maximum load, and the second loading stage applied in displacement control at a rate of 0.5 mm/min until the load decreased to 40% of the maximum load.

Test results and discussions

Failure modes

For the push-out specimens with UHPC-2% (i.e. FUC-1-2-16-1,2, FUC-2-2-13-1,2, FUC-2-2-16-1,2 and FUC-2-2-19-1,2), the same failure mode was observed: all the headed studs ruptured at the roots of studs, while there was no crack or crushing observed in UHPC, as shown in Figure 11. For the specimens with UHPC-1% (i.e. FUC-2-1-16-1,2) and specimens with UHPC-0% (i.e. FUC-2-0-16-1,2 and SUC-2-0-16-1,2), not only all the headed studs ruptured at the roots of studs, but also the crushing or cracks of UHPC were observed near the positions of headed studs, as shown in Figure 12. It can be also seen from Figure 12(b) that the UHPC layer severely damaged in specimens with UHPC-0%. The above phenomenon can be explained as the cracking resistance of UHPC was largely improved by the steel fibers. For specimens with UHPC-0%, headed studs can not be effectively confined due to the massive cracking of the UHPC. Therefore, the shear failure of studs was affected by the bending moment. It was found that no cracking, crushing, or splitting was observed on the outer surface of the UHPC, and the steel tube did not exhibit any local buckling. The push-out specimens with normal concrete (i.e. NC-2-0-16-1,2) failed by concrete crushing and rupture of part of the headed studs. The crushed concrete attached on the headed studs can be seen in Figure 13.

Figure 11.

Failure mode 1: rupture of headed studs at roots of specimens with UHPC-2%.

Figure 12.

Failure mode 2: cracks or crushing of UHPC and rupture of headed studs of specimens with UHPC-1% and UHPC-0%.

Figure 13.

Failure mode 3: concrete crushing and rupture of part of headed studs of specimens with NC.

Load-slip responses

The main test results, including the ultimate load of specimens (P_u), the shear capacity of per stud (P_s) and the average value of two companion specimens (P_s,avg), the interfacial slip corresponding to ultimate load (S_u) and the average value of two companion specimens (S_u,avg), the maximum slip (S_max) and the average value of two companion specimens (S_max,avg), the friction between the steel tube and the UHPC layer (F_u), are summarized in Table 6. The P_s is calculated by first subtracting the friction (F_u) and then dividing by the number of studs. The friction is measured by two push-out specimens without studs. However, the P_s of the specimens NC-2-0-16-1,2 were not available because only part of the headed studs ruptured. The maximum slip (S_max) was defined as the slip corresponding to the load dropping to 90% of the P_u, according to the Eurocode-4 (2004).

Table 6.

Key results of push-out tests.

Specimen	P_u (kN)	P_s (kN)	P_s,avg (kN)	S_u (mm)	S_u,avg (mm)	S_max (mm)	S_max,avg (mm)	F_u (kN)
FUC-1-2-16-1	1128.5	95.0	82.3	5.83	4.73	7.19	6.50	368.8
FUC-1-2-16-2	924.9	69.5	82.3	3.63	4.73	5.80	6.50
FUC-2-2-13-1	1387.9	63.7	69.0	3.98	4.29	5.95	6.11
FUC-2-2-13-2	1556.3	74.2	69.0	4.60	4.29	6.26	6.11
FUC-2-2-16-1	1792.4	89.0	92.0	4.65	5.17	7.03	6.93
FUC-2-2-16-2	1888.3	95.0	92.0	5.68	5.17	6.82	6.93
FUC-2-2-19-1	2171.3	112.7	105.7	3.81	3.11	6.54	6.29
FUC-2-2-19-2	1948.7	98.7	105.7	2.41	3.11	6.04	6.29
FUC-2-1-16-1	1772.2	87.7	93.0	6.16	6.50	6.84	7.66
FUC-2-1-16-2	1940.1	98.2	93.0	6.83	6.50	8.48	7.66
FUC-2-0-16-1	1690.0	82.6	82.1	7.10	6.45	7.98	7.59
FUC-2-0-16-2	1674.8	81.6	82.1	5.80	6.45	7.19	7.59
SUC-2-0-16-1	1625.9	78.6	79.5	5.19	5.20	7.44	7.36
SUC-2-0-16-2	1654.8	80.4	79.5	5.21	5.20	7.27	7.36
NC-2-0-16-1	1442.2	——	——	9.11	8.73	10.70	10.18
NC-2-0-16-2	1342.4	——	——	8.35	8.73	9.66	10.18

Figure 14 is the load-slip curves of all tested specimens. The slips were obtained by the average readings of four LVDTs. The load-slip curves can be divided into three phases, including elastic phase, plastic-strengthening phase and descending phase. In the first phase, the slip of the specimens was very small and the load-slip curves increased linearly, indicating that the headed studs had a high shear stiffness at this stage. Then, the load-slip curves entered the plastic-strengthening phase, the slip significantly increased and the shear stiffness of headed studs continuously decreased. After the ultimate load, the headed studs gradually failed and the load-slip curves entered the descending phase until reaching the maximum slip (S_max).

Figure 14.

Load-slip curves of all specimens.

Strain responses and confinement effect of GFRP tubes

The readings of hoop strain gauges on the GFRP tubes of the push-out specimens are shown in Figure 15, in which the tensile strains are defined to be positive. It can be seen from the figures that the hoop strain on the GFRP tubes approached zero during the initial loading stage for all specimens. In the subsequent loading stage, the strain on the GFRP tubes began to increase rapidly at a specific load level, indicating that the concrete near the studs had started to deform and develop cracks. After that, the hoop strain increased continuously even after the load began to decrease, suggesting that the crack width of concrete increased continuously. The minimum and maximum hoop strains are observed in the first and third rows, respectively, indicating that the lower portion of the concrete/UHPC sustained a higher degree of compression. This is attributable to the pressure exerted on the concrete/UHPC layer, which comprises the shear capabity of studs and the friction between the steel tubes and UHPC/normal concrete layer, transmitted from the top to the bottom. Consequently, the closer the area is to the base of the specimens, the greater the shear capabity of studs and friction.

Figure 15.

Typical load-strain curves of GFRP tubes in push-out specimens.

Table 7 summarizes the hoop strain on the GFRP tube when the specimens reach their peak load. The ε _row1, ε _row2 and ε _row3 represent the average values measured by the four strain gauges arranged on the first, second, and third rows of the GFRP tube, respectively. The strain distribution on the GFRP tube is shown in Figure 10. The ε _avg1 is the average value of ε _row1, ε _row2 and ε _row3, and the ε _avg is the average value of ε _avg1 for two nominally identical specimens. The push-out specimens FUC-2-0-16, FUC-2-1-16, and FUC-2-2-16 differ in their steel fiber volume fractions in UHPC, which are 0%, 1%, and 2%, respectively. As indicated in Table 7, an increase in the fiber volume fraction leads to a decrease in the average strain (ε _avg) of the GFRP tube. This occurs because adding steel fibers not only boosts the compressive and tensile strength of UHPC but also enhances its crack resistance, which helps to limit crack propagation within the UHPC. As a result, the confinement effect of the UHPC layer on the studs is improved, which reduces the confinement effect of the GFRP tube. Additionally, it is important to note that the ε _avg of the UHPC specimens is considerably lower than that of the NC specimens, following the same principle.

Table 7.

Hoop strain of GFRP tubes.

Specimen	ε row 1 (με)	ε row 2 (με)	ε row 3 (με)	ε avg 1 (με)	ε avg (με)
FUC-1-2-16-1	4.52	27.79	133.32	55.21	61.15
FUC-1-2-16-2	68.16	92.94	40.15	67.08	61.15
FUC-2-2-13-1	−28.91	−3.98	36.66	1.26	16.48
FUC-2-2-13-2	5.40	12.62	77.07	31.70	16.48
FUC-2-2-16-1	55.69	71.93	100.50	76.04	139.61
FUC-2-2-16-2	134.26	220.77	254.50	203.18	139.61
FUC-2-2-19-1	16.57	64.93	74.69	52.06	74.25
FUC-2-2-19-2	14.50	112.82	162.00	96.44	74.25
FUC-2-1-16-1	196.61	500.83	375.31	357.58	172.44
FUC-2-1-16-2	−118.52	−1.43	81.86	−12.70	172.44
FUC-2-0-16-1	33.93	206.20	327.06	189.06	252.60
FUC-2-0-16-2	60.31	459.65	428.47	316.14	252.60
SUC-2-0-16-1	117.89	461.88	638.69	406.15	353.92
SUC-2-0-16-2	40.38	296.51	568.14	301.68	353.92
NC-2-0-16-1	294.91	548.96	677.02	506.96	525.01
NC-2-0-16-2	300.34	434.78	894.06	543.06	525.01

From Table 7, it can be seen that the average strain (ε _avg1) on the GFRP tubes of the two nominally identical specimens, FUC-2-1-16-1 and FUC-2-1-16-2, are 357.58 με and −12.70 με, respectively. This indicates that the GFRP tube of specimen FUC-2-1-16-1 provided a confining effect, whereas the GFRP tube of specimen FUC-2-1-16-2 did not. However, as shown in Figure 7(e), the peak load of the unconstrained specimen FUC-2-1-16-2 is actually greater than that of FUC-2-1-16-1. Using the same comparison method, it can also be found that in the FUC-1-2-16 group and the FUC-2-2-19 group, specimens with a smaller confining effect exhibited greater load-bearing capacity than those with a larger confining effect. In contrast to the aforementioned situation, specimens in the FUC-2-2-13 group and the FUC-2-2-16 group that exhibited a smaller confining effect demonstrated a lower load-bearing capacity than those with a larger confining effect. Additionally, the average strain (ε _avg1) on the GFRP tubes of the two nominally identical specimens, SUC-2-0-16-1 and SUC-2-0-16-2, are 406.15 με and 301.68 με, respectively. However, as seen in Figure 7(g), the load-slip curves of the two specimens show no significant difference. In summary, the influence of GFRP tubes on the shear performance of studs is completely irregular. Therefore, it can be inferred that the confinement provided by the GFRP tube in this research has no impact on the shear strength of the studs and the load-slip behavior. The reason may be attributed to the UHPC utilized in this study, which not only possesses high strength but also features a relatively thick UHPC cover layer. The constraint provided by the UHPC layer alone is adequate, which explains why the average strain (ε _avg) of the GFRP tubes is varied between 0.0016% and 0.0525%, which was much smaller than the rupture strain obtained from the coupon tests (i.e. about 1.75%).

Effect of stud diameter

Figure 16 presents a comparative analysis of the load-slip curves of specimens FUC-2-2-13-1, 2, FUC-2-2-16-1, 2, and FUC-2-2-19-1, 2. These specimens differ in the diameter of headed studs, specifically 13 mm (FUC-2-2-13-1, 2), 16 mm (FUC-2-2-16-1, 2), and 19 mm (FUC-2-2-19-1, 2). As depicted in Figure 16, an obvious trend emerges: the P_u of the specimens augments with the increase of the studs diameter. During the elastic phase, the load-slip curves of the three sets of specimens exhibit remarkable similarity. However, upon entering the plastic-strengthening phase, notable differences emerge. Specifically, specimens FUC-2-2-19-1, 2 exhibit the smallest slip, followed by specimens FUC-2-2-16-1, 2, while specimens FUC-2-2-13-1, 2 display the largest slip for a given load. According to Table 6, the average shear capacity of single headed stud for specimens FUC-2-2-16-1, 2 and FUC-2-2-19-1, 2 surpasses that of specimens FUC-2-2-13-1, 2 by 33.3% and 53.2%, respectively. The above reveals the significant role played by the diameter of studs in determining the shear performance of these specimens.

Figure 16.

Effect of stud diameter on load-slip curves.

Effect of concrete strength

Figure 17 offers a comparative analysis of the load-slip curves of specimens FUC-2-0-16-1, 2, SUC-2-0-16-1, 2, and NC-2-0-16-1, 2. The three pairs of specimens differ in their concrete types, specifically, first batch of UHPC (FUC-2-0-16-1, 2), second batch of UHPC (SUC-2-0-16-1, 2), and normal concrete (NC-2-0-16-1, 2).

Figure 17.

Effect of concrete strength on load-slip curves.

As evident from Figure 17, the specimens utilizing normal concrete (NC-2-0-16-1, 2) exhibit the lowest ultimate load. Compared to specimens FUC-2-0-16-1, 2 and SUC-2-0-16-1, 2, the average ultimate load of the normal concrete specimens (NC-2-0-16-1, 2) decreased by 17.2% and 15.1%, respectively. Furthermore, Figure 17 and Table 6 reveal that the S_u,avg and S_max,avg of the UHPC specimens (FUC-2-0-16-1, 2 and SUC-2-0-16-1, 2) are significantly less than those of the normal concrete specimens (NC-2-0-16-1, 2). The significant difference in the load-slip curves between NC specimens and UHPC specimens can be attributed to their distinct failure modes. During the loading process of the test, the concrete near the root of the studs in specimens NC-2-0-16 sustained considerable damage, which prevented the studs from being completely destroyed and resulted in increased slip. Consequently, the load-bearing capacity of specimens NC-2-0-16 is lower than that of the UHPC specimens, although the slip is greater. The difference between specimens SUC-2-0-16 and FUC-2-0-16 lies in the tensile strength of the UHPC. As presented in Tables 4 and 6, the tensile strengths of UHPC for specimens SUC-2-0-16 and FUC-2-0-16 are 6.5 MPa and 8.0 MPa, respectively, while the shear capacities of the studs are 79.5 kN and 82.1 kN, respectively. This indicates that the tensile strength of UHPC has a slight effect on the shear performance of the studs.

Effect of fiber volume ratio

Figure 18 shows the load-slip curves of specimens FUC-2-0-16-1, 2, FUC-2-1-16-1, 2, and FUC-2-2-16-1, 2. The three pairs of specimens have the same testing parameters but different fiber volume ratio of UHPC, specifically 0% (i.e. FUC-2-0-16-1, 2), 1% (i.e. FUC-2-1-16-1, 2) and 2% (i.e. FUC-2-2-16-1, 2), respectively. It can be seen from Figure 18 and Table 6 that the average shear capacity per headed stud of specimens FUC-2-1-16-1, 2 and specimens FUC-2-2-16-1, 2 are close and both greater than that of specimens FUC-2-0-16-1, 2. As illustrated in Figures 11 and 12, the crack control capability of UHPC in specimens FUC-2-0-16 is significantly lower than that of specimens FUC-2-1-16 and FUC-2-2-16. Consequently, for specimens with UHPC-0%, headed studs can not be effectively confined due to the massive cracking of the UHPC, leading to shear failure of studs was affected by the bending moment. This is the primary reason why the shear capacity of the studs in specimen FUC-2-0-16 is lower than that of specimens FUC-2-1-16 and FUC-2-2-16.

Figure 18.

Effect of fiber volume ratio on load-slip curves.

Effect of stud layout

Figure 19 provides a comparative analysis of the load-slip curves for specimens FUC-1-2-16-1, 2 and FUC-2-2-16-1, 2. The two pairs of specimens feature varying numbers of headed stud rows, specifically single row for FUC-1-2-16-1, 2 and two rows for FUC-2-2-16-1, 2. As depicted in Figure 19, the load-slip curves of all four specimens are almost the same until the load reaches approximately 700 kN, which corresponds to the onset of the plastic-strengthening phase for specimens FUC-1-2-16-1, 2. According to Table 6, the average shear capacity per headed stud in specimens FUC-2-2-16-1, 2 exceeds that of FUC-1-2-16-1, 2 by 11.8%. However, the shear capacity per headed stud of specimen FUC-1-2-16-1 is strikingly similar to the average shear capacity per headed stud of specimens FUC-2-2-16-1, 2. Furthermore, the shear capacity per headed stud of specimen FUC-1-2-16-2 is notably lower, specifically 26.8% less than that of specimen FUC-1-2-16-1, likely attributed to the substandard welding quality of individual headed studs in the specimen FUC-1-2-16-2. In terms of interfacial slip, the (S_u,avg) and the (S_max,avg) of specimens FUC-1-2-16-1, 2 are comparable to but slightly smaller than those of specimens FUC-2-2-16-1, 2.

Figure 19.

Effect of stud layout on load-slip curves.

Shear capacity of headed studs

The shear capacity of headed studs is an important index to estimate the mechanical property of shear connectors. Up to now, many formulas for calculating the shear capacity of headed studs have been proposed. In the present paper, four of the representative calculation formulas are selected for comparison with the test results.

Theoretical prediction of shear capacity

In Eurocode-4 (2004), the shear capacity of single stud is given by the following equation:

P_{u} = 0.29 α d^{2} \sqrt{E_{c} {f_{c}}^{'}} / γ_{v} \leq 0.8 A_{s} f_{u} / γ_{v}

(1)

where P_u is the shear capacity of single stud; d is the diameter of the stud shank; E_c is the elastic modulus of concrete; f_c’ is the cylinder compressive strength of concrete; A_s is the cross-sectional area of the stud shank; f_u is the tensile strength of the stud; γ_v is the design safety factor, taken as 1.25; α is a factor determined by equation (2), where h is the height of stud.

α = 0.2 (h / d + 1) \leq 1

(2)

In AASHTO (2012), the stud shear capacity is defined as:

P_{u} = ϕ_{sc} 0.5 A_{s} \sqrt{E_{c} {f_{c}}^{'}} \leq ϕ_{sc} A_{s} f_{u}

(3)

where ϕ_sc is the design safety factor and equals to 0.85.

Zhao (2017) pointed out that existing calculation methods of the shear capacity for studs in traditional push-out tests were unsuitable for studs in DSTMs. Therefore, the novel shear capacity equation of stud was proposed as:

P_{u} = 10.8 d^{1.29} {f_{c}^{'}}^{0.35} {f_{u}}^{0.65}

(4)

Since headed studs in UHPC have different mechanical behavior compared with studs in NC, Kruszewski et al. (2018) proposed the calculated model for estimating the shear capacity of headed studs in UHPC based on the model of Doinghaus et al. (2003), as shown in equation (5). In this model, the effect of weld collars is considered.

P_{u} = A_{s} f_{u} + 0.16 η f_{cu} d^{2}

(5)

where f_cu is the compressive strength of UHPC; η is the shear capacity enhancement coefficient caused by the weld collar shown by equation (6), where β is taken as 0.0119.

η = β f_{cu} - 0.983

(6)

Comparison of shear capacity

The shear capacity of headed studs predicted by different theoretical methods are compared with experimental results, as shown in Table 8. For consistency, the safety factors γ_v and ϕ_sc in equation (1) and (3) were set to 1. The values in parentheses represent the predicted shear capacity of the studs in the event of concrete failure.

Table 8.

Comparison the predicted values with tested values of shear capacity of headed stud.

Specimen	P_s,avg (kN)	P_Eucode (kN)	P_Eucode/P_s,avg	P_AASHTO (kN)	P_AASHTO/P_s,avg	Equation (4)(kN)	P_{Equation (4)}/P_s,avg	Equation (5)(kN)	P_{Equation (5)}/Ps,avg
FUC-1-2-16	82.3	74.8 (184.5)	0.91	93.4 (262.8)	1.13	115.3	1.40	96.6	1.17
FUC-2-2-13	69.0	49.4 (128.2)	0.72	61.7 (173.5)	0.89	88.2	1.28	63.7	0.92
FUC-2-2-16	92.0	74.8 (184.4)	0.81	93.4 (262.8)	1.02	115.3	1.25	96.6	1.05
FUC-2-2-19	105.7	105.4 (227.3)	1.00	131.8 (370.6)	1.25	143.9	1.36	136.2	1.29
FUC-2-1-16	93.0	74.8 (175.9)	0.80	93.4 (250.6)	1.00	108.8	1.17	95.0	1.02
FUC-2-0-16	82.1	74.8 (169.8)	0.91	93.4 (242.0)	1.14	107.6	1.31	94.8	1.15
SUC-2-0-16	79.5	74.8 (170.5)	0.94	93.4 (242.9)	1.17	109.2	1.37	95.1	1.20
Mean			0.87		1.09		1.31		1.11
Standard deviation			0.09		0.11		0.07		0.12
Coefficient of variation			0.10		0.10		0.05		0.11

As depicted in Table 8, the shear capacity of headed stud predicted by Eurocode-4 (2004) is generally lower than the tested values, except for specimens FUC-2-2-19. The deviation between the tested values and the predicted values increases with the decrease of the studs diameter. Among them, the average predicted value of specimens FUC-2-2-13 is 28.4% lower than the tested value, but the average predicted value of specimens FUC-2-2-19 is very close to the tested value. Except for the specimens FUC-2-2-13, the shear capacity of single stud predicted by AASHTO (2012) is generally higher than the tested values, as shown in Table 8. The predict-to-test ratio increases with the increase of studs diameter. The predicted results by AASHTO (2012) underestimates the shear capacity of headed stud for specimens FUC-2-2-13 by 10.6%, but overestimates that for specimens FUC-2-2-19 by 24.7%. Zhao’s (2017) model overestimates the headed stud shear capacity for all specimens. The main reason is that the influence of concrete strength and stud strength were considered simultaneously for calculating the shear capacity of headed studs in Zhao’s (2017) model, while it was found from the present study that all the tested specimens with UHPC failed by headed stud rupture rather than concrete failure. Therefore, only the stud strength need to be considered for calculating the shear capacity of headed studs embedded in UHPC. When comparing the predicted values from Kruszewski et al.'s (2018) model to the test results, the trend closely resembles the comparative analysis between AASHTO’s (2012) estimates and the tested values, as evident in Table 8.

In summary, various existing theoretical methods for calculating the shear capacity of studs were deemed inappropriate for studs embedded in double-skin tubular members (DSTMs) with an UHPC layer. Consequently, there is a need to propose a novel shear capacity equation tailored specifically for headed studs. Furthermore, the present study reveals that the incorporation of steel fibers in UHPC significantly affects the shear behavior of studs. Therefore, this critical factor should be taken into account in future research.

Conclusions

The shear behavior of headed stud in FRP-UHPC-steel DSTMs was investigated by push-out tests of 16 specimens. The effects of stud diameter, concrete strength, steel fiber volume ratio and stud layout on the behaviors of headed studs were investigated. The following conclusions can be drawn:

(1) The failure modes of push-out specimens with UHPC and normal concrete are different, as the former mainly failed by rupture of headed studs (although some cracks were observed in UHPC for specimens with UHPC-0%), while the latter failed by concrete crushing and rupture of headed studs. The crack resistance of UHPC increased significantly with the increase of steel fiber volume ratio.

(2) The load-slip curves can be divided into three distinct phases: the elastic phase, the plastic-strengthening phase, and the descending phase. During the elastic phase, the load-slip curves exhibited a linear increase. As the curves transitioned into the plastic-strengthening phase, the slip increased significantly, while the shear stiffness of the headed studs continuously decreased. When the headed studs gradually failed, the load-slip curves began to enter the descending phase.

(3) In the initial loading stage, the hoop strain on the GFRP tubes approached zero. Subsequently, the strain on the GFRP tubes began to increase rapidly at a specific load level due to the deformation and cracking of the UHPC. An increase in the fiber volume fraction and concrete strength leads to a decrease in the hoop strain of the GFRP tube. The hoop strain on the GFRP tube was significantly smaller than the rupture strain of the GFRP. By comparing and analyzing the strain and load-slip curves of nominally identical specimens, it can be inferred that the confinement provided by the GFRP tube in this study has no effect on the shear strength of the studs or the load-slip behavior.

(4) Stud diameter and concrete strength have prominent influence on the load-slip responses of specimens, the ultimate load of specimens increased with the increase of stud diameter and concrete strength, and the S_u and S_max decreased with the increase of concrete strength. In addition, the number of rows of headed studs and tensile strength of UHPC showed less effects on the load-slip responses of specimens.

(5) The shear capacities of test specimens predicted by four of the representative theoretical formulas were compared with the test data. The predictions for the shear capacity of the studs, based on Eurocode-4 (2004) and AASHTO (2012), were the closest to the experimental findings. However, the Eurocode 4 formula tended to underestimate the shear capacity, while the AASHTO’s (2012) formula tended to overestimate it. Furthermore, neither of these methods considered the effect of steel fibers, highlighting the necessity for a more precise approach to predict the shear capacity of studs in UHPC.

(6) This research is a groundbreaking examination of the shear performance of headed studs in FRP-UHPC-steel DSTMs. The experimental findings from this study can serve as a basis and reference for future experiments, numerical analyses, and other related investigations. In this experiment, the primary reason the FRP tube was unable to fully exert its confinement effect may be attributed to the excessive thickness of the UHPC cover layer. This issue will be a focal point in subsequent research.

Footnotes

ORCID iDs

Yan Ke

Shishun Zhang

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors are grateful for the financial support received from the National Natural Science Foundation of China (Project No. 52078231) and the Key Research and Development Program of Hubei Province of China (Project No. 2021BCA150).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

AASHTO (2012) AASHTO LRFD Bridge Design Specifications. Washington, D.C: American Association of State Highway and Transportation Officials.

ASTM D3039/D3039M-17 (2014) Standard Test Method for Tensile Properties of Polymer Matrix Composite Materials. West Conshohocken. PA: American Society of Testing Materials.

ASTM E8 (2016) Standard Test Methods for Tension Testing of Metallic Materials. West Conshohocken, PA: American Society of Testing Materials.

Cao

Shao

Deng

, et al. (2017) Static and fatigue behavior of short-headed studs embedded in a thin ultrahigh-performance concrete layer. Journal of Bridge Engineering 22(5): 04017005.

Collings

(2005) Steel Concrete Composite Bridges. London, England: Thomas Telford.

Deng

Huang

Xue

(2024) Experimental study on shear behavior of short studs in Steel-UHPC light composite structure. Journal of Tianjin University (Science and Technology) 57(1): 42–52, (in Chinese).

Ding

Zhu

Kang

, et al. (2021) Experimental study on grouped stud shear connectors in precast steel-UHPC composite bridge. Engineering Structures 242: 112479.

Ding

Zhu

Shi

(2023) Performance of grouped stud connectors in precast steel-UHPC composite bridges under combined shear and tension loads. Engineering Structures 277: 115470.

Doinghaus

Goralski

Will

(2003) Design rules for composite structures with high performance steel and high performance concrete. In: Proceedings of the International Conference United Engineering Foundation on High Performance Materials in Bridges. Reston, VA: ASCE, 139–149.

10.

Duan

Zou

Bao

, et al. (2022) Experimental investigation of headed studs in steel-ultra-high performance concrete (UHPC) composite sections. Engineering Structures 270: 114875.

11.

ECS Eurocode-4 (2004) Design of Composite Steel and Concrete Structures. European Committee for Standardization. Brussels, Belgium: European Committee for Standardization.

12.

Gao

Wang

Yan

(2020) Experimental studies of headed stud shear connectors in UHPC Steel composite slabs. Structural Engineering and Mechanics 74(5): 657–670.

13.

GB-50017 (2003) Code for Design of Steel Structure. Beijing, China: China Planning Press. (in Chinese).

14.

GB/T-2975 (2018) Steel Mechanical Properties and Process Performance Test Sampling Regulations. Beijing, China: Standards Press of China. (in Chinese).

15.

GB/T-50081 (2019) Standard for Test Methods of Concrete Physical and Mechanical Properties. Beijing, China: China Architecture & Building Press. (in Chinese).

16.

Hollaway

Teng

(2008) Strengthening and Rehabilitation of Civil Infrastructures Using Fibre-Reinforced Polymer (FRP) Composites. Cambridge, England: Woodhead Publishing.

17.

Yin

Ding

, et al. (2020) Shear behavior of large stud shear connectors embedded in ultra-high-performance concrete. Advances in Structural Engineering 23(16): 3401–3414.

18.

Hung

El-Tawil

Chao

(2021) A review of developments and challenges for UHPC in structural engineering: behavior, analysis, and design. Journal of Structural Engineering 147(9): 03121001.

19.

Idris

Ozbakkaloglu

(2014) Flexural behavior of FRP-HSC-steel composite beams. Thin-walled Structures 80: 207–216.

20.

Johnson

(1994) Composite Structures of Steel and Concrete, Volume 1: Beams, Slabs, Columns, and Frames for Buildings. Oxford, England: Blackwell Scientific Publication.

21.

Kang

Lee

Park

, et al. (2010) Tensile fracture properties of an Ultra High Performance Fiber Reinforced Concrete (UHPFRC) with steel fiber. Composite Structures 92(1): 61–71.

22.

Kim

Park

Joh

, et al. (2013) Push-out test on shear connectors embedded in UHPC. Applied Mechanics and Materials 351-352: 50–54.

23.

Kim

Kwark

Joh

, et al. (2015) Headed stud shear connector for thin ultrahigh-performance concrete bridge deck. Journal of Constructional Steel Research 108: 23–30.

24.

Kruszewski

Wille

Zaghi

(2018) Push-out behavior of headed shear studs welded on thin plates and embedded in UHPC. Engineering Structures 173: 429–441.

25.

Lai

Weng

Yang

, et al. (2024) Shear behavior and design of headed studs embedded in steel-UHPC composite structures. Structures 59: 105788.

26.

Shao

Cao

, et al. (2021) Performance of experimental and theoretical analysis on shear short headed studs embedded in UHPC. China Journal of Highway and Transport 34(8): 191–204, (in Chinese).

27.

Chen

, et al. (2023a) Calculation of shear bearing capacity, slip and stiffness of headed studs in steel-UHPC composite slab. Engineering Mechanics 40(6): 110–121, (in Chinese).

28.

Wang

Zhao

, et al. (2023b) Shear behavior of short studs in steel-thin ultrahigh-performance concrete composite structures. Case Studies in Construction Materials 19: e02423.

29.

Liu

Zhang

Meng

, et al. (2019) Transverse fatigue behaviour of steel-UHPC composite deck with large-size U-ribs. Engineering Structures 180: 388–399.

30.

Magureanu

Sosa

Negrutiu

, et al. (2012) Mechanical properties and durability of ultra-high-performance concrete. ACI Materials Journal 109(2): 177–184.

31.

Zeng

Liao

, et al. (2022) Flexural behavior of hybrid FRP-concrete-steel double-skin tubular beams with PBL shear connectors. Engineering Structures 254: 113840.

32.

Nie

(2011) Steel-concrete Composite Bridges. Beijing, China: China Communications Press. (in Chinese).

33.

Oehlers

Bradford

(1999) Elementary Behaviour of Composite Steel and Concrete Structural Members. Oxford, England: Butterworth-Heinemann.

34.

Wang

, et al. (2019) Behavior and strength of headed stud shear connectors in ultra-high performance concrete of composite bridges. Frontiers of Structural and Civil Engineering 13(5): 1138–1149.

35.

Russell

Graybeal

(2013) Ultra-high Performance Concrete: A State-Of-The-Art Report for the Bridge Community. Federal Highway Administration. Publication No. FHWA-HRT-13-060.

36.

Semendary

Stefaniuk

Yamout

, et al. (2022) Static performance of stud shear connectors and UHPC in deck-to-girder composite connection. Engineering Structures 255: 113917.

37.

Sun

Nie

, et al. (2017) Experimental research on tensile and shear behaviour of the interface between non-steam-cured UHPC and steel plate structure. Engineering Mechanics 34(9): 167–174, (in Chinese).

38.

T/CBMF37/T/CCPA7 (2018) Fundamental Characteristics and Test Methods of Ultra-high Performance Concrete. Beijing, China: China Building Material Industry Publishing House. (in Chinese).

39.

Teng

Wong

(2004) Behavior of hybrid FRP concrete-steel tubular columns. In: Proceedings of 2nd International Conference on FRP Composites in Civil Engineering, Adelaide, Australia. Hoboken, NJ: Taylor & Francis, 811–818.

40.

Teng

Wong

, et al. (2007) Hybrid FRP concrete-steel tubular columns: concept and behavior. Construction and Building Materials 21(4): 846–854.

41.

Teng

Fernando

(2012) Strengthening of steel structures with fiber-reinforced polymer composites. Journal of Constructional Steel Research 78: 131–143.

42.

Wang

Guo

Jia

, et al. (2017) Push-out tests of demountable headed stud shear connectors in steel-UHPC composite structures. Composite Structures 170: 69–79.

43.

Wang

Yao

, et al. (2018) Static behavior of grouped large headed stud-UHPC shear connectors in composite structures. Composite Structures 206: 202–214.

44.

Wang

Tong

, et al. (2019) Static behavior of large stud shear connectors in steel-UHPC composite structures. Engineering Structures 178: 534–542.

45.

Wei

Zhang

Zhou

, et al. (2022) Static and fatigue behaviors of short stud connectors embedded in ultra-high performance concrete. Engineering Structures 273: 114888.

46.

Feng

Dai

, et al. (2022) Study on mechanical properties of stud shear connectors in steel-UHPC composite structures. Engineering Mechanics 39(2): 222–234, (in Chinese).

47.

Xiao

Song

Sun

, et al. (2022) Design and experimental study of a replaceable steel UHPC composite bridge deck. Structures 40: 1107–1120.

48.

Wille

(2015) Fracture energy of UHP-FRC under direct tensile loading applied at low strain rates. Composites Part B 80: 116–125.

49.

Xiao

Yang

(2022a) Analysis of shear behavior of short group studs in ultra-high performance concrete composite bridge deck. Journal of Central South University (Science and Technology) 53(11): 4359–4371, (in Chinese).

50.

Sebastian

, et al. (2022b) Development and performance of innovative steel wedge block-crossed inclined stud-UHPC connections for composite bridge. Journal of Structural Engineering 148(9): 04022128.

51.

Xue

Briseghella

Huang

, et al. (2020) Review of ultra-high performance concrete and its application in bridge engineering. Construction and Building Material 260: 119844.

52.

Yoo

Lee

Yoon

(2013) Effect of fiber content on mechanical and fracture properties of ultra high performance fiber reinforced cementitious composites. Composite Structures 106: 742–753.

53.

Wong

Teng

, et al. (2006) Flexural behavior of hybrid FRP-concrete-steel double-skin tubular members. Journal of Composites for Construction 10(5): 443–452.

54.

Zhao

(2017) Behaviour and Modelling of Largescale Hybrid FRP-Concrete-Steel Double-Skin Tubular Beams with Shear Connectors. PhD Thesis. Hong Kong, China: The Hong Kong Polytechnic University.

55.

Zhao

Teng

, et al. (2016) Behavior of large-scale hybrid FRP-concrete steel concrete steel double-skin tubular beams with shear connectors. Journal of Composites for Construction 20(5): 4016015.

56.

Zhao

Huang

, et al. (2021) Push-out behavior of short headed stud connectors in steel-ultra high performance concrete composite deck. KSCE Journal of Civil Engineering 25(7): 2640–2650.

57.

Zhou

Song

, et al. (2018) Application of ultra-high performance concrete in bridge engineering. Construction and Building Materials 186: 1256–1267.