Experimental study of precast beam-to-column joints with steel connectors under cyclic loading

Abstract

This article investigated the seismic performance of a new type of precast concrete beam-to-column joint with a steel connector for easy construction. Five interior beam-to-column joints, four precast concrete specimens, and one monolithic joint were tested under reversed cyclic loading. The main variables were the embedded H-beam length, web plate or stiffening rib usage, and concrete usage in the connection part. The load–displacement hysteresis curves were recorded during the test, and the behavior was investigated based on displacement ductility, deformability, skeleton curves, stiffness degradation, and energy dissipation capacity. The results showed that the proposed beam-to-column joint with the web plate in the steel connector exhibited satisfactory behavior in terms of ductility, load capacity, and energy dissipation capacity under reversed cyclic loading, and the performance was ductile because of the yielding of the web plate. Therefore, the proposed joint with the web plate could be used in high seismic regions. The proposed joint without the web plate exhibited similar behavior to the monolithic specimen, indicating that this joint could be used in low or moderate seismic zones. Furthermore, the utilization of the web plate was vital to the performance of this system.

Keywords

beam-to-column joint precast structure quasistatic cyclic test seismic performance steel connector

Introduction

With the development of the national economy and the advancement of society, precast concrete (PC) structures have been extensively applied in many countries due to high quality control, fast construction, and improved cost efficiency, as PC structures generate less waste material than conventional cast-in-place reinforced concrete (RC) structures (Ghayeb et al., 2017). Therefore, the modern requirements of national rapid construction and mass production are satisfied using PC structures. However, when subjected to earthquakes, PC structures exhibit extensive damage and catastrophic failure, which is mainly due to connection failures and inadequate ductility, highlighting the significance of ductile joints in PC frames (Mitchell et al., 1995). Thus, the main problem facing PC structures is increasing the ductility of the beam-to-column connections.

At present, various studies have been conducted by domestic and foreign scholars on beam-to-column joints in precast buildings within earthquake areas, which have greatly promoted the development of PC structures (Alcocer et al., 2002; Chisari and Amadio, 2014; Li et al., 2009; Nzabonimpa et al., 2018; Restrepo et al., 1995; Shufeng et al., 2018; Xiao et al., 2017; Xue and Yang, 2010). Ersoy and Tankut (1993) studied the use of welding to connect PC beam-to-column joints and concluded that the welded joints were very similar in strength, stiffness, and energy consumption to monolithic joints. Kim et al. (2013) designed and tested two welded sleeve connection specimens, and their results showed that the welded cylinder was prone to high stresses under normal stress conditions. In addition, Vidjeapriya and Jaya (2012) tested J-bolts and angle cleats and concluded that the bolts had better ductility and energy consumption. Vidjeapriya and Jaya (2013) also proposed the connection of a corbel and angle steel with a stiffening rib. Their results show that the connection with two stiffening ribs had higher strength and that the connection with a single stiffening rib had better ductility. Li et al. (2013) designed three high-efficiency ductile joints, and the tested ductility of the ductile joints was higher than that of the monolithic joint. In addition, Pan et al. (2017) designed an experimental research on three new hybrid joints with bolted connections and moved the plastic hinge from the end of the beam to the weak part of the joint to avoid adverse impacts on the joint. Their study found that these new hybrid joints had high bearing capacity and good energy dissipation capacity. Eom et al. (2016) achieved the ideal seismic performance by changing the type and shape of the reinforcement in the core zone or by reshaping the plastic hinge by reinforcing the core zone. Moreover, Yang et al. (2016) tested the flexural strength and shear properties of hybrid H-steel–PC beams. MacRae et al. (2009) reported a friction moment joint that dissipated seismic energy by adding energy-consuming metal plates in the joint zone to counteract seismic forces. To reduce distortion, Wang et al. (2018) and Kurama and Morgen (2007) introduced prestressing into fabricated joints.

Furthermore, Im et al. (2013) tested precast beam-to-column joints with U-shaped beam shells and analyzed the effect of the shelved length of the prefabricated beam end and the steel angle setting at the end of the beam on the performance of the joints. Ha et al. (2014) proposed a semi PC joint with U-shaped prestressed steel strands, which was applied in moderate seismic zones. Yuksel et al. (2015) compared and studied prefabricated joints used in industrial building structures and residential structures, wherein U-shaped reinforcement was adopted for the upper end and bottom of the two joints. Guan et al. (2016) designed a connection system with double reinforcement of anchoring and lap joints. Their experiments showed that the structure had better shear performance than monolithic joints under reversed cyclic loading. Guan et al. (2016) also designed a new joint with a grouted steel sleeve and a U-shaped steel bar. This joint removed welding, bolting, and prestressing, which greatly accelerated the construction speed. Yan et al. (2018) tested the new type of steel bar in a PC joint using a sleeve connection. Their results showed that the grouting sleeve, the steel bar diameter, and the cast-in-place zone had an important impact on the seismic resistance of the joint. Khaloo and Bakhtiari Doost (2018) created four precast reinforced concrete column-to-steel beam (RCS) joints and tested the connections under cyclic loading. The extended face bearing plates or extended cover plates were welded to shear keys embedded in the core zone. The best performance in terms of strength and joint distortion was obtained using shear keys and extended cover plates. In addition, Aninthaneni et al. (2018) studied the nonlinear hysteretic behavior of prestressed end-plate joints.

Existing studies have focused on the seismic behavior of PC structures through wet and dry connections. The results showed that the PC beam-to-column joints showed similar seismic behaviors to conventional cast-in-place concrete beam-column connections. However, the existing PC connections with many cast-in-place concrete or prestressing techniques in precast frames are complicated and time-consuming. Therefore, beam-to-column joints with easy connections should be developed. The connections in steel structures are convenient to realize, but there have been a few studies on steel connections in PC structures. Therefore, the performance of PC joints connected by steel connectors requires further study.

This article proposes a new precast interior RC beam-to-column joint used in seismically active regions. The major benefits of this new joint are faster construction, higher economic benefits, and more satisfactory seismic performance than traditional precast joints. A total of five full-scale beam-to-column joints, four of the proposed PC joints, and one conventional cast-in-place concrete joint were tested under cyclic loading. In this study, the main purpose is to evaluate the seismic behavior of the four PC joints under reversed cyclic quasistatic loading. Based on the test results, the behavior of the experimental PC joints was investigated in terms of the failure mode, load–displacement curves, ductility, skeleton curves, stiffness degradation, and energy dissipation.

Experimental program

Test specimens

Figure 1 shows the details of four PC interior beam-to-column connection specimens and one monolithic concrete joint. Figure 2 shows the skeletal pictograph of specimens. The specimen parameters were the method of connection, the embedded length of the H-beam, the usage of web plates, stiffening ribs or headed studs, and the usage of concrete in the connection part, as shown in Table 1. The joints were subjected to reversed cyclic loading. To study the behavior of the joint, the specimen was designed for the occurrence of joint failure.

Figure 1.

Details of specimen for: (a) J1–J5, (b) J2, (c) J3, (d) J4, and (e) J5.

Figure 2.

Skeletal pictograph of specimen for: (a) J1, (b) J2, (c) J3, (d) J4, (e) J5, (f) J2/J3, and (g) J2/J3.

Table 1.

Details of the test specimens.

Specimen	J1	J2	J3	J4	J5
Connection method	Monolithic	Precast	Precast	Precast	Precast
Embedded H-beam length (mm)	X	300	300	200	200
Web plate	X	PL-250 × 318 × 12	X	PL-250 × 318 × 12	X
Stiffening rib	X	X	X	X	PL-120 × 318 × 10
Headed stud	X	Headed stud	X	Headed stud	X
Pouring concrete in connection part	X	X	X	Concrete	Concrete

X indicates a test specimen without the parameter mentioned.

Figure 1 shows that the proposed PC beam-to-column joint consists of a concrete beam with an embedded H-steel beam, a concrete column with steel connectors, and a beam-to-column connection. The column sectional was 350 mm × 350 mm, the column length was 2800 mm, and 10 HRB600 reinforcement bars were used as the column longitudinal reinforcement bars. The beam length was 3550 mm, the beam sectional area was 250 mm × 400 mm, and four HRB600 reinforcement bars were used as the top and bottom beam longitudinal reinforcement bars. Grade HRB400 transverse reinforcements were placed at an interval of 100 mm in the beams, columns, and connections.

The production process of specimen J2 is described as follows. Two steel plates were connected with two horizontal connection plates and one web plate through the welding method to form a steel skeleton embedded in the precast column. There was a hole with a diameter of 100 mm in the web plate to pour the concrete of the column, which was slight influence on the joint performance due to provision of hole (Li et al., 2013). The upper longitudinal reinforcement bars of the column were welded at the top of the horizontal connection plate. The lower longitudinal reinforcement bars were welded at the bottom of the horizontal connection plate. The H-steel beam was embedded in a PC beam with a length of 300 mm inside and 200 mm outside. Then, the PC beam was connected on site to the PC column via welding to the flange and bolting to the web.

The production process of joint J3 was similar to that of specimen J2, except that there were no web plates in the PC column or headed studs in the PC beam. In specimen J4, the H-steel beam was embedded in the PC beam with a length of 200 mm both inside and outside. After prefabricating the precast column and beam, the connection part was cast from concrete on site. The other processes were the same as those used for joint J2.

In the case of joint J5, two steel plates were connected with two horizontal connection plates and four stiffening ribs through the welding method to form a steel skeleton embedded in the PC column. The upper and lower longitudinal reinforcement bars of the column were welded at the top and bottom of horizontal connection plates, respectively. The H-steel beam was embedded in a PC beam with a length of 200 mm both inside and outside. The PC beam was connected on site with the precast column via welding to the flange and bolting to the web.

Test material parameters

Concrete cubes (150 mm × 150 mm × 150 mm) were cast, which were subjected to compression tests in accordance with GB/T50152-2012. On average, the 28-day cubic strength of the concrete was 47.3 MPa for the precast columns and beams and 46.6 MPa for the connection parts.

Table 2 lists the strength and percent elongation of the reinforcement bars and steel plates, which were determined from tensile tests conducted in accordance with GB/T228-2010; the results listed in the table are an average of three samples.

Table 2.

Material properties of the reinforcement bars and steel plates.

Material	Thickness/diameter (mm)	Yield strength (MPa)	Tensile strength (MPa)	Elastic modulus (GPa)	Percent elongation
Reinforcement bar	10	502.6	636.5	207.0	24.0
	18	700.8	875.9	211.0	20.4
	22	648.5	821.8	210.0	21.2
Steel plate	6	298.3	445.3	211.2	33.8
	8	298.3	466.7	212.6	35.5
	10	312.0	450.0	212.6	32.7
	12	264.7	400.0	235.7	37.0
	16	263.0	431.3	225.1	31.8
	20	279.3	454.7	239.3	30.0

Experimental test setup and instrumentation

Figure 3 displays the test setup. During the loading process, both column ends were restrained with lateral bracings to prevent unexpected out-of-plane instability of the column. The specimens had a spherical hinge constraint, which allowed rotation only at the bottom of the column end. At the top of the column, the specimens had roller supports to simulate the unidirectional hinge condition. First, a vertical hydraulic jack with a 1000 kN capacity was used to apply a constant axial load of N = 460 kN, which was connected to the top of the column. Second, two 100T actuators were used to apply the reversed cyclic loading at the beam ends, which were installed on the rigid beam.

Figure 3.

Test setup.

The cyclic loading history adopted is presented in Figure 4. The loading history was based on the specification JGJ 101-2015 and employed load–deformation mixed control laws. The loading was repeated only once in the load-controlled stage, and then each level of displacement was repeated three times in the displacement-controlled stage. The increase in displacement adopted the integral multiple of the yield displacement, which was determined by the yielding of the beam longitudinal bars. Finally, the test was stopped when the applied load fell below 85% of the maximum load.

Figure 4.

Cyclic loading history.

The vertical force and displacement at the end of the beam were recorded by transducers inside the actuator. The axial load was measured by the load cell. The reinforcement bar strain and steel plate strain were measured by strain gauges during the test.

Test results and discussion

Failure modes of the specimens

The failure modes of each specimen are shown in Figure 5. The crack patterns of all joint specimens at failure are given in Figure 6. All of the joints exhibited shear failure in the core region according to the design purpose in this study.

Figure 5.

Failure modes of all specimens: (a) J1, (b) J2, (c) J3, (d) J4, (e) J5, (f) J2 at 69 mm, and (g) J4 at 76 mm.

Figure 6.

Crack patterns of all specimens: (a) J1, (b) J2, (c) J3, (d) J4, and (e) J5.

In the monolithic joint specimen J1, the initial cracks formed at the beam ends when loading to 50 kN, and then flexural cracks extended transversely along the beams. Diagonal cracks formed in the core area when the load reached 90 kN and then propagated with increasing load. There number of flexural cracks and diagonal shear cracks in the beams and joint zone increased after the specimen yielded. When the displacement increased, the concrete splits at the surface of the beam and column to obtain the maximum load point. Then, the concrete cover crushed severely in the core region of the joint when the displacement reached 80 mm. Finally, the joint exhibited a shear failure because of crushing of the joint at a displacement of 86 mm.

Joint specimens J2 and J4, which have web plates in the core region, exhibited similar performance and failure modes. The initial flexural cracks were found between the embedded H-steel beam and concrete beam at a load of 60 kN, and then the existing flexural cracks propagated and widened toward the beams. Shear cracks formed in the core area when loading to 90 kN. As the displacement increased, more hairline diagonal cracks were observed in the core zone. When the displacements were 70 mm for joint J2 and 76 mm for J4, concrete cover spalling occurred at the core area. When the displacement reached nearly 105 mm for specimens J2 and J4, severe concrete crushing was observed, resulting in shear failure of the specimens.

In the case of specimen J3, the initial flexural cracks were observed at 60 kN. The initial diagonal cracks formed at 90 kN in the negative direction. Specimen J3 showed similar failure modes to the monolithic joint J1. However, in the case of specimen J5, which had stiffening ribs, the concrete crushing at the column ends was less than substantial than that observed in the other precast specimens. This difference attributed to the fact that the stiffening ribs welded at the end of the column reduced the concrete spalling.

Before the displacement reached 40 mm in specimens J4 and J5, there were no cracks between the beam and column face and the connection part due to the embedded H-steel beam. Afterward, there were fewer horizontal cracks at the connection part in joints J4 and J5 than in the monolithic joint specimen J1. Moreover, a vertical crack formed in joints J4 and J5 at the interface of the precast beam and connection part after reaching the maximum load. Furthermore, there were fewer cracks in the embedded H-steel beam part in specimens J2–J4 than in the cast-in-place joint J1. However, specimen J3, which did not have a web plate or stiffening ribs, exhibited fewer cracks in the joint region than the other precast joints; hence, specimen J3 possessed the lowest shear capacity among the precast joints. At the failure stage, the concrete crushing was more severe in the precast joints J2 and J4 than in the monolithic joint J1 due to the much higher shear capacity in specimens J2 and J4, which have web plates. The precast joints J3 and J5, which do not have web plates, achieved similar shear capacity and exhibited similar failure modes to the cast-in-place specimen J1. Moreover, the embedded H-beam length, headed stud usage, and concrete usage in the connection part had little influence on the failure modes of the precast specimens.

The displacement of specimen J4 was 76 mm and the load was 210 kN, which was 1.4% and 37.2% higher than those of the specimen J5 at the failure stage. The specimen J2 exhibited 2.4% higher displacement and 39.0% larger load at 69 mm than specimen J3 at the failure stage. However, the concrete crushing in the precast joints J2 and J4 (Figure 5(f) and (g)) was slightly less severe than that in the precast joints J3 and J5 due to the much higher shear capacity in specimens J2 and J4, which have web plates, respectively. At the failure stage, the concrete crushing was more severe in the precast joints J2 and J4 than that in the precast joints J3 and J5 due to the much higher shear capacity in specimens J2 and J4, which was mainly attributed to the use of the web plates.

Hysteresis curves

The hysteresis curves provide important indexes—such as energy dissipation capacity, stiffness degradation and strength degradation—to evaluate the seismic behavior of joint specimens under reversed cyclic loading. The hysteresis curves for all joints are shown in Figure 7.

Figure 7.

Load–displacement hysteresis curves: (a) J1, (b) J2, (c) J3, (d) J4, and (e) J5.

The hysteresis curves of all the precast joints and the monolithic joint changed linearly before cracking. The hysteresis loop area was small, and all specimens dissipated little energy. After specimen yielding, the hysteresis loop area became wider with increasing load, indicating that energy dissipation increased. Then, the loads decreased after reaching the maximum bearing capacity, and the hysteresis curve became obviously pinched. The load value of the hysteresis curve in the first cycle was larger than the load value in the two following cycles, indicating strength degradation.

Figure 7 shows that the hysteresis loop area of joints J2 and J4 (i.e. specimens with web plates) was much larger than that of the monolithic joint J1 and the precast specimens J3 and J5, indicating that the web plates provide larger energy dissipation under reversed cyclic loading. Moreover, the pinching effect was less substantial in the hysteresis curves of specimens J2 and J4 than that in the curves of the other specimens. Figure 7 also shows that the rate of stiffness degradation in precast joints J2 and J4 decreased due to the role of the web plate in stitching cracks and enhancing stiffness. Similarly, the web plate appeared to make stiffness and strength loss more gradual after the maximum load. This finding indicated that the web plate enabled the specimens to exhibit more ductile failures. However, the hysteresis loop area, stiffness, and strength of specimens J2 and J4 were similar, indicating that concrete usage in the connection part and the embedded H-steel beam length had little influence on the energy dissipation and stiffness degradation.

Specimens J3 and J5 showed slightly larger hysteresis loop area than monolithic specimen J1, and the pinching phenomenon was slightly improved. These results showed that the energy dissipation of joints J3 and J5 was increased due to the steel connectors. However, there was little difference in specimens J3 and J5, indicating that the stiffening rib had a small effect on the hysteretic behavior of the joints.

The width of the hysteresis curves in the PC joints J2 and J4 with the web plate was the largest, followed by that of PC joints J3 and J5 without the web plate in steel connectors and finally by that of the monolithic joint J1. Thus, specimens J2 and J4 exhibited the best seismic performance.

Skeleton curves

The load–displacement skeleton curves of all joint joints are shown in Figure 8(a). The yield load, maximum load, yield displacement, and failure displacement are listed in Table 3.

Figure 8.

Load–displacement skeleton curves (a) load-displacement skeleton curves of all joints and (b) definitions of yield displacement and failure displacement.

Table 3.

Test results.

Specimen	Direction	P_y (kN)	P_m (kN)	Δ_y (mm)	Δ_u (mm)	μ
J1	Positive	178.63	188.00	45.15	88.19	1.95
J1	Negative	145.10	171.05	30.70	83.63	2.72
J2	Positive	189.62	216.55	35.95	91.55	2.55
J2	Negative	195.98	214.40	36.27	118.54	3.27
J3	Positive	149.08	176.45	35.19	76.72	2.18
J3	Negative	151.45	176.45	24.58	67.60	2.75
J4	Positive	203.50	224.15	40.13	101.24	2.52
J4	Negative	197.08	216.55	34.33	106.66	3.11
J5	Positive	173.94	193.40	44.73	76.61	1.71
J5	Negative	141.85	166.80	28.26	73.29	2.59

P_y: yield load; P_m: maximum load; Δ_y: yield displacement; Δ_u: failure displacement; μ: ductility (Δ_u/Δ_y).

Figure 8(a) shows that the load–displacement skeleton curves of J3 and J5 were similar to those of J1, but J5 exhibited much more rapid strength degradation than J1 and J3 in the positive direction. In the negative direction, specimen J3 showed much larger stiffness than specimens J1 and J5 before reaching the maximum load. The load–displacement skeleton curves of the tested specimen J2 were similar to those of J4. However, the maximum strength, deformation, and stiffness of specimens J2 and J4 were much greater than those of the precast joints J3 and J5 and the monolithic joint J1, especially in the negative direction, due to the web plate in the steel connectors.

Table 3 shows that the average yield load and maximum load of specimen J1 were 161.9 and 179.5 kN, respectively. The average yield loads for specimens J2, J3, J4, and J5 were 192.8, 150.3, 200.3, and 157.9 kN, respectively. The average maximum loads for specimens J2, J3, J4, and J5 were 215.5, 176.5, 220.4, and 180.1 kN, respectively. Thus, the average maximum loads of specimens J2 and J4 were 20%–25% higher than those of the other specimens. This increased maximum load was mainly attributed to the use of the web plate in the steel connector, which enhanced the load resistance capacity. Moreover, the PC joints J3 and J5 exhibited similar average yield and maximum load to specimen J1, indicating that the precast joints without the web plate in the steel connector were comparable to the monolithic joint.

Displacement ductility and deformability

The displacement ductility is a vital parameter in seismic performance, which is expressed as the ratio between the failure displacement and the yield displacement. The yield displacement is calculated by the principle of equivalent elastoplastic energy and determined by setting the equal area, as shown in Figure 8(b). The failure displacement Δ_u corresponds to 85% of the maximum load capacity. The yield displacement, failure displacement, and displacement ductility are shown in Table 3.

The average failure displacement of precast specimen J2 was 22.3% larger than that of joint J1. When the web plate was welded in the steel connector and the embedded H-steel length with headed studs increased to 300 mm, the largest increase in deformation capacity was observed. This phenomenon mainly occurred because the use of the web plate increased the deformation of the steel connector in the PC joint, thereby improving the deformation capacity of the PC joint. The average failure displacement of specimen J4 was nearly 1.2 times that of joint J1 and 1.4 times that of specimens J3 and J5. Therefore, whether concrete was poured in the connection part or the embedded H-steel length was 200–300 mm, the web plate had a major effect on the deformation capacity. However, the average yield displacements of specimens J3 and J5 were 29.9 and 36.5 mm, respectively, which were 21.2% and 3.8% lower than those of joint J1, respectively. Furthermore, the average yield displacements of specimens J3 and J5 were 16.0% and 12.8% lower than those of joint J1, respectively. This finding indicated that the PC joint connected by a steel connector without a web plate decreased the deformation behavior. The precast specimen J5 with a stiffening rib had little effect on the deformation capacity.

The average displacement ductility of joint specimen J1 was 2.34. The average displacement ductility for specimens J2, J3, J4, and J5 was 2.91, 2.47, 2.82, and 2.15, respectively, which indicated the ductile behavior of the joints. The precast joints J2 and J4 with the web plate exhibited 24.6% and 20.6% higher displacement ductility than the monolithic joint J1 because of larger deformation capacity of the former after yielding, indicating that the two tested precast joints exhibited more ductile behavior because of the web plate in the steel connector. The yield displacement of PC joint J5 with stiffening ribs, which exhibited a higher maximum load in the positive direction than in the negative direction, was larger than that of precast joint J3 when satisfying the equal energy principle. This phenomenon led specimen J5 to exhibit a much lower displacement ductility in the positive direction than specimen J3. The yield displacement of precast specimen J3 was approximately 8 mm smaller than that of the monolithic joint J1, resulting in higher by 5.6% displacement ductility on average.

Stiffness degradation

The stiffness degradation is assessed by the variation in secant stiffness and expressed as the slope of the secant line connecting the average of the maximum positive and negative load and corresponding displacement points at each load cycle (Sucuoglu, 1995). Figure 9(a) gives the secant stiffness of all specimens calculated from the hysteresis curves.

Figure 9.

Stiffness degradation of all joints: (a) stiffness degradation curves and (b) normalized stiffness degradation curves.

The initial stiffness of the monolithic joint was 8.4 kN/mm, whereas the initial stiffness of PC specimens J2, J3, J4, and J5 was 7.1, 9.8, 9.3, and 7.2 kN/mm, respectively. The PC specimens J2 and J5 exhibited lower initial stiffness than specimen J1. However, the initial stiffness of PC specimens J3 and J4 was higher than that of joint J1. After yielding, the precast joints J2 and J4 exhibited higher stiffness than the other specimens. Moreover, the trend of stiffness degradation was similar in joints J2 and J4, which was slower than that of precast specimens J3 and J5; this slowed stiffness degradation was because of the web plate in the steel connectors embedded in the joint specimen. The embedded H-steel length and concrete usage in the connection part had little influence on the stiffness degradation. The stiffness degradation of the precast joint J3 was more rapid than that of the precast joint J5 with stiffening ribs and the monolithic joint J1, especially in the early stage. The precast joint J5 exhibited similar stiffness degradation to the monolithic joint J1, but the stiffness degradation of the former became more rapid in the failure stage. Therefore, the stiffening rib could not effectively delay the stiffness degradation.

Each secant stiffness value was divided by the initial secant stiffness and normalized as a nondimensional parameter. Figure 9(b) shows the normalized stiffness degradation of all joints.

The specimen J3 exhibited more severe stiffness degradation than the other precast specimens and the monolithic joint. Furthermore, the stiffness of joint J5 deteriorated obviously faster than that of specimens J2 and J4 but slower than that of the monolithic specimen. The stiffness degradation ratio of specimen J2 was the slowest among all tested specimens, followed by precast joint J4. These results showed that the web plate had an obvious influence on the stiffness degradation of the PC joint. However, the embedded H-steel beam length, stiffening rib usage, and concrete usage in the connection part had a minor influence on the stiffness degradation.

Energy dissipation capacity and damping ratio

The joints should exhibit greater energy dissipation capacity in the inelastic stage to ensure satisfactory seismic behavior, which is also an important factor for the joint specimen under seismic loading. The cumulative energy dissipation and equivalent viscous damping ratio are calculated on the basis of hysteresis curves. The area enclosed by the hysteresis loop during a cycle represents the energy dissipation in that cycle. The cumulative energy dissipation during a particular cycle can be identified as the sum of the areas enclosed by each hysteresis loop up to that cycle. Figure 10(a) shows the equivalent viscous damping ratio (h_e) computed by equation (1)

h_{e} = \frac{S_{ABCD}}{2 π \cdot (S_{OBE} + S_{ODF})}

(1)

where S_ABCD is the area of the enclosed curve, S_OBE is the area of the triangle OBE, and S_ODF is the area of the triangle ODF. The cumulative energy dissipation versus displacement curves are shown in Figure 10(b). The equivalent viscous damping ratio versus displacement curves are given in Figure 10(c).

Figure 10.

Energy dissipation: (a) definition of energy dissipation, (b) cumulative energy dissipation curves, and (c) equivalent viscous damping ratio curves.

Figure 10(b) shows that all the joints exhibited similar energy dissipation in the early stage. However, the precast specimens J2–J5 dissipated greater energy than the monolithic specimen J1 in the late stage. Joints J3 and J5 exhibited less energy dissipation than specimens J2 and J4, which could be attributed to the lack of the web plate. Furthermore, the cumulative energy dissipation of specimens J2 and J4 was approximately three times that of the monolithic joint J1 in the failure stage.

In Figure 10(c), similar conclusions could be made for the tested joints. All precast specimens except specimen J5 dissipated more energy than the monolithic joint J1. However, in the late stage of the test, joint J5 exhibited a larger equivalent viscous damping ratio than the monolithic joint J1. The equivalent viscous ratio of the precast joints J2 and J4 always increased before being completely damaged. Furthermore, the equivalent viscous damping ratio of joints J2 and J4 was approximately 1.5 times that of the monolithic specimen J1 in the failure stage. The joint J3 exhibited higher damping ratios than the precast joint J5 and the monolithic joint J1 but lower damping ratios than the precast joints J2 and J4.

Conclusion

In this research, the seismic performance of PC beam-to-column connections with steel connectors was evaluated under reversed cyclic loading. Based on the limited research presented in this article, the following conclusions were drawn:

All the joints exhibited shear failure in the core region, which was the intended joint failure. The concrete crushing was more severe in the precast joints with the web plate than that in the monolithic joint due to the former having much higher shear capacity. The precast joints without web plates exhibited similar failure modes to the monolithic joint.

The PC beam-to-column joint with the web plate had wide hysteresis curves. In addition, the hysteresis curves of specimens J2 and J4 had much less pinching than those of the other specimens. Moreover, specimens J3 and J5 exhibited slightly larger hysteresis loop area than the monolithic connection J1, and the pinching phenomenon was slightly less substantial.

The PC beam-to-column joint with the web plate exhibited higher strength and slower stiffness degradation than the PC joints without the web plate and the monolithic specimen. The precast joints without the web plate exhibited similar strength as well as stiffness degradation to the cast-in-place joint.

Considering the ductility and deformability, the displacement ductility of the specimens with the web plate was 24.6% and 20.6% higher than that of the monolithic joint due to former exhibiting larger deformation, illustrating that these specimens can resist high seismic loading. However, the failure displacement of the PC joint with stiffening ribs was too small, which resulted in lower displacement ductility.

The equivalent viscous damping ratio of the specimens with the web plate was between 0.214 and 0.230 in the failure stage, which was approximately 1.5 times that of the monolithic joint. Thus, this joint had satisfactory energy dissipation ability. The precast joint without the web plate exhibited slightly higher equivalent viscous damping ratio, indicated a better energy dissipation ability.

The embedded H-beam length, headed stud usage, and concrete usage in the connection part had little influence on the seismic behavior of the precast joints. However, the web plate was very vital to the proposed joint.

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: This research was financially supported by the Natural Science Foundation of Hebei Province, China (E2017202278 and E2018202290).

References

Alcocer

Carranza

Perez-Navarrete

, et al. (2002) Seismic tests of beam-to-column connections in a precast concrete frame. PCI Journal 47: 70–89.

Aninthaneni

Dhakal

Marshall

, et al. (2018) Nonlinear cyclic behaviour of precast concrete frame sub-assemblies with “dry” end plate connection. Structures 14: 124–136.

Chisari

Amadio

(2014) An experimental, numerical and analytical study of hybrid RC-encased steel joist beams subjected to shear. Engineering Structures 61: 84–98.

Eom

Hong-Gun

Park AMA

Hwang

, et al. (2016) Plastic hinge relocation methods for emulative PC beam–column connections. Journal of Structural Engineering 142: 04015111.

Ersoy

Tankut

(1993) Precast concrete members with welded plate connections under reversed cyclic loading. PCI Journal 38: 94–100.

Ghayeb

Razak

Sulong

NHR

(2017) Development and testing of hybrid precast concrete beam-to-column connections under cyclic loading. Construction and Building Materials 151: 258–278.

Guan

Guo

Xiao

, et al. (2016) Experimental study of a new beam-to-column connection for precast concrete frames under reversal cyclic loading. Advances in Structural Engineering 19: 529–545.

Guan

Jiang

Guo

, et al. (2016) Development and seismic behavior of precast concrete beam-to-column connections. Journal of Earthquake Engineering 22: 234–256.

Kim

Lee

, et al. (2014) Performance evaluation of semi precast concrete beam-column connections with U-shaped strands. Advances in Structural Engineering 17: 1585–1600.

10.

Park

Eom

(2013) Cyclic loading test for reinforced-concrete-emulated beam-column connection of precast concrete moment frame. ACI Structural Journal 110: 115–126.

11.

Khaloo

Bakhtiari

Doost R

(2018) Seismic performance of precast RC column to steel beam connections with variable joint configurations. Engineering Structures 160: 408–418.

12.

Kim

Cho

Lee

(2013) Structural performance evaluation of circular steel bands for PC column–beam connection. Magazine of Concrete Research 65: 1377–1384.

13.

Kurama

Morgen

(2007) Seismic design of friction-damped precast concrete frame structures. Journal of Structural Engineering 133: 1501–1511.

14.

Kulkarni

Lai Leong

(2009) Seismic performance of precast hybrid-steel concrete connections. Journal of Earthquake Engineering 13: 667–689.

15.

Wan

, et al. (2013a) Experimental study on connections of RC beam-composite column with openings on steel tube. Journal of Harbin Institute of Technology 45: 63–70.

16.

Gao

(2013b) Experimental study on high ductile joints for precast RC frame. Journal of Central South University (Science and Technology) 44: 3453–3463.

17.

MacRae

Clifton

Mackinven

(2009) The sliding hinge joint moment connection. Bulletin of the New Zealand National Society for Earthquake Engineering 43: 202–212.

18.

Mitchell

DeVall

Saatcioglu

, et al. (1995) Damage to concrete structures due to the 1994 Northridge earthquake. Canadian Journal of Civil Engineering 22: 361–377.

19.

Nzabonimpa

Hong

Kim

(2018) Experimental and non-linear numerical investigation of the novel detachable mechanical joints with laminated plates for composite precast beam-column joint. Composite Structures 185: 286–303.

20.

Pan

Zhou

, et al. (2017) Experimental and numerical investigations of seismic performance of hybrid joints with bolted connections. Journal of Constructional Steel Research 138: 867–876.

21.

Restrepo

Park

Buchanan

(1995) Tests on connections of earthquake resisting precast reinforced concrete perimeter frames of buildings. PCI Journal 40: 44–61.

22.

Shufeng

Qingning

Hao

, et al. (2018) Experimental study of a fabricated confined concrete beam-to-column connection with end-plates. Construction and Building Materials 158: 208–216.

23.

Sucuoglu

(1995) Effect of connection rigidity on seismic response of precast concrete frames. PCI Journal 40: 94–103.

24.

Vidjeapriya

Jaya

(2012) Behaviour of precast beam-column mechanical connections under cyclic loading. Asian Journal of Civil Engineering 13: 233–245.

25.

Vidjeapriya

Jaya

(2013) Experimental study on two simple mechanical precast beam-column connections under reverse cyclic loading. Journal of Performance of Constructed Facilities 27: 402–414.

26.

Wang

Marino

Pan

, et al. (2018) Experimental study of a novel precast prestressed reinforced concrete beam-to-column joint. Engineering Structures 156: 68–81.

27.

Xiao

Ding

Zhang

(2017) Structural behavior of a new moment-resisting DfD concrete connection. Engineering Structures 132: 1–13.

28.

Xue

Yang

(2010) Seismic tests of precast concrete, moment-resisting frames and connections. PCI Journal 22: 102–121.

29.

Yan

Chen

Xie

(2018) Seismic experimental study on a precast concrete beam-column connection with grout sleeves. Engineering Structures 155: 330–344.

30.

Yang

Seo

Hong

(2016) Cyclic flexural tests of hybrid steel–precast concrete beams with simple connection elements. Engineering Structures 118: 344–356.

31.

Yuksel

Karadogan

Bal

, et al. (2015) Seismic behavior of two exterior beam–column connections made of normal-strength concrete developed for precast construction. Engineering Structures 99: 157–172.