Experimental investigation of steel-concrete composite beams with different degrees of shear connection under monotonic and fatigue loads

Abstract

Steel-concrete composite beams have been extensively used in buildings and bridges in recent years to combine these two construction materials. The behavioral characteristics of steel-concrete composite beams are related to the degree of shear connection. In this article, eight steel-concrete composite beams were investigated to examine the effects of different degrees of shear connection on the static and fatigue performances of the beams. First, the ultimate capacity, ductility coefficient, failure modes, strain distribution at the midspan section, and relative slip distribution of the steel-concrete composite beams under monotonic loads were investigated with different shear connection degrees. The effect of the shear connection degree on the fatigue life, failure mode, and other mechanical properties under fatigue loading was studied. The results show that all mechanical properties of the steel-concrete composite beams under monotonic and fatigue loads decrease with a decrease in shear connection degree, and their failure modes change with the decreasing shear connection degree. Second, based on the experimental results, the fatigue life of composite beams was analyzed and predicted under different assumptions of failure modes according to EC4, AASHTO, and the Chinese code. The result shows that different failure modes should be considered according to different shear connection degrees in the fatigue design of composite beams. Finally, the issues on the difference between static and fatigue capacities with different shear connection degrees and the critical point of the failure mode were discussed.

Keywords

failure mode fatigue shear connection degree slip steel-concrete composite beams

Introduction

Steel-concrete composite (SCC) beams have been extensively used in buildings and bridges in recent years to combine these two construction materials (Bing et al., 2017; Xiang et al., 2015; Zhou et al., 2016). The shear connector is the key component that assures shear connection between the steel girder and the concrete deck, which enables the composite action to contribute to the shear transfer and prevent uplift (Selvi, 2016; Vasdravellis and Uy, 2014). Therefore, the shear connection can be obtained by reducing or preventing the relative displacement of the concrete and steel sections at their interfaces (Rodrigues and Laím, 2014; Zhan et al., 2016). The degrees of shear connection are generally determined by the arrangement and strength of the shear connectors in SCC beams. In addition, it affects the structural behavior and failure mode of the composite beam. As a result, the design of SCC beams requires specific attention to the degree of shear connection (Zona and Ranzi, 2014).

The degree of shear connection in a composite beam is mainly characterized by the shear strength, shear stiffness, and number of installed shear connectors. With different degrees of shear connection, composite beams can be divided into three types: no-shear-connection beams, partial-shear-connection beams, and full-shear-connection beams. Many studies focus on the latter two types of beams (Ban and Bradford, 2014; Loh et al., 2006; Queiroz et al., 2008; Titoum et al., 2009). Several authors have experimentally or numerically studied the structural behavior of composite beams with different arrangements of shear connectors in recent years.

Sanghyo et al. (2011) investigated the relationship between the structural behavior and the degree of shear connection using three composite beam specimens with different stud diameters. The results showed that the elastic stiffness, yield load, and ultimate load were affected by the degree of shear connection. Experimental and numerical studies by Vasdravellis and Uy (2014) were performed on the shear strength and moment–shear interaction in simply-supported SCC beams. In total, 14 composite beams and 1 steel beam were tested under combined bending and shear. It is shown that the main factors that affect the shear capacity of a composite beam are the slab thickness and degree of shear connection. Fanaie et al. (2015) studied the differences between face-to-face and back-to-back arrangements of channel shear connectors in composite beams by finite element analyses. In addition, the effects of the channel size, and length and spacing of channels were investigated. Based on elastic concrete, Xing et al. (2016) investigated the static behavior of elastic SCC beams with different section sizes, studs, and degrees of shear connection. The test results showed that a larger degree of shear connection can lengthen the elastic stage and delay the development and spread of slip but may decrease the ductility.

Previous studies mainly focused on the static performance of SCC beams with different degrees of shear connection. However, it is also worth noting that due to population and economic growth, traffic flow has substantially increased in the last few decades, subjecting the steel girder and shear connectors to high-cycle fatigue loading by the vehicles and resulting in the fatigue of SCC beams, particularly in bridges (Nagy et al., 2016; Wannenburg et al., 2009). In addition, some researchers focused on the vibration of the SCC beam with different connection degrees for damage detection (Dilena and Morassi, 2009; Jimbo et al., 2012). However, few studies have systematically examined the effect of the shear connection on the fatigue behaviors of composite beams. Thus, this article attempts to investigate the fatigue performance of SCC beams with different degrees of shear connection under repeated loading.

In this article, we designed and fabricated eight simply-supported SCC beam specimens. The headed studs were selected, which are the most widely used shear connectors. The SCC beam specimens were designed for four different shear connections by varying the number of stud connectors in the longitudinal direction. Four specimens were tested under monotonic loads, and four similar beam specimens were subjected to cyclic loads. From the experimental results, the ultimate strength, failure mode, fatigue resistance, and so on were evaluated and discussed.

Experimental investigation

Shear connection degree

The shear connection degree reflects the equilibrium of the axial strength between the shear connection and the concrete or steel element in the composite section (Oehlers et al., 1997; see Figure 1).

Figure 1.

Degree of shear connection in composite section.

As shown in Figure 1, the axial strength of the concrete element is F_c = A_cf_c, the axial strength of the steel element is F_s = A_sf_y, and the shear strength of the shear connection in the shear span is V_s = n_sV_u, where n_s is the number of headed studs and V_s is the shear capacity of each individual stud. Thus, full shear connection occurs when the strength of shear connection is greater than the minimum between F_c and F_s (V_s ≥ min(F_c, F_s)). In addition, partial shear connection occurs when the strength of shear connection does not fulfill the previous criterion (V_s < min(F_c, F_s)). No shear connection, which is a special case of partial shear connection, is when V_s = 0.

Test scheme and specimen configuration

Eight simply-supported SCC beam specimens were designed based on the Chinese code (GB50017-2003, 2003) as shown in Figure 2. Four beam specimens (SCB-1 to SCB-4) were subjected to monotonic loads, and the remaining four beam specimens (FCB-1 to FCB-4) were tested under fatigue loads.

Figure 2.

Dimension and reinforcement of specimen (unit = mm).

The total length of each beam specimen was 3200 mm, and the span was 3000 mm. A concrete deck that was 300-mm wide and 80-mm thick was used. The I-shaped steel girder was welded using a 10-mm-thick steel plate with a 120-mm-wide top flange, a 160-mm-wide bottom flange, and a 150-mm web height. Stiffening ribs were established at the midspan, quarter points, and support sections to improve the local buckling stability of the beam specimens. The type of longitudinal reinforcement and stirrup in concrete slab was 6-mm-diameter hot-rolled reinforced bar. And the design strength of reinforcement is 300 MPa. In the cross section of concrete slab, there were six longitudinal reinforcements. The longitudinal reinforcement ratio in cross section is 0.707%. In the elevation of concrete slab, the stirrup spacing was 100 mm, and the transverse reinforcement ratio was 0.353%.

Two rows of stud connectors with the diameter of 13 mm were welded on the top flange of the steel girder with a transverse spacing of 60 mm. The longitudinal spacing was a variable parameter for the specimens, as shown in Table 1. The shear connection degree η was defined as follows

η = \frac{n_{s}}{n_{f}}

(1)

n_{f} = ⌈ \frac{min {A_{s} f_{y}, A_{c} f_{c}}}{V_{u}} ⌉

(2)

where n_s is the actual number of stud connectors in a shear span, n_f is the least number of stud connectors in a shear span in a full-shear-connection SCC beam, and V_u is the shear capacity of a single stud, which is defined by the corresponding push-out tests in section “Push-out tests for the stud connector.”

Table 1.

Summary of the experimental program.

Group	Specimens	No. of studs	Stud spacing (mm)	η	P_max/P_u	P_min/P_u	ΔP/P_u
Monotonic loading specimens	SCB-1	28	13×215	1.00	–	–	–
	SCB-2	20	9×310	0.71	–	–	–
	SCB-3	16	7×400	0.57	–	–	–
	SCB-4	32	15×186	1.14	–	–	–
Fatigue specimens	FCB-1	28	13×215	1.00	0.60	0.35	0.25
	FCB-2	20	9×310	0.71	0.60	0.35	0.25
	FCB-3	16	7×400	0.57	0.60	0.35	0.25
	FCB-4	32	15×186	1.14	0.60	0.35	0.25

The fabrication procedures of the SCC beam specimens are illustrated in Figure 3.

Figure 3.

Fabrication procedures of the test specimens: (a) stud welding, (b) finished steel girders, (c) fabrication of mold, (d) concrete placement, (e) concrete curing, and (f) completed specimen.

The material properties of the eight specimens were identical. The material of the steel girder was Q345 steel, and the tested mean yield strength was 352 MPa. At the time of the monotonic load test, the concrete strength was measured using three standard prismatic specimens. The mean value obtained was 38.5 MPa.

Push-out tests for the stud connector

Before the SCC beam specimen design, push-out tests were performed to evaluate the ultimate capacity of the stud connector, which would be installed in the composite beam specimens (Bing et al., 2017). Three push-out specimens were made with the identical materials as the composite beam. The average ultimate capacity of a single stud in three specimens was 70.2 kN, which was obtained for further analysis.

Figure 4 presents the load–slip relationships of single stud connector in three push-out specimens. The experimental data were fitted by the commonly used load–slip constitutive model (Ollgard et al., 1971) in equation (3). The fitting determination coefficient is 0.973, which indicates a good fitting effect

V = V_{u} {(1 - e^{- 1.75 s})}^{0.85}

(3)

Figure 4.

Load–slip curves of specimens SCP-1 to SCP-3.

Test set-up and instrumentation

All beam specimens were loaded in a three-point bending test scheme with a span length of 3000 mm. The beam test was performed on the computer-controlled, two-channel, electro-hydraulic servo static and dynamic loading test system (JAW-500K), as shown in Figure 5. The maximum loading capacity of the actuator is 500 kN. A displacement sensor and a force sensor were fixed on the actuator to capture the displacement and load at the loading end during the test.

Figure 5.

Test set-up and instrumentation: (a) loading device for experiments, (b) electronic control cabinet, and (c) strain collection device.

In total, 10 linear voltage displacement transducers (LVDT) were used to measure the displacements during the test, as shown in Figure 6. Among them, three LVDTs were installed at the quarter points and midspan sections of each beam to measure the vertical displacements, two LVDTs were fixed on both ends of the beam specimen to investigate the uplift displacements, and the remaining five LVDTs were uniformly distributed in a shear span to measure the interface slip of the concrete deck and steel girder. Moreover, three concrete strain gauges and four steel strain gauges were distributed along the beam depth in the midspan section (see Figure 6). The size of steel strain gauges was 3 mm × 2 mm, and the gage factor was 2.06. The size of concrete strain gauges was 80 mm × 3 mm, and the gage factor was 2.06.

Figure 6.

Measuring-point layout of the specimen.

Monotonic loading tests

Load–deflection relationships and failure modes

As shown in Table 1, the experimental tests were divided into two groups: monotonic loading tests and fatigue tests. In the monotonic loading test, the loading was increased at an incremental rate of 10 kN/min until the load reached 60% of the theoretical ultimate load. Subsequently, the displacement control was applied at the rate of 0.5 mm/min. This group included four SCC beam specimens (SCB-1 to SCB-4).

Figure 7 presents the load–deflection curves of four beam specimens with different shear connection degrees under monotonic loading. All curves show three stages in a typical failure process including elastic stage, elastic–plastic stage, and plastic stage. In addition, a comparison of the four curves in Figure 8 shows that the ultimate capacity and ductility of the beams decrease with a reduction in the degree of shear connection.

Figure 7.

Load–deflection responses of the specimens with different shear connection degrees.

Figure 8.

Failure modes of the beam specimens under monotonic loading: (a) SCB-1, (b) SCB-4, (c) SCB-2, and (d) SCB-3.

The failure modes of the SCC beam specimens vary with different degrees of shear connection. When η ≥ 1.0, which corresponds to the composite beams SCB-1 and SCB-4, these two beams experienced flexural failure modes. The specific process was that the steel girder yielded, and the concrete deck crushed in the midspan, as shown in Figure 8(a) and (b). When η = 0.71, which refers to the beam SCB-2, specimen SCB-2 had a similar failure mode to the first two specimens, except more cracks appeared on the bottom of the concrete deck because of the decrease in shear connection (Figure 8(c)). When η = 0.57, the composite beam (SCB-3) experienced shear connection failure. The specific process was that the studs suddenly fractured before the concrete deck crushed, and the concrete deck clearly separated from the steel girder, as shown in Figure 8(d).

The ultimate capacity and ductility are the most significant concepts in the design of composite beams. The ductility coefficient μ is defined as the ratio of the maximum deflection Δ_u and the elastic limit deflection Δ_p, that is, μ = Δ_u/Δ_p. Table 2 shows the ultimate capacity P_u and ductility coefficient of each beam specimen.

Table 2.

Ultimate capacity and ductility of the composite beams.

η	Specimens	P_u (kN)	Ratio 1	Δ_p (mm)	Δ_u (mm)	μ (Δ_u/Δ_p)	Ratio 2
1.14	SCB-4	240	1.05	11.47	84.66	7.38	1.10
1.00	SCB-1	228	1.00	11.98	80.19	6.69	1.00
0.71	SCB-2	220	0.96	11.11	71.98	6.48	0.97
0.57	SCB-3	204	0.89	11.78	55.64	4.72	0.71

Ratio 1 = P_u of each case/P_u of SCB-1; Ratio 2 = μ of each case/μ of SCB-1.

In Table 2, the ultimate capacity and ductility coefficient decrease with the decrease in degree of shear connection. In addition, the SCB-3 specimen with the smallest degree of shear connection had a larger ductility reduction than the others.

Load–strain relationships and load–relative slip relationships

Figure 9 shows the strain distribution at the midspan section of SCC beam specimens with different degrees of shear connection before the yield load of the composite beams. As observed in Figure 9, only the strain distributions of SCB-1 and SCB-4 specimens (η ≥ 1.0) are consistent with the assumption of a plane section. For SCB-2 and SCB-3 specimens (η < 1.0), the tensile stress appears at the bottom of the concrete deck, and two neutral axes occur. Obviously, the shear connection degree determines the stress state of the beam, which should be considered in the design.

Figure 9.

Load–strain distributions at the midspan section: (a) SCB-1, (b) SCB-2, (c) SCB-3, and (d) SCB-4.

The load–relative slip relationship, which is another important parameter, affects the mechanical properties of composite beams. Figure 10 shows the distribution of the relative slip along the length in the shear span under different load levels, where δ is the relative slip between the concrete deck and the steel girder, and L is the distance from the support to the measured section. It can be seen that the slip linearly distributes along the length under the same load with the full-shear-connection composite beams (SCB-1 and SCB-4), except for the midspan. For the partial-shear-connection composite beams (SCB-2 and SCB-3), the relative slip along the length shows a nonlinear distribution and the maximum slip under each level of load appears at 1/8 point. Moreover, the relative slip under the same load ratio increases with the decrease in shear connection degrees.

Figure 10.

Relative-slip distributions according to the load steps: (a) SCB-1 (η = 1.00), (b) SCB-2 (η = 0.71), (c) SCB-3 (η = 0.57), and (d) SCB-4 (η = 1.14).

Fatigue test

Fatigue life and failure modes

In the fatigue test, a repeated force of sine wave was adopted with a loading frequency of 4 Hz. The control mode was force control. This group consisted of four SCC beam specimens of FCB-1 to FCB-4. Generally, the expected fatigue life of structures in civil engineering is 2 million cycles (Chen et al., 2011). For welded steel girder, its fatigue strength at 2 million cycles is approximately 100 MPa according to Chinese specifications (JTG D64-2015, 2015).

In this study, the SCC beam specimens were similar to those in the case of a welded steel girder. For the full-shear-connection beam, that is, FCB-1, the loading range ΔP can be calculated according to the designed stress range and ΔP = 56 kN, which was 0.25 times of P_u. And P_u is the static ultimate capacity defined by the corresponding monotonic loading test of SCB-1.

The values of maximum load P_max and minimum load P_min will be determined later. Normally, the ratio between the service load and the ultimate load on the SCC beam bridge is 0.5–0.6. Hence, for FCB-1, P_max was defined as 0.6 P_u. As a result, P_min = (0.6–0.25) and P_u = 0.35. To make the beam tests more comparable, all fatigue beam specimens used the same fatigue parameters as FCB-1, which are listed in Table 1.

Table 3 summarizes the actual fatigue parameters and the test results of the fatigue test specimens (FCB-1 to FCB-4). In Table 3, P_max, P_min, and ΔP of each specimen are different, but the ratios of fatigue parameters, which correspond to the ultimate load of each specimen, are identical. The test result shows that the fatigue life of the beam specimen increases with the increase in shear connection degree. When η = 1.00, the fatigue life was approximately 2 million cycles. When η = 1.14, the fatigue life only increased by 7.1% of that in FCB-1. However, when η < 1, the fatigue life of the beam specimens sharply decreased. For example, the fatigue life of SCB-3 is only 0.20 times of 2 million cycles.

Table 3.

Fatigue test results of the SCC beam specimens.

η	Specimens	P_max (kN)	P_min (kN)	ΔP (kN)	No. of fatigue loading (N)	Failure mode
1.14	FCB-4	142	84	58	2,238,413	Steel girder failure
1.00	FCB-1	136	80	56	2,075,600	Stud failure
0.71	FCB-2	132	78	54	759,410	Stud failure
0.57	FCB-3	122	72	50	394,565	Stud failure

Regarding the failure mode, the specimen FCB-4 (η > 1) shows steel girder failure (Figure 11(d)), while the other three specimens FCB-1 to FCB-3 (η ≤ 1) were performed as stud failure (Figure 11(a) to (c)). The stud failure position first appeared near the row of the studs at the end of beam, closest to the beam end of the row of studs, and then the second and third rows of studs failed. The stiffness of the test beam suddenly dropped, which further leads to the actuator unloading. Concerning the failure mode of the studs under the fatigue loads, Hallam (1976) summarized three types of fatigue cracks. According to the observation, the failure mode of the studs in this test should be type A, that is, the beginning of crack formed at the stud shank and the successive crack forms through the shank and the weld collar, respectively; it is shown in Figure 11(a). In the case of the steel girder failure, an initial crack formed at the edge of a bottom flange at midspan where a defect existed due to a defect in the steel, which is shown in Figure 11(d).

Figure 11.

Failure modes of the beam specimens under fatigue loading: (a) FCB-1, (b) FCB-2, (c) FCB-3, and (d) FCB-4.

This difference occurred because when the shear connection degree was small, the shear stress amplitude of each stud was large, which easily causes a stud fatigue failure. Instead, when the shear connection degree is large, each stud is subjected to a lower shear stress condition. Thus, compared to the studs, the steel girder is more easily to fail.

Variation in mechanical properties

In this section, we mainly investigated the variation in mechanical properties of SCC beams with different shear connection degrees under fatigue loading, such as the relative slip and elastic stiffness of the beam.

Figure 12 describes the relative slip versus the number of cycles for all four types of beams under fatigue loading. The test results clearly show that the relative slip of the beam with a low shear connection degree (η ≤ 1.0) can be divided into three typical stages. The relative slip rapidly increases in the first and last stages and slowly increases in the second stage, which is consistent with the push-out fatigue test results of studs in relevant studies (Hanswille et al., 2007). For the FCB-4 specimen, as it has more shear connectors, and the failure mode is not stud failure, the relative slip is notably small in the entire process. Furthermore, a comparison of the relative slip at the support, quarter points, and midspan sections shows that for a beam with high shear connection degree (η ≥ 1.0), the relative slip at the support section is the largest; for a beam with lower shear connection degree (η < 1.0), the largest relative slip occurs at the quarter points section. In addition, the relative slip at the midspan section remains small. This variation rule is consistent with the static test results.

Figure 12.

Relative-slip variation under fatigue loading: (a) FCB-1 (η = 1.00), (b) FCB-2 (η = 0.71), (c) FCB-3 (η = 0.57), and (d) FCB-4 (η = 1.14).

The elastic stiffness of a composite beam is another important performance index in the normal service stages. The elastic stiffness is commonly defined as the load increment that corresponds to the unit deflection, that is, K = dP/dΔ. In the actual test, the measured load–deflection curve in the elastic stage is not a linear line because of the effect of the equipment error, material inhomogeneity, and other factors, resulting in the stiffness undetermined. In this article, we define it as the secant stiffness that corresponds to the upper limit value point of fatigue loads, as shown in Figure 13.

Figure 13.

Elastic stiffness variation under fatigue loading.

Figure 13 demonstrates the variation of the elastic stiffness for four types of beams with numbers of cycles. The stiffness of composite beams first increases and then slightly decreases, and finally, it sharply decreases with the increasing fatigue loading number. Moreover, the elastic stiffness has obviously positive correlation with the shear connection degree.

Analysis and prediction of the fatigue test results

Determination of the fatigue stress range

From these test results, we can divide the fatigue failure modes of the specimens into two types: steel failure and stud failure. Therefore, it is vital to discuss the fatigue life of the SCC beams with these two types of failure modes.

Various design equations of the fatigue strength of a steel girder or a stud are shown according to the individual design codes of each country.

In the design equations, the variables are the fatigue life and fatigue stress range. Other parameters can be determined based on the code standards. To determine the fatigue life, the fatigue stress range must be known.

For the welded steel girder, the fatigue stress range can be calculated from the strain gauge data at the bottom of the test beam.

For the stud connector, the acquisition of the stress amplitude is more complicated because of the lack of a strain gauge. Previous studies have indicated that the shear force of the stud is related to the slip and can be calculated using equation (3). The slip can be obtained by checking the corresponding loading stress amplitude of the test beams. Then, the shear stress of the stud can be obtained. A point to emphasize is that the selected slip value is the maximum value among all measuring points along the beams under the same loading level. Table 4 shows the stress range of the steel girder and stud connector under fatigue loading.

Table 4.

Fatigue stress range of the steel girder and stud connector.

Specimens	Steel girder		Stud connector
	Δμε	Δσ (MPa)	δ_max (mm)	δ_min (mm)	ΔV = V(δ_max) − V(δ_min) (kN)	Δτ = ΔV/(πd²/4) (MPa)
FCB-1	497	102.38	0.38	0.21	12.22	92.06
FCB-2	535	110.21	0.56	0.28	15.68	118.13
FCB-3	635	130.81	0.65	0.32	16.33	123.01
FCB-4	478	98.47	0.33	0.19	10.84	81.68

δ_max and δ_min are the measured values with the upper and lower limits of fatigue load, respectively.

Analysis and prediction of the fatigue life

Figure 15 shows the comparison between the fatigue test results and design values according to EC3 (1993-1-9: 2005), EC4 (EN 1994-2: 2005), AASHTO (2014), and Chinese code (JTG D64-2015, 2015). It is not difficult to observe that for both steel girder and stud, the AASHTO specification results in a lower estimation of their fatigue capacity than the Eurocode and Chinese code.

In Figure 14(a), the results are controlled based on the assumption of a steel girder failure; the results of FCB-1 and FCB-4 specimens are in good agreement with the design codes. However, FCB-3 and FCB-2 specimens failed to achieve the expected fatigue life because the studs failed before the steel girder acquired the fatigue strength. In Figure 14(b), we assume that the failure mode of all beam specimens is stud failure. Only FCB-4 did not acquire the expected fatigue life. The other three beams results show that the test fatigue life was longer than the calculated fatigue life. Therefore, the design codes to calculate the fatigue life of the specimens are conservative. Equation (4) is a linear interpolation analysis of the test results of the stud failure data (FCB-1 to FCB-3). The fitting determination coefficient is 0.927, which indicates a good fitting effect. According to the calculation of equation (4), the actual fatigue life of FCB-4 specimen was 2,200,000 number of cycles fewer than the expected value. Its early failure occurred because of the steel beam fracture

\log N + 5.62 \log Δ τ = 17.41

(4)

Figure 14.

Comparison between the fatigue test results and design values: (a) design codes for steel girder and (b) design codes for stud.

Both failure modes should be considered in the fatigue design of composite beams. When the shear connection degree is small, the fatigue strength of studs is considered because they will likely fail before the steel girder; when the shear connection degree is large, the actual internal stress of studs is relatively small, and the control design is the fatigue life of the steel girder. Therefore, we suggest an equation to calculate the fatigue life of composite beams considering the different connection degrees. And the equation is presented below

{\begin{matrix} {\begin{matrix} Δ σ_{R}^{m} N_{R} = Δ σ_{c}^{m} \times 2 \times 10^{6} \begin{matrix} (N_{R} \leq 5 \times 10^{6}, m = 3) \end{matrix} \\ Δ σ_{R}^{m} N_{R} = Δ σ_{D}^{m} \times 5 \times 10^{6} \begin{matrix} (5 \times 10^{6} < N_{R} \leq 10^{8}, m = 5) \end{matrix} \end{matrix} \begin{matrix} (η > 1.14) \end{matrix} \\ \log N_{R} + 5.62 \log Δ τ = 17.41 \begin{matrix} \begin{matrix} \begin{matrix} \begin{matrix} (η \leq 1.00) \end{matrix} \end{matrix} \end{matrix} \end{matrix} \end{matrix}

(5)

where N_R is the number of stress range cycles, Δσ_R is the stress range of the steel girder, Δσ_c is the fatigue strength at 2 million cycles, Δσ_D = 0.767Δσ_c, Δτ is the stress range of the stud, and η is the shear connection degree.

Discussion

This work focuses on the effect of different degrees of shear connection on the static and fatigue performances of SCC beams independently. Meanwhile, the static and fatigue capacities should be compared in detail.

Figure 15 shows the comparison between the static bearing capacity and the fatigue capacity with different shear connection degrees. It can be seen that both of them tend to decrease with the decrease in shear connection degree. However, compared with the static bearing capacity, the fatigue capacity is more sensitive to the shear connection degree, and the decreasing rate is faster and more obvious. Therefore, in the design of composite beams with partial connection, their fatigue problems require more attentions.

Figure 15.

Comparison between static and fatigue results with different shear connection degrees.

Another interesting finding is that the degree of shear connection can change the failure modes of the test beams. Figure 16 demonstrates the static failure modes and fatigue failure modes of test beams with different shear connection degrees. The static failure modes include two types: concrete crush at midspan and stud fracture. When η ≥ 0.71, the failure mode of the test beams is concrete crush; when η = 0.57, the failure mode of the test beam is stud fracture. Thus, the critical value of two types of failure modes is 0.57-0.71, which should be obtained from more tests. Similarly, there are two types of fatigue failure modes: steel girder failure and stud failure. When η = 1.14, the beam fails with the steel girder fracture. When η ≤ 1.00, the failure mode of the test beam is stud failure. Therefore, the critical range of fatigue failure modes is 1.00–1.14. Additional tests are required to determine the specific value.

Figure 16.

Static failure modes and fatigue failure modes of test beams with different shear connection degrees: (a) static failure modes of beam specimens and (b) fatigue failure modes of beam specimens.

It is generally known that the fatigue failure mode is related to the preset loading amplitude. In this study, the preset loading amplitude is deduced according to the Chinese specification and actual stress state on bridges. Therefore, the conclusions of this article are practical and universal.

Conclusion

In this article, eight SCC beams were investigated to examine the effects of different degrees of shear connection on the static and fatigue performances of the beams. The following conclusions are drawn:

In static tests, the ultimate capacity, ductility coefficient, failure modes, strain distribution of midspan section, and relative-slip distribution of SCC beams were investigated with different shear connection degrees. The ultimate capacity and ductility coefficient decrease with a decrease in shear connection degree, and the reduction is larger when the shear connection degree is smaller. The failure modes change from the concrete crush failure to stud failure with the decrease in shear connection degree. The strain distribution at the midspan section is consistent with the assumption of plane section only for the full-shear-connection composite beams. The relative slips linearly distribute when the shear connection degree is larger than 1.00. For partial-shear-connection composite beams, the relative slip along the length shows a nonlinear distribution, and the maximum slip under each level of load appears at the 1/8 point.

In fatigue tests, the effect of the shear connection degree on the fatigue life, failure mode, and other mechanical properties is studied. The fatigue life of the composite beams increases with the increase in shear connection degree. When the shear connection degree is smaller than 1.00, the fatigue life of the composite beams sharply decreases. The failure modes change from the steel girder fracture to stud failure with the decrease in shear connection degree. The relative slip of the beam with a low-shear connection degree can be divided into three typical stages under fatigue loading. However, for the beam with a shear connection degree larger than 1.00, the relative slips are notably small during the entire process. The elastic stiffness of all composite beams first increases and then slightly decreases, and finally, it sharply decreases with the increasing fatigue loading number. Moreover, the elastic stiffness has an obviously positive correlation with the shear connection degree.

The fatigue life of composite beams was analyzed and predicted under different assumptions of failure modes according to EC4, AASHTO, and the Chinese code. The result shows that both failure modes should be considered in the fatigue design of composite beams. When the shear connection degree is small, the fatigue strength of the studs will be considered because they probably fail before the steel girder; when the shear connection degree is large, the actual internal stress of studs is relatively small, and the control design is the fatigue life of the steel girder.

The comparison of the static and fatigue capacities with different shear connection degrees is discussed. The results of the comparison show that the fatigue capacity is more sensitive to the shear connection degrees, and the decreasing rate is faster and more obvious than the static bearing capacity. Therefore, in the design of composite beams with partial connection, their fatigue problems require more attentions. In addition, the issue on the critical point of the failure mode is also discussed. For static tests, the critical range of shear connection degree for two types of failure modes (concrete crush and stud fracture) is 0.57–0.71, whereas the critical range for two failure modes (steel girder failure and stud failure) in fatigue tests is 1.00–1.14. The specific value needs to be determined by more tests.

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 research described in this paper was financially supported by the Natural Science Foundation (grant no. 51278119).

References

AASHTO (2014) LRFD Bridge Design Specifications. 7th ed.Washington, DC: AASHTO.

Ban

Bradford

(2014) Flexural behaviour of composite beams with high strength steel. Engineering Structures 56(6): 1130–1141.

Bing

Qiao

Xiaoling

(2017) Deterioration in strength of studs based on two-parameter fatigue failure criterion. Steel and Composite Structures 23(2): 239–250.

Chen

Liu

Tang

et al . (2011) Influence of high-cycle fatigue on the tension stiffening behavior of flexural reinforced lightweight aggregate concrete beams. Structural Engineering and Mechanics 40(6): 847–866.

Dilena

Morassi

(2009) Vibrations of steel–concrete composite beams with partially degraded connection and applications to damage detection. Journal of Sound and Vibration 320(1–2): 101–124.

EN1993-1-9:2005 (2005) Eurocode 3: design of steel structures—Part 1-9: fatigue.

EN1994-2:2005 (2005) Eurocode 4: design of composite steel and concrete structures—Part 2: general rules and rules for bridges.

Fanaie

Esfahani

Soroushnia

(2015) Analytical study of composite beams with different arrangements of channel shear connectors. Steel and Composite Structures 19(2): 485–501.

GB50017-2003 (2003) Code for design of steel structures.

10.

Hallam

(1976) The behaviour of stud shear connectors under repeated loading. Research report no. 281. Sydney, NSW, Australia: School of Civil Engineering.

11.

Hanswille

Porsch

Ustundag

(2007) Resistance of headed studs subjected to fatigue loading: Part I: Experimental study. Journal of Constructional Steel Research 63(4): 475–484.

12.

Jimbo

Morassi

Nakamura

et al . (2012) A non-destructive method for damage detection in steel-concrete structures based on finite eigendata. Inverse Problems in Science and Engineering 20(2): 233–270.

13.

JTG D64-2015 (2015) Specifications for design of highway steel bridge.

14.

Loh

Bradford

(2006) The effects of partial shear connection in composite flush end plate joints Part I—experimental study. Journal of Constructional Steel Research 62(4): 378–390.

15.

Nagy

Schotte

Bogaert

et al . (2016) Fatigue strength application of fracture mechanics to orthotropic steel decks. Advances in Structural Engineering 19(11): 1696–1709.

16.

Oehlers

Nguyen

Ahmed

et al . (1997) Partial interaction in composite steel and concrete beams with full shear connection. Journal of Constructional Steel Research 41(2): 235–248.

17.

Ollgard

Slutter

Fisher

(1971) Shear strength of stud connectors in lightweight and normal weight concrete. AISC Engineering Journal 8: 55–64.

18.

Queiroz

Vellasco

PCGS

Nethercot

(2008) Finite element modelling of composite beams with full and partial shear connection. Steel Construction 63(4): 505–521.

19.

Rodrigues

JPC

Laím

(2014) Experimental investigation on the structural response of T, T-block and T-Perfobond shear connectors at elevated temperatures. Engineering Structures 75: 299–314.

20.

Sanghyo

Chiyoung

Jinhee

(2011) Ultimate strength of composite structure with different degrees of shear connection. Steel and Composite Structures 11(5): 375–390.

21.

Selvi

(2016) Numerical evaluation of deformation capacity of laced steel-concrete composite beams under monotonic loading. Steel and Composite Structures 20(1): 167–184.

22.

Titoum

Tehami

Achour

(2009) Effects of partial shear connection on the behavior of semi-continuous composite beams. International Journal of Steel Structures 9(4): 301–313.

23.

Vasdravellis

(2014) Shear strength and moment-shear interaction in steel-concrete composite beams. Journal of Structural Engineering 140(11): 191–209.

24.

Wannenburg

Heyns

Raath

(2009) Application of a fatigue equivalent static load methodology for the numerical durability assessment of heavy vehicle structures. International Journal of Fatigue 31(10): 1541–1549.

25.

Xiang

Yang

Zhao

(2015) Stochastic creep and shrinkage effect of steel-concrete composite beam. Advances in Structural Engineering 18(8): 1129–1140.

26.

Xing

Han

et al . (2016) Experimental and numerical study on static behavior of elastic concrete-steel composite beams. Journal of Constructional Steel Research 123: 79–92.

27.

Zhan

Zhao

et al . (2016) Interface behavior between steel and concrete connected by bonding. Journal of Bridge Engineering 21(6): 04016026.

28.

Zhou

Zhang

Zhong

et al . (2016) Shear stress calculation and distribution in variable cross sections of box girders with corrugated steel webs. Journal of Structural Engineering 142(6): 04016022.

29.

Zona

Ranzi

(2014) Shear connection slip demand in composite steel–concrete beams with solid slabs. Journal of Constructional Steel Research 102(11): 266–281.