Experimental study on friction performance of damaged interface in steel-concrete composite beam connected by high-strength bolt

Abstract

Although traditional steel-concrete composite beam has excellent structural characteristics, it cannot meet the requirement of quick disassembly and repair in the bridge. This article presents an experimental study on friction performance of damaged steel-concrete interface in recoverable composite beam connected by high-strength frictional bolts. A total of 21 specific split tests were carried out using different concrete strength, steel strength, and surface treatment of steel. The experimental results showed that the demountable high-strength frictional bolt used in composite beam has similar performance as in the bare steel structures. The initial friction coefficient and slip stiffness were measured to be 0.34–0.47 and 52.3–116.1 kN/mm, respectively. Friction performance of damaged interface was obtained, proving that friction coefficient and slip stiffness will not decrease after the first slip damage. It is also confirmed that shot blasted steel and concrete with higher strength were more suitable in the demountable composite beam.

Keywords

damaged interface friction coefficient high-strength frictional bolt recoverable structures steel-concrete composite beams

Introduction

Nowadays, the demand of bridge design is changing from preventing damage to maintainable and restorable structural functions. Rapid repair of bridges after fatigue damage plays a very important role in traffic support. Once a bridge cannot be restored for a long time, it will cause huge economic loss and negative social influence.

For the composite beam bridge, fatigue damage always occurs on shear connectors first, while the concrete slab and steel beam remains good, as shown in Figure 1. However, it is difficult to just replace a failed connector without concrete broken, because the most widely used shear connector, headed stud, was cast in the concrete slab (Han et al., 2015; Jurkiewiez and Hottier, 2005). As a result, the conventional repair method is to remove part of the concrete slab, weld the new studs, and then pour the concrete again, which leads to more work, longer repairing time, and more material waste. Considering demand of “recoverable cities” and “recoverable structures,” it is necessary to ensure the disassembly and replaceable of shear connectors on composite beam. A new composite beam with prefabricated concrete slab, steel beam, and high-strength frictional bolt (HSFB) shear connectors may provide certain solution, as the failed bolts can be unbolted and replaced easily without concrete broken (shown in Figure 2) (Ataei et al., 2016a; Kwon et al., 2010, 2011).

Figure 1.

Fatigue failed mode of composite beam (Xing et al., 2016).

Figure 2.

Recoverable HSFB connected steel-composite beam.

The shear resistance of HSFB is provided by the friction between steel and concrete without any bonding. After fatigue failure of HSFB, relative slip must have occurred between the concrete-steel interface, which leads to a worn interface and different friction performance. Therefore, the friction performance after the slip damage directly determines the mechanical performance of the composite beam bridge after repair.

As HSFB is a new shear connector of composite beam, there is only a small amount of literatures forces on this issue. Berthet et al. (2011) carried two groups of push-out tests to evaluate the adhesion resistance of the composite structure with different connection types. A total of 16 groups of specimens were tested by Su et al. (2016) to measure the shear bond strength and friction coefficient of steel-concrete. Moynihan and Allwood (2014) obtained the performance of composite beams using M20 bolts as demountable shear connectors, and suggested that longer demountable beam had performance similar to welded connector composite beam. Pathirana et al. (2016) and Mirza et al. (2010) carried out research on demountable studs using blind bolts. It was found that blind bolts behaved very similar to welded headed studs in terms of stiffness and strength but the blind bolt had a relative brittle behavior. Ataei et al. (2016a, 2016b) presented full-scale composite beams as part of a novel deconstructable and sustainable structural system, and the bolts were proved to provide sufficient frictional resistance between the precast slabs and steel beam to ensure the full shear interaction of composite beams.

From the literature review, it is found that previous research concentrated on the performance of new concrete-steel interface, while the research of worn interface after slip damage is quite limited. In this research, a series of 21 specific split tests were conducted systematically, considering the effect of concrete strength, steel strength, surface treatment of steel, and the polypropylene fiber (PP fiber) added in the concrete.

Experimental study

Test specimens

To assess the interface properties, seven groups of tests were carried out at the Taiyuan University of Technology. In general, the test specimens were designed specifically to reflect the friction state of steel-concrete composite beam connected by high stress bolt. It consisted of a 340 × 120 × 10 mm steel plate and two identical concrete slabs of size 200 × 100 × 45 mm. The concrete slabs were connected to the steel plate by M16 high stress bolts through well-placed bolt holes. On one hand, the slip of bolts would not occur at a same moment, so the more bolts, the less accurate of the measurement. On the other hand, only one bolt cannot constrain the in-plane rotation of concrete slabs. As a result, two 8.8 grade bolts were designed in the test with the tensile strength of over 800 MPa and yield ratio of 0.8. Details of the specimen are shown in Figure 3.

Figure 3.

Details of specimen (mm).

Before the formal test, predrilled holes in the concrete slabs were used. However, local cracks near bolt holes were observed under tremendous bolt preload because of the micro-damage during drilling. As a result, all the bolt holes in concrete must be formed during casting to avoid any post-processing. Steel tubes with the diameter of 19 mm were inserted into the steel mold before casting, and be separated from concrete slabs within 2 days after casting. The steel tubes were tightly covered by thin heat-shrinkable film in order to reduce the interface bounding. In this way, can steel tube be removed easily and hole’s shape be remained. Concrete slabs were cast in the uptight position to ensure the friction surface flat and smooth. For each kind of the concrete, three standard cubes with the dimensions of 150 × 150 × 150 mm were prepared at the time that specimens were cast. All the specimens were cured in standard curing condition for more than 28 days.

Considering the lower accuracy of casting, enlarged bolt holes were employed on the concrete slabs to ensure the installation of bolts. The clearance between the hole in steel plate and shank of the bolt is 1.5 mm, while the clearance between concrete hole and shank is 3 mm. Correspondingly, larger and thicker washers with the size of Ø60 × 6 mm were adopted to protect concrete slab from local crushing according to Chinese standard for design of steel structures (GB50017:2017, 2018).

In total, 21 specimens were divided into seven groups as presented in Table 1. In each group, three replicate samples were tested to obtain the average results and eliminate accidental error. Four controlled variables were considered in the test including the compressive strength of concrete, the yield strength of steel, the PP fiber added in the concrete, and the surface treatment of steel.

Table 1.

Summary of the test specimens.

Group	Specimen	Concrete	Steel	Surface treatment of steel
Group 1	S1–S3	C40	Q235	Shot blasted
Group 2	S4–S6	C50
Group 3	S7–S9	C60
Group 4	S10–S12	C60 + PP fiber
Group 5	S13–S15	C60	Q355
Group 6	S16–S18	C60	Q235	Loose rusty
Group 7	S19–S21	C60	Q235	Metal brushed

Specimen assembling

Pretension of the bolt is another important factor affecting the friction of concrete-steel interface except friction coefficient. The nuts were tightened by wrench from which the torque could be read. The pretension of bolt can be calculated by equation (1)

T_{c} = k P_{t} d

(1)

where T_c is the final torque of a high-strength bolt, k is the torque coefficient which determined by geometry of the thread, P_t is the pretension stress of high-strength blot, d is the nominal diameter of the bolt. For the HSFBs in the tests, k was 0.15.

All the concrete slabs could not be reinforced in this study on account of their small thickness of 45 mm. Different pretension was applied before the formal test to show that C40 concrete was easily crushed when the pretension was greater than 45 kN. As a result, the pretension load was determined as 40 kN in order to ensure a high success rate of the test. High-strength friction bolts were screwed in two steps, initial screwing and final screwing, to ensure the uniform pretension and closed interface. The torque of initial screwing was about 50% of final screwing. What’s more, bolts and washers must not be turned with nuts during the final screwing, otherwise the bolt had to be changed (Yang et al., 2018).

Test setup and loading procedures

Double repeated loading was applied by a 600-kN electro-hydraulic servo testing machine. The specimen was set on a base plate that can rotate as spherical hinge so as to collimate actuator and steel plate. Two concrete blocks were put under the concrete slabs, so the steel plate did not touch the base and could be pushed-out. The blocks were to create space for slippage of steel plate, so they were fabricated with high strength and rigidity, and were not replaced or moved during a test. The friction created by vertical load would be big enough to prevent any horizontal slip of the two blocks without any lateral support. The loading setup is shown in Figure 4.

Figure 4.

Loading setup.

At once the friction was overcome, the HSFB was considered failed, and the first loading was stopped. After that, only the two failed bolts were replaced, then the concrete slabs and steel plate with damaged interface were reassembled by new HSFB for second loading. All the specimens were preloaded before formal test to remove the gap between testing machine and specimen. The preload was about 20% of the estimated slip load. Load control with the rate of 0.5 kN/s was used for the formal monotonic test.

The load and displacement produced by the actuator were measured by built-in transducer continuously throughout the entire test, so the load–displacement curves could be shown timely. The displacement jump in load–displacement curve indicated that the friction had been overcome and the test should be stopped.

Experimental results and discussion

Mode of failure

Two main failure modes were observed in these tests. The first mode of failure is slip jump between concrete-steel interface. In this type of mode, slip jump means the overcome of friction, and then the high-strength friction-type bolt turned to be pressure type. Although shear load could still be transferred by pressure of shank and hole instead of friction, the test of interface performance was over.

The second mode of failure is slip jump with concrete crack. Vertical cracks all began at the edge of the lower hole because of the pressure between bolt shank and concrete hole after slipping, as shown in Figure 5(a). If the friction of one bolt was overcome first but another bolt did not, the different slip trend might lead to vertical tensile stress in concrete and horizontal cracks as shown in Figure 5(b). Effective sliding loads can be obtained in both failure modes in the first loading. However, some serious cracks in the first loading directly led to failure of the second loading; as a result, the second sliding loads of S1, S6, and S20 had not been obtained.

Figure 5.

Cracks in concrete: (a) vertical crack; (b) horizontal crack.

Since the loading stopped as soon as the slip occurred, the stress of the bolt was much less than its tensile strength, so no shear or bending deformation was found on the bolt shank.

Load–slip relationship

Double load versus slip behavior of all test specimens is presented in Figure 6. In this figure, S2 and S2′ in the legend are the first loading curve and second loading curve of specimen S2 relatively, as an example. As shown in Figure 6(a), the whole process of test can be divided into three stages: friction stage, slip stage, and pressure stage. In the friction stage (O-A), the shear load is entirely resisted by the static friction force, so the relative slip between steel and concrete is quite small and the load–slip curves show an almost linear relationship for all the specimens.

Figure 6.

Load–slip curves: (a) Group 1; (b) Group 2; (c) Group 3; (d) Group 4; (e) Group 5; (f) Group 6; and (g) Group 7.

Apparent slip jump stared at point A when the load overcomes the maximum static friction force, which means the beginning of slip stage (A-B). The curves in this stage are almost horizontal, indicating that slipping occurred suddenly and developed rapidly. In these tests, almost all the slip at the end of slip stage was less than 3 mm, because the value of slip depends on the clearance between bolt shank and bolt hole. If the bolt was ideally in the middle of the bolt holes, the slip at point B ought to be 2.25 mm in theory, which is half of the concrete clearance plus steel clearance. Considering the deviation of bolt position, the theoretical slip range is between 0 and 4.5 mm. Due to the high installation accuracy of this test, the slip was all controlled within 3 mm.

Once the bolt shank contacted with the concrete hole, slip jump ended at point B and pressure stage (after B) began. In this stage, high-strength friction-type bolt turned to be pressure type (shown in Figure 7). The slowly increased relative slip was mainly caused by the deformation of bolts or the cracking of concrete slab. For HSFBs, friction stage is the normal service state allowed in the design, while the bearing capacity provided in slip stage and pressure stage is only used as strength reserve. Therefore, tests were stopped in the pressure stage, and the load of point A was regarded as the shear bearing capacity $N_{V}$ of the specimen.

Figure 7.

Load transfer mechanism.

Discussion of experimental results

Effect of concrete strength

Table 2 summarizes the maximum shear capacity, calculated friction coefficient, and failure modes of all the specimens. In Table 2, $\sum P$ and $\sum P'$ are the total pretension stress of tow bolts in two loading, relatively; $N_{V}$ and $N'_{V}$ are the load at the moment slip jump occurs in two loading, relatively; µ and $μ'$ are the friction coefficient in two loading, relatively; $\bar{μ}$ and $\bar{μ'}$ are the mean value of friction coefficient in two loading, respectively. The friction coefficient µ, as a function of load $N_{V}$ , can be calculated by equation (2). The concrete slab of specimen S7 cracked accidentally before slippage of steel plate, so the test of S7 is a failure with no slip load collected

μ = \frac{N_{v}}{2 \sum P}

(2)

Table 2.

Summary of friction coefficient and failure modes.

Group	Specimen	$\sum P$ (kN)	$N_{V}$ (kN)	µ	$\bar{μ}$	$\sum P'$ (kN)	$N'_{V}$ (kN)	$μ'$	$\bar{μ'}$	$\bar{μ'} / μ'$	Concretecrack
Group 1	S1	83.3	59.5	0.357	0.377	–	–	–	0.449	1.19	First loading
	S2	85.0	73.2	0.431		85.0	78.9	0.464			Second loading
	S3	85.0	58.4	0.344		88.3	89.1	0.434			No crack
Group 2	S4	82.5	61.6	0.373	0.390	85.0	70.3	0.414	0.411	1.05	No crack
	S5	83.3	65.2	0.391		83.3	68.1	0.409			No crack
	S6	84.2	68.2	0.405		–	–	–			First loading
Group 3	S7	–	–	–	0.444	–	–	–	0.489	1.10	–
	S8	83.3	74.1	0.445		83.3	88.0	0.528			Second loading
	S9	83.3	73.9	0.443		83.3	74.9	0.449			Second loading
Group 4	S10	85.0	64.4	0.379	0.420	85.8	50.9	0.296	0.389	0.93	Second loading
	S11	85.0	69.7	0.410		84.2	69.9	0.415			First loading
	S12	81.7	76.8	0.470		83.3	76.1	0.456			First loading
Group 5	S13	85.0	70.8	0.416	0.458	84.2	75.5	0.448	0.489	1.07	No crack
	S14	83.3	85.6	0.514		86.7	95.3	0.550			Second loading
	S15	85.0	75.6	0.445		84.2	79.1	0.470			No crack
Group 6	S16	83.3	54.0	0.324	0.358	83.3	43.6	0.261	0.395	1.12	First loading
	S17	84.2	60.0	0.356		82.1	82.4	0.502			No crack
	S18	84.2	64.1	0.381		85.0	71.9	0.423			No crack
Group 7	S19	83.3	53.8	0.323	0.285	85.0	64.9	0.382	0.350	1.23	No crack
	S20	83.3	49.3	0.296		–	–	–			First test
	S21	84.2	39.3	0.237		84.2	53.7	0.319			Second test

The test results of Group 1, Group 2, and Group 3 were demonstrated in Figure 8. The surface roughness of all the concrete is basically the same because they were cast in the same mold and method. However, it can be seen that $\bar{μ}$ increases with the compressive strength of concrete, indicating that surface roughness in not the only factor that affects the shear strength of concrete-steel interface. The $\bar{μ}$ of C60 concrete is 13.8% and 17.8% higher than that of C50 and C40 concrete, relatively. This is mainly due to the better crack resistance of C60 concrete. Under the action of bolt pretension stress, micro-cracks are more likely to occur in low-strength concrete as seen in Figure 9, leading to the loss of actual bolt pretension stress and lower failure load N_v.

Figure 8.

Effect of concrete strength on µ.

Figure 9.

Micro-cracks of concrete (×2000): (a) C50 concrete; (b) C60 concrete.

When the HSFB failed in the first loading, relative slip occurs on the contact interface, resulting in slip damage on the smooth concrete surface. For each specimen in the Group 1–3, the interface friction coefficient of damaged interface $μ'$ is all higher than its $μ$ of new interface. In order to explain the mechanism of this change, the microscopic damage of concrete surface was observed by scanning electron microscope (SEM) with the magnification of 250 times.

Due to the much less surface hardness, slip damage scratches caused by slip were found only on the concrete surface instead of steel plate, as shown in Figure 9. The width of the scratches is about 0.1 mm, and spacing between scratches is about 0.4 mm. Benefits from the scratches, roughness of the damaged surface was obviously increased compared to the undamaged concrete surface in Figure 10. During the second loading, the interlocking and friction stress of the steel-concrete interface increased correspondingly.

Figure 10.

Surface damage of C60 concrete (×250): (a) undamaged concrete surface; (b) damaged concrete surface after first loading.

Some conclusions can be drawn that higher strength and slip damaged surface of concrete can improve the shear stress of HSFB connected concrete-steel interface. Moreover, the strength of the concrete should be well matched with the pretension force of the HSFB.

Effect of steel strength

Q235 and Q355 steel are two of the most commonly used materials in the Chinese composite structures. Figure 11 compares the results of Group 3 and Group 5, whose only difference is strength of steel.

Figure 11.

Effect of steel strength on µ.

Both the first loading and second loading results show little difference between the two groups. In the first loading, the mean friction coefficient of the specimens using Q355 steel plate is only 3.2% larger than that of Q235 steel plate, while the ones obtained from the second loading are exactly same. Moreover, the small difference of result can be attributed to the dispersion of the tests.

This result also confirms the conclusion above from another aspect, that it is the concrete strength instead of steel strength plays a controlling role in the friction performance of damaged interface, because concrete is more vulnerable.

Effect of surface treatment of steel

Surface treatment of steel has a significant effect to the friction performance of concrete-steel interface. As shown in Figure 12, three different surface treatments used in the Group 3, Group 6, and Group 7 were shot blasted steel, loose rusty steel, and metal brushed steel, respectively. It is obviously that the metal brushed steel had the smoothest surface and lowest roughness. The roughness of the first two steel plates were much higher, and the shot blasted surface was harder than the loose rust. Correspondingly, the $\bar{μ}$ of Group 6 and Group 7 was 19.4% and 35.8% lower than shot blasted steel in Group 3, as show in Figure 13.

Figure 12.

Surface treatments of steel plate.

Figure 13.

Effect of steel surface treatment on µ.

In the second loading, friction coefficient $\bar{μ'}$ of these three groups were all higher than the first loading result $\bar{μ}$ , proving increment of concrete surface roughness caused by slip damage. However, the $\bar{μ'}$ of Group 3, Group 6, and Group 7 was still significantly different, indicating diverse damage ability of three surface treatments of steel.

In order to reveal the effect of steel surface treatment on concrete damage, concrete surface of three groups was observed by SEM at a magnification of 1000 times. As shown in Figure 14, the concrete plates of Group 3 (S8 and S9) with shot blasted steel were the most damaged ones with the scratches’ width of about 0.1 mm. For the concrete plates in Group 6 (S16–S18), the scratches are much smaller with the width of about 0.03 mm, which mainly attribute to the poor hardness of loose rust. It is quite different that the scratches on concrete surface damaged by metal brushed steel in Group 7 (S19–S21) have no obvious depth and width.

Figure 14.

Surface of C60 concrete damaged by steel: (a) undamaged (×1000); (b) damaged by shot blasted steel (×1000); (c) damaged by rusty steel (×1000); and (d) damaged by metal brushed steel (×1000).

From what has been found above, the steel plate with low friction coefficient has relatively small damage to the concrete surface, so the friction coefficient of the damaged interface $μ'$ is also relatively small. As a result, shot blasted steel is suggested to be used in the HSFB connected demountable steel-composite beam.

Effect of PP fiber

PP fiber concrete is commonly used in the road and bridge deck due to its good deformability. The results of Group 3 and Group 4 were compared in Figure 15 to show the effect of PP fiber. For the specimens made of C60 PP fiber concrete, mean friction coefficient $\bar{μ}$ of the interface was 0.420, decreasing by 5.4% than the common C60 concrete. The $μ'$ of damaged interface was even 7.4% lower than the undamaged one, which is completely different from the rule of common concrete.

Figure 15.

Effect of polypropylene fiber on µ.

The reason might be found from the macro- and micro-picture of surface of C60PP concrete. As PP fibers are softer, some of them were exposed out of the concrete surface shortly, which lead to a gap between concrete and steel and prevent them from sticking together. Therefore, fibers have a negative effect on new interface’s friction performance, but the effect is limited as the fibers are tiny. After the first loading, even though some scratches were produced to increase the concrete surface roughness, micro-cracks also appeared along the direction of the fibers to decrease the pretension stress of HSFB (shown in Figure 16). As a result, the shear capacity of the specimens in Group 4 was not as good as that of Group 3, proving PP fiber concrete is not suitable in the recoverable HSFB connected steel-concrete composite beam.

Figure 16.

Surface of C60PP concrete surface.

Friction coefficient and stiffness of damaged interface

In total, 17 specimens of totally 21 ones were loaded twice successfully in the tests. It can be seen from Table 2 that except for Group 4, the $\bar{μ'}$ of damaged concrete-steel interface in other groups is higher than the $\bar{μ}$ of new interface by 5%–23%. The relationship of $μ$ and $μ'$ of all the 17 specimens is shown in Figure 17. Only three specimens locate below the diagonal, indicating that the damaged concrete-steel interface has a higher friction coefficient than before in more than 80% of the cases. Regardless the different material strength and surface treatment of steel, most friction coefficient of new concrete-steel interface ranged from 0.34–0.47, while the friction coefficient of damaged concrete-steel interface ranged from 0.40–0.53.

Figure 17.

Relationship of $μ$ and $μ'$ .

The slip stiffness of HSFB connected concrete-steel interface is another important performance, which will determine the deformation of concrete beam in the real project. To calculate the stiffness, 0.7N_v is usually used as the ending point of linear elastic stage (Shim et al., 2004). Therefore, in this study, 0.7N_v and corresponding slip was used to calculate the secant stiffness K, and the same method was also applied to the calculation of K′.

The highest stiffness in first loading was observed in specimen S13 with 116.1 kN/mm, and the lowest stiffness was observed in specimen S16 with 52.3 kN/mm. The gap between actuator and specimen leads to the systematic error of the displacement measurement. Compare to larger scale specimen as supported composite beam, this error has a greater impact on smaller specimens in this study. As a result, the discrete of stiffness is higher than that of friction coefficient.

In spite of the system error, about 65% of the specimens show an increased stiffness in the second loading. For the other 6 of 17 specimens, their K′ were only slightly less than 5% lower than the K, as the data points located very closely to the diagonal (see Figure 18).

Figure 18.

Relationship of K and K′.

It can come to the conclusion that the friction coefficient and slip stiffness of the concrete-steel interface will not decrease basically after the primary friction damage. Therefore, it is possible to replace the failed bolts and repair the composite structure without decreasing any mechanical property.

Conclusion

A total of 21 specifically designed tests had been conducted to investigate the friction coefficient and slip stiffness of damaged interface in steel-concrete composite beam connected by high stress frictional bolt. The results obtained from the experimental work have led to the following conclusions:

The initial friction coefficient of concrete-steel interface ranged from 0.34–0.47, which is similar to the friction coefficient of steel-steel interface. The initial secant stiffness is 52.3–116.1 kN/mm.

The concrete strength affects the friction of concrete-steel interface. It appears that the ultimate shear resistance increases with the increase in concrete strength. On the contrary, the steel strength had little effect.

Friction coefficient of concrete-steel interface is greatly affected by the surface treatment of steel. Shot blasted steel is suggested in concrete-steel composite project, while the metal brushed steel should be forbidden.

The use of PP fiber is not suitable in the recoverable HSFB connected steel-concrete composite beam.

The demountable HSFB can be used to connect concrete and steel members. It is possible to replace the failed bolts and repair the composite structure without decreasing any mechanical property, as friction coefficient and slip stiffness will not decrease basically after the primary friction damage.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work in this article was supported by the National Natural Science Foundation of China (No. 51708384), Natural Science Foundation of Shanxi Province, (No. 201801D221216 and No. 201901D211017), Science and Technology Project for Oversea Scholars in Shanxi Province (No. DC1900000602), and Science & Technology Project of State Grid Shandong Electric Power Company (No. 52062519001A).

ORCID iDs

Ying Xing

Qi Guo

References

Ataei

Bradford

Liu

(2016a) Experimental study of composite beams having a precast geopolymer concrete slab and deconstructable bolted shear connectors. Engineering Structures 114: 1–13.

Ataei

Bradford

Valipour

, et al. (2016b) Experimental study of sustainable high strength steel flush end plate beam-to-column composite joints with deconstructable bolted shear connectors. Engineering Structures 123: 124–140.

Berthet

Yurtdas

Delmasand

, et al. (2011) Evaluation of the adhesion resistance between steel and concrete by push out test. International Journal of Adhesion and Adhesives 31(2): 75–83.

GB50017:2017 (2018) Standard for design of steel structures (in Chinese).

Han

Xing

, et al. (2015) Static push-out test on steel and recycled tire rubber-filled concrete composite beams. Steel and Composite Structures 19(4): 843–860.

Jurkiewiez

Hottier

(2005) Static behaviour of a steel-concrete composite beam with an innovative horizontal connection. Journal of Constructional Steel Research 61(9): 1286–1300.

Kwon

Engelhardt

Klingner

(2010) Behavior of post-installed shear connectors under static and fatigue loading. Steel Construction 66(4): 532–541.

Kwon

Michael

Richard

(2011) Experimental behavior of bridge beams retrofitted with post-installed shear connectors. Journal of Bridge Engineering 16(4): 536–545.

Mirza

Patel

(2010) Behavior and strength of shear connectors utilizing blind bolting. In: Proceedings of the 4th international conference on steel and composite structures, Sydney, NSW, Australia, 26–30 July.

10.

Moynihan

Allwood

(2014) Viability and performance of demountable composite connectors. Journal of Constructional Steel Research 99: 47–56.

11.

Pathirana

Mirza

, et al. (2016) Bolted and welded connectors for the rehabilitation of composite beams. Journal of Constructional Steel Research 125: 61–73.

12.

Shim

Lee

Yoon

(2004) Static behavior of large stud shear connectors. Engineering Structures 26(12): 1853–1860.

13.

Q-T

C-X

, et al. (2016) Tests of basic physical parameters of steel concrete interface. Journal of Tongji University (Natural Science) 44(4): 499–506 (in Chinese).

14.

Xing

Han

, et al. (2016) Experiment study on the fatigue behavior of partial shear connected composite beam with crumb rubber concrete. Journal of Hunan University (Natural Science) 43(7): 32–42 (in Chinese).

15.

Yang

Liu

Jiang

, et al. (2018) Shear performance of a novel demountable steel-concrete bolted connector under static push-out tests. Engineering Structures 160: 133–146.