Seismic performance of fabricated continuous girder bridge with grouting sleeve-prestressed tendon composite connections

Abstract

The seismic performance of a fully fabricated bridge is a key factor limiting its application. In this study, a fiber element model of a fabricated concrete pier with grouting sleeve-prestressed tendon composite connections was built and verified. A numerical analysis of three types of continuous girder bridges was conducted with different piers: a cast-in-place reinforced concrete pier, a grouting sleeve-fabricated pier, and a grouting sleeve-prestressed tendon composite fabricated pier. Furthermore, the seismic performance of the composite fabricated pier was investigated. The results show that the OpenSees fiber element model can successfully simulate the hysteresis behavior and failure mode of the grouted sleeve-fabricated pier. Under traditional non-near-fault ground motions, the pier top displacements of the grouting sleeve-fabricated pier and the composite fabricated pier were less than those of the cast-in-place reinforced concrete pier. The composite fabricated pier had a good self-centering capability. In addition, the plastic hinge zones of the grouting sleeve-fabricated pier and the composite fabricated pier shifted to the joint seam and upper edge of the grouting sleeve, respectively. The composite fabricated pier with optimal design parameters has good seismic performance and can be applied in high-intensity seismic areas; however, the influence of pile-soil interaction on its seismic performance should not be ignored.

Graphical abstract

Keywords

seismic performance / continuous girder bridge / grouting sleeve-prestressed tendon composite connections / grouted sleeve connection / design parameters

1 Introduction

Assembly technology has been promoted to shorten the construction period, improve construction quality, and reduce construction cost [1,2]. Assembly technology is widely applied to bridge superstructures such as prefabricated bridge decks, T-beams, and box girders [3]. The implementation of prefabrication and assembly technologies in the substructure of bridges is an essential requirement for realizing a fully fabricated bridge. However, presently, prefabricated pier systems mostly adopt socket and other wet joint connections and are mostly used in low-intensity seismic regions.

Dry connection methods (such as prestressed, grouted sleeve (GS), and grouting bellows connections) can achieve the advantages of fabricated concrete structures in a very real sense, including rapid and green construction, and factory fabrication. However, the seismic performance of fabricated components with dry connections is significantly different from that of cast-in-place (CIP) concrete components. The GS connection has a simple structure and unambiguous force transfer, and the seismic performance of prefabricated concrete components with GS connections has been widely studied. Qu et al. [4] verified that a fabricated pier with a GS connection (SP) and a fabricated pier with a grouted bellows connection behaved similarly in quasi-static cyclic tests. The energy consumption capacity, stiffness, and residual displacement of the two piers are very similar. Li et al. [5] and Wang et al. [6] indicated that the seismic performance of an SP has poor energy consumption capacity and ductility. The introduction of a GS in plastic hinge areas can easily cause local stiffness changes in the GS area, which leads to upward or downward movement of the plastic hinge [7]. To reduce the impact of the introduction of a GS, Haber et al. [8] described a method of changing the plastic hinge location using a new type of connection element. Al-Jelawy et al. [9] proposed transferring plastic hinges to components with GS connections to avoid adverse earthquake damage. Nishiyama and Wei [10], Li et al. [11], and Wang et al. [12] introduced prestressed tendons (PTs) into GS connections and investigated the seismic performance of prefabricated concrete piers with GS-PT composite connections (PSPs). They found that the bearing and self-reset capacity of the SP increased.

The stress conditions of the SP and PSP are much more complex than those of the CIP pier, owing to the introduction of the prefabricated joints. Meanwhile, the complex stress conditions increase the difficulty of a theoretical analysis of their seismic performance. The proposed analysis methods, including the single plastic hinge model [13–15], two-plastic-hinge model [16], and simplified single-opening and double-opening calculation models [17], have certain limitations [18,19]. The refined finite element method can clearly describe the local cracking, crushing of the concrete, and time-dependent stress of the components, particularly the opening and closing processes at the joint seam. Nevertheless, many unsolved problems restrict the accuracy and efficiency of a refined finite element analysis, including the contact nonlinearity at the joint seam, bond-slip nonlinearity between the concrete and reinforcement, material nonlinearity of the concrete, and geometric nonlinearity caused by failure at the joint seam [20–23]. Compared with the refined finite element model, the fiber element model (FEM) has the advantages of a good convergence rate and calculation efficiency. Wang et al. [24] simulated a joint seam using plain concrete columns of equal heights. This simulation method accurately predicted the hysteresis behavior of the prefabricated pier before reaching the peak load. However, it does not consider the crushing and failure of the concrete. Hence, a non-length element was used to simulate the opening and closing of the joint seam. However, for a prefabricated pier with bonded prestressed connections, the non-length element cannot effectively model the bond-slip effect. Therefore, some scholars have modified the elasticity modulus of the reinforcement to simulate the bond-slip relationship [25–27]. Tazarv and Saiidi [26] adopted a series of springs to construct a modified elasticity modulus to simulate the bond-slip relationship between the pier and basement. Previous studies show that the FEM can simulate the load–displacement hysteresis behavior of a fabricated pier. The key point in using the FEM is to effectively simulate the damage process at the joint seam and the bond-slip effect between the concrete and reinforcement.

A comparative finite element analysis of the seismic response of continuous girder bridges with fabricated piers was conducted [28]. Zhao et al. [29] conducted a series of numerical simulations to study the seismic response of bridges with fabricated and CIP piers and found significant differences. Royero et al. [30] also established a three-dimensional numerical model of a fabricated bridge and conducted a finite element analysis. They reported that the constitutive model of the material, simulation of the connection of fabricated components, and boundary conditions among the components had a significant effect on the accuracy of the simulation. Relevant achievements can provide a reference for the analysis, design, and investigation of fully fabricated bridges. However, the seismic response of fabricated bridges to the laws of earthquake parameters and the design parameters of fabricated piers or bridges, remain to be investigated.

In this study, the FEM of PSPs was first built and verified. Subsequently, numerical analysis models of three continuous girder bridges with different types of piers were built: CIP piers (RC-CBs), GS-fabricated piers (SP-CBs), and GS-PT composite fabricated piers (PSP-CBs). Several non-near-fault ground motions (non-NFGMs) and near-fault ground motions (NFGMs) were chosen as inputs. The seismic responses of the three types of bridges were analyzed through an elastoplastic time-history analysis. The results were used to systematically investigate the impact of the parameters on the seismic performance of the PSP-CB.

2 Finite element model

2.1 Fabricated pier with grouting sleeve-prestressed tendon composite connection

The effective pier height was 3.2 m. C40 concrete was used for the pier with a design axial compressive strength of 19.1 MPa. HRB400 was used for the longitudinal rebar and HRB300 for the stirrup. The cross-section of the pier was 530 mm × 530 mm. The effective shear-span ratio of the pier was 6.04. The relevant axial compression ratio (ACR) was 0.06 (the equivalent axial force was 322 kN). The diameters of the longitudinal rebar and stirrup were 16 and 8 mm, respectively. The longitudinal rebars were arranged symmetrically around the sides of the section. The stirrups at the bottom of the pier were encrypted. GSs with a length of 328 mm were used to connect the pier body and basement and were reserved at the bottom of the pier body. Four 1 × 7 (seven strands) steel strands were applied to the unbonded PTs.

According to recent findings [31], when ACR > 0.2, the deformation capacity of the component may decrease with increasing ACR. Moreover, when the additional axial compression ratio (AACR) provided by the PT is relatively small (< 0.03), the effective role of the PT decreases [32]. Accordingly, the AACR generated by the unbonded PTs was set to 0.10. Therefore, the PT not only has self-resetting capability, but also has little impact on the failure mechanism of the components [7]. In addition, when the PT is arranged near the outer edge of the section, it provides a more significant self-resetting force. In other words, under a certain ACR and arrangement of the outer PT, a large self-resetting force can be obtained [33]. Thus, in this study, unbonded PTs were arranged at the inner edges of the stirrups and longitudinal rebars.

The nonlinear FEM was established using OpenSees 3.3.0 software. The FEM and fiber section divisions of the pier are shown in Fig.1. Concrete02 and Concrete01 were used to simulate the mechanical behaviors of the concrete of the pier segment and segment joint, respectively [12]. Steel02 was used to simulate the mechanical behavior of the rebars and PTs. The mechanical behaviors of the materials of the FEM are listed in Tab.1 and Tab.2.

The unbonded PTs were anchored at the two ends (pier top and bottom) of the model pier and were not in contact with the pier body. Therefore, the strain of the unbonded PTs did not correlate with that of the concrete. Consequently, the model pier and unbonded PTs were individually established [34,35]. Truss elements were selected to simulate the effectiveness of the unbonded PTs. In the FEMs, four rigid outriggers with the actual distance between the anchorage points and the geometric center of the pier body were first built at the pier top and bottom. One end of each of the outriggers was connected to the pier top or bottom. The truss elements were then constructed and anchored to the other end of each outrigger. Finally, the initial stress of each of the truss elements was set to be consistent with the prestressing force of the PTs, as illustrated in Fig.1.

Moreover, recent findings have shown that the GS can provide effective connection behavior, and can remain elastic during the entire loading process [36–38]. Hence, the mechanical model of the GS was chosen as the elastic model with a constant elastic modulus of 2.0 × 10⁵ MPa. The GS section area was equal to the actual area. In addition, no obvious sliding occurred in the GS, grouting material, or rebars [1,19]. Therefore, the bond-slips among the grouting materials, rebars, GSs, and concrete were not considered. The GS was considered to be part of the fiber members of the pier element. The pier element was simulated using the nonlinear beam−column element. Furthermore, a zero-length element was inserted at the upper edge of the GS (UE-GS) and at the joint seam of the PSP to model the bond-slip relationship between the concrete and longitudinal rebars. The Bond SP01 model was used to simulate the mechanical behavior of a zero-length element [39]. The skeleton and hysteresis curves of the Bond SP01 model are shown in Fig.2. The meaning of each parameter in Fig.2 are explained in Ref. [39].

The parameters that are required to construct the rebar stress versus slip relationship in OpenSees platform are σ_y, σ_u, S_y, S_u, b, and R. {\sigma }_{\mathrm{y}} and {\sigma }_{\mathrm{u}} represent the yield and ultimate strengths of the rebar, respectively. S_{\mathrm{y}} is the rebar slip at the member interface under the yield stress, which can be calculated using Eq. (1).

(1)S_{\mathrm{y}}=2.54{\left[\frac{d_{\mathrm{b}}{f}_{\mathrm{y}}}{8437\sqrt{f_{\mathrm{c}}^{\prime }}}\left(2\alpha +1\right)\right]}^{\frac{1}{\alpha }}+0.34,

where d_b is the diameter of the rebar; f_y is the yielding strength of the rebars; f_{\mathrm{c}}^{\prime } is the compressive strength of concrete for adjacent connection components; \alpha is a parameter used to simulate the local bond-slip relationship between concrete and the rebars, which can be set to 0.4 [40]. The main parameters of the constitutive model of Bond SP01 are presented in Tab.3. S_{\mathrm{u}} is the rebar slip when the rebars are fractured (recommended value: 30S_y–40S_y); b is the initial hardening ratio (recommended value: 0.3–0.5); and R is the pinching factor (recommended value: 0.5–1.0).

To simplify the FEM, the weight of the pier body and dead loads at the pier top are applied to the joint on the pier top as a concentrated mass.

2.2 Verification test design

The design parameters of the PSP specimen were identical to those of the FEM described in Subsection 2.1. The configuration of the PSP specimen is illustrated in Fig.3. The dimensions of the loading end were 800 mm × 800 mm × 400 mm. The dimensions of the cover beam were 1500 mm × 1200 mm × 560 mm. The layout of the loading device is illustrated in Fig.4.

To verify the material performance of the specimens, the mechanical properties of the concrete, longitudinal rebar, stirrup, and GS were investigated. The compressive strength of the concrete cube [41] is 48.65 MPa. The test results for the rebar and GS-rebar components are listed in Tab.4. The prism compressive strength [41] of the grouting material is 115.3 MPa. According to the experimental results, the mechanical behavior of all the materials satisfied the design requirements.

During the experiment, strain and displacement sensors were used to measure the stress and displacement at the measurement points, respectively. The measurement points are shown in Fig.5. A vertical force of 322 kN was loaded on the pier top to provide a stable ACR using a hydraulic jack. Quasi-static experiments were then performed on the specimens. The displacement was loaded according to the drift level. The loaded drift levels were 0.125%, 0.25%, 0.50%, 0.75%, 1.0%, 1.25%, 1.5%, 2.0%, 2.5%, 3.0%, 3.5%, 4.0%, 4.5%, 5.0%, 5.5%, and 6.0%. Three loading cycles were performed at each drift level. The experiment was stopped when the longitudinal rebar fractured or when the bearing load was less than 85% of the peak load.

2.3 Model verification

2.3.1 Load–displacement curve

The numerical load–displacement hysteresis and skeleton curves are compared with the experimental results in Fig.6. The numerical and experimental hysteresis curves are almost the same, particularly at low loading displacements. During the positive loading process, the numerical and experimental results for the bearing capacity of the PSP were almost the same. However, during the negative loading process the horizontal load capacity calculated from the numerical analysis differed significantly from the experimental values owing to the asymmetry of the experimental results. In addition, the numerical hysteresis curves were plumper, and the pinch effect was not significant, compared with the experimental results. This may be because the numerical analysis model only carried out one loading cycle.

2.3.2 Ductility coefficient

The positive and negative displacement ductility coefficients obtained from the numerical analysis and experiments are listed in Tab.5 [42].

As shown in Tab.5, the ductility coefficient of PSP can satisfy the requirements of the code [41]. During the positive loading process, the displacement ductility coefficients obtained from numerical analysis were similar to the experimental values. However, the numerical model cannot accurately predict the cumulative damage to the material in the negative loading direction. This may be because the constitutive model of the rebar adopted in this study is uniaxial ideal elastoplastic, and the change in the performance of the rebar after degradation is not considered.

2.3.3 Rigidity degradation

The K_S/K₀–displacement curves of the PSP are illustrated in Fig.7, which shows that the rigidity of the specimen decreases rapidly, and the descending speed decreases. When the loaded drift is 5%, the rigidity of the specimen decreases to 10% of its initial value and the bearing capacity declines significantly. Similar to the results shown in Tab.5, during the positive loading process the numerical K_S/K₀–displacement curves of PSP match well with the experimental results. However, during the negative loading process, the degradation speed of the rigidity and the declining speed of the bearing capacity of the numerical results are more rapid than those of the experimental results. This may be because the constitutive model of the concrete adopted in this study degrades faster than the experimental model after reaching the peak strength. In addition, compared with the experimental result, the yielding displacement of the numerical result is smaller, and the numerical model reaches the yield point sooner, therefore, the rigidity of the numerical model degrades faster.

2.3.4 Cumulative energy consumption

A comparison of the cumulative energy consumption is shown in Fig.8.

As Fig.6 and Fig.8 show, the cumulative damage to the material cannot be accurately reflected in the FEM. The constitutive model of the rebar is a uniaxial ideal elastoplastic. The numerical hysteresis curves are plumper than the experimental curves, which leads to higher cumulative energy consumption/dissipation.

Moreover, the proposed numerical method was verified through the results from a shaking table test of a 1/6 scaled PSP model [12]. The results show that the proposed numerical method can be used to analyze the seismic behavior of PSP.

3 Seismic performance

3.1 Continuous girder bridge model

The bridge prototype was a 3 m × 25 m continuous girder bridge. The section of the main girder was a single-box, double-cell box girder built with C40 prestressed concrete. The beam height was 1.5 m. The widths of the top and bottom box girders were 12 and 8 m, respectively. The substructure of a continuous girder bridge is a double-column pier with a rectangular section. The pier height was 10 m, and the cap was 2.2 m. C40 concrete was used for the pier. HRB400 was used for the longitudinal rebars and stirrups of the pier. The reinforcement ratio and volume stirrup ratio of the pier were 1.5% and 1.79%, respectively. To study the seismic performance of the RC-CBs, SP-CBs, and PSP-CBs, a numerical analysis model of three fabricated continuous girder bridges with different types of piers was built. Except for the introduction of the fabricated construction (including the joint seam, GS, and PT), the design parameters of the piers were the same. The superstructure and bearing systems were also identical.

Using Subsection 2.1, the FEM of the RC, SP, and PSP piers were established. The FEM of the RC and SP piers are shown in Fig.9. For simplification, the parameters of the three bridge piers were selected as design values. The determination or selection method of the parameters of the bridge piers was the same as that of PSP, as described in Subsection 2.1. Every 1 m of the pier was considered as one element. The confined concrete of the pier was divided into 10 × 10 elements. The fabricated pier was divided into two segments with lengths of 5 m each. Zero-length elements were inserted at the UE-GS, at the joint seam for SP and PSP, and at the pier bottom for the RC pier, to model the bond–slip relationship between the concrete and the rebar. The Bond SP01 model was used to simulate the mechanical behavior of the zero-length elements. The main parameters of the Bond SP01 model used in the bridge piers are listed in Tab.3. The load–displacement hysteresis curves of the three bridge piers are shown in Fig.10. The numerical results are consistent with the existing experimental results [7], and the proposed numerical method was applied to simulate the RC, SP, and PSP piers.

An elastic beam−column element was used to simulate the main girder and cover beam, and the cross-sectional eigenvalues of the box girder and cover beam were obtained. The main girder and gap were simulated by uniaxial elastic materials. Every 1 m of the main girder was considered as one element. A concentrated mass was adopted to model the weight of the main girder and other dead loads. The secondary dead load was 30 kN/m and the mass of each element was assigned to the nodes on both sides. The main girder was set at the center of the cover beam and connected to the cover beam through supports. Each cover beam was divided into four elements. This study has concentrated mainly on the seismic response of prefabricated bridges without considering the effect of the pile−soil interaction (PSI). First, the bottom of each pier is fixed.

All the supports of each continuous girder bridge are pot-type rubber bearings, among which the supports of piers 1#, 3#, and 4# were sliding supports along the longitudinal direction, whereas the support of pier 2# was a fixed bearing. A zero-length element with a bilinear hysteresis model (hardening material) is used to simulate the support. The constructive model of the sliding support is shown in Fig.11, where F_max is the critical friction of the support that is calculated as follows:

(2)F_{max}=\mu N=x_{y}K,

where \mu is the sliding friction coefficient, N is the gravity of the superstructure of the bearing support, x_y is the critical horizontal displacement of the support, and K is the initial stiffness of the support. The parameters of each bearing are listed in Tab.6. The FEM of the PSP-CB is shown in Fig.11.

To verify the FEMs, a comparison of the first three dynamic parameters between the design value (based on the Midas platform) and the OpenSees value is presented in Tab.7. The first three dynamic parameters of the RC-CB calculated by the OpenSees platform are consistent with the design parameters. The frequency error of the bridges between the OpenSees model and the Midas model is within 5%. The dynamic characteristics of the three different bridges calculated by the OpenSees platform are also consistent with the previous studies [43]. Moreover, as shown in Fig.12, the first three modes of the Midas and OpenSees models were the same. Hence, the OpenSees model can be used for further seismic analysis.

3.2 Analysis cases

The bridge prototype was located at site type II. The fortification intensity and category of the bridge were 8° and B-type, respectively. Based on JTG/T 2231-01-2020 [44], six traditional non-NFGMs, and six NFGMs were used to clarify the seismic response of bridges with CIP and prefabricated piers. The traditional non-NFGMs were the Taft-21, mel_90, Sun_10_nor, Tal_280, KOBE_90, and artificial waves (generated using the code’s design response spectrum). The NFGMs were the Imperial, Northridge-01, Parkfield-02_CA3, San Salvador, Coyote Lake, and CHICHI waves. The peak ground acceleration (PGA) of each earthquake record was determined using Eq. (3).

(3)a_{\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{k}}g=C_{\mathrm{i}}{C}_{\mathrm{s}}{C}_{\mathrm{d}}A,

where C_{\mathrm{i}}, C_{\mathrm{s}}, and C_{\mathrm{d}} are the seismic importance, site coefficient, and damping adjustment factors, respectively; A is the peak acceleration of the basic ground motion; g is the gravity acceleration, which is equal to 9.8 m/s²; and a_{\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{k}} is the PGA of the earthquake wave. The response spectra of each earthquake record and code response spectrum are shown in Fig.13. After the ground motions were selected, an elastoplastic time-history analysis of the three continuous girder bridges was conducted under uniform excitation. For convenience, all the seismic waves were input in the longitudinal direction.

3.3 Seismic performance of continuous girder bridges

3.3.1 Non-near-fault ground motions

3.3.1.1 Hysteresis curves

Fig.14 shows the hysteresis curves of the fixed pier for the RC-CB, SP-CB, and PSP-CB under typical non-NFGMs, including the shear force-displacement curve of the pier bottom and the moment-curvature curves of the joint seam, GS position, and UE-GS.

As Fig.14 shows, the initial stiffnesses of the specimens improves with the introduction of GS and PTs. The stiffnesses of the fixed SP-CB and PSP-CB piers are higher than those of the RC-CB pier. The maximum shear forces of the fixed pier bottoms of the RC-CB, SP-CB, and PSP-CB are 1188, 1251, and 1600 kN, respectively. The pier-bottom shear forces of the SP-CB and PSP-CB increase. The residual displacements of the fixed piers of the RC-CB and SP-CB are higher than that of the PSP-CB. This indicates that the fabricated pier with GS-PT composite connections has a certain self-reset capability.

In addition, as Fig.14 shows, plastic deformation mainly occurs at the pier bottom and UE-GS. The traditional plastic-hinge zone shifts compared with the fixed pier of the RC-CB, however, concrete at the GS position remains elastic. In addition, the curvatures of the joint seams for the SP-CB and PSP-CB increase by 163.9% and 290.8%, respectively, compared with that of the fixed pier of the RC-CB, and the curvatures of the UE-GS for SP-CB and PSP-CB increase by 124.5% and 154.6%, respectively. By contrast, the curvatures of the GS positions for SP-CB and PSP-CB decrease by approximately 98%. Therefore, the shifting of the plastic hinge area and the concentrated damage caused by the introduction of the GS should not be ignored when prefabricated piers with GS connections (including GS-PT composite connections) are used.

3.3.1.2 Displacement response of pier top

Fig.15 shows the displacement response of the pier top for the three bridges along the longitudinal direction under a typical non-NFGM.

As Fig.15 shows, the peak displacements of the fixed piers of the three bridges exhibit similar behavior under a typical non-NFGM. The displacement response of the fixed PSP-CB pier decreases after 25 s, compared with those of the RC-CB and SP-CB. Furthermore, the residual displacements of the RC-CB, SP-CB, and PSP-CB fixed piers are 8.8, 16.8, and 8.5 mm, respectively. The residual displacement of the SP-CB fixed pier increases significantly. In contrast, the residual displacement of the PSP-CB fixed pier was the lowest. This may be because the damage to the SP-CB fixed pier is more concentrated and severe, leading to a large residual displacement. Meanwhile, the PSP-CB has a self-reset capability. The introduction of PTs can increase the reparability of the PSP-CB and reduce the degree of damage. Tab.8 lists the peak displacements of the fixed pier for the three bridges under non-NFGMs.

As shown in Tab.8, the mean value of the peak displacement of the fixed pier decreases (< 8%) for SP-CB and PSP-CB under non-NFGMs, compared to that of the RC-CB. However, the local degrees of damage at the joint seam and UE-GS for SP-CB and PSP-CB are more severe than that for RC-CB (shown in Fig.14). The probable reason for this is that the formation of the plastic hinge results in the main lateral plastic deformation of the CIP pier. Nevertheless, the lateral plastic deformation of the fabricated concrete pier comprises the plastic deformation of the joint seam and the UE-GS. Although the local damage to the joint seam and UE-GS is much more severe, the plastic-hinge length is much shorter. Hence, although the degrees of local damage of the SP-CB and PSP-CB are more severe, the displacements are smaller. This indicates that the ductility of the fabricated pier is significantly lower than that of the CIP pier. Moreover, the increase in the initial stiffness of the SP-CB and PSP-CB (due to the GS) leads to a decrease in the elastic pier top displacement. In summary, concentrated damage at the joint seam, local destruction, and shifting of the plastic hinge should be noted in the actual application of fabricated piers.

3.3.2 Near-fault ground motions

3.3.2.1 Hysteresis curves

Fig.16 presents the hysteresis curves of the fixed piers of the three bridges under a typical NFGM, including the shear force−displacement curve of the pier bottom, and moment−curvature curves of the joint seam, GS position, and UE-GS.

As Fig.16 shows, the initial stiffnesses of the SP-CB and PSP-CB fixed piers are higher than those of the RC-CB, which is similar to the dynamic response law of bridges under non-NFGMs. In addition, for the SP-CB and PSP-CB fixed piers, plastic deformation mainly occurs at the joint seam and UE-GS, whereas the concrete at the GS position remains elastic. The plastic-hinge area is shifted for the SP-CB and PSP-CB. The curvature of the joint seams for the SP-CB and PSP-CB increase by 35.6% and 82.9%, respectively, compared with that of the RC-CB. The curvatures of the UE-GS for the SP-CB and PSP-CB increase by 106% and 186%, respectively. The plastic-hinge area shift and concentrated damage caused by the introduction of the GS connection should not be ignored. Under NFGMs, the shear–displacement hysteresis curves and moment−curvature curves of the SP-CB and PSP-CB fixed piers are plumper. Local plastic damage is much more severe. It should be noted that when the plastic damage is minor (the pier top displacement is less than 50 mm, and the pier bottom curvature is less than 1 × 10⁻⁵ mm⁻¹), the curvature of the joint seam and UE-GS are significantly reduced after setting the PTs. By contrast, when the plastic damage is severe (plastic displacement > 50 mm), the PTs increase the pier bottom curvature. This is because when the degree of damage is minor, the PTs provide a certain bending resistance moment and decrease the pier deformation. However, when the degree of damage is severe, the PTs increase the degree of damage of the pier (similar to the impact of the ACR increase).

3.3.2.2 Displacement response of pier top

Fig.17 shows the displacement response of the pier top for the three bridges along the longitudinal direction under a typical NFGM.

As Fig.17 shows, the peak displacements of the SP-CB and PSP-CB fixed piers are smaller than that of the RC-CB under a typical NFGM. The residual displacements of the SP-CB and PSP-CB fixed piers decrease significantly. The residual displacements of the RC-CB, SP-CB, and PSP-CB fixed piers are 11.31, 8.15, and 2.83 mm, respectively. The PSP-CB has a good self-reset capability. Tab.9 lists the peak displacements of the fixed piers of the three bridges under NFGMs.

As shown in Tab.9, under NFGMs, the mean values of the peak displacements of SP-CB and PSP-CB decrease by 9.48% and 18.86%, respectively, compared with the peak displacement of RC-CB. It should be noted that there are significant differences in the pier top peak displacements among the three bridges under the different NFGMs. As shown in Fig.13(b), when the spectrum value of the NFGMs increases significantly for periods longer than 1 s (Imperial wave, CHICHI wave), the pier top displacements of the fabricated bridges (SP-CB and PSP-CB) decrease significantly. This indicates that the degree of damage of SP-CB and PSP-CB is less under NFGMs with a higher spectrum over a long period. The degree of damage of the RC-CB is much more severe under NFGMs with a higher spectrum over a long period. Conversely, the pier-top displacements of SP-CB and PSP-CB are higher than that of RC-CB under excitation with a smaller spectrum over a long period. In this case, the damage to the fabricated bridges is substantially more severe because of the severe damage occurring in the joint seam.

Based on the above analysis, the damage mode of the fabricated piers changes and the plastic hinge zone shifts compared to that of the RC-CB. The plastic deformation at the joint seam and UE-GS is more obvious for the fabricated piers. Concentrated damage occurs at the pier bottom under the NFGMs. In addition, the local damage at the joint seam and the plastic hinge development at the UE-GS for the fabricated piers are much more evident under NFGMs compared with those under the non-NFGMs. The dynamic response of the fabricated bridges is significantly influenced by the spectral characteristics of the NFGM. The NFGM with a higher spectrum over a long period has a greater influence on the response of the RC-CB, compared with the SP-CB and PSP-CB. The PSP-CB exhibits better seismic performance under non-NFGM and NFGM conditions.

4 Influence of design parameters on seismic performance

4.1 Prestress

Fig.18 shows the variation curves of the main seismic response of the PSP-CB with different AACRs.

As Fig.18 shows, the pier-top displacement of the PSP-CB decreases slightly as the prestress increases. When the AACR increases from 5% to 15%, the mean value of the pier top peak displacement decreases from 84.7 to 79.7 mm (−5.9%) under non-NFGMs. The mean value of the pier top peak displacement decreases from 148.4 to 138.8 mm (−6.5%) under NFGMs. The bending moment increases with an increase in the prestress. The mean value of the pier bottom bending moment increases from 1.59 × 10⁴ to 1.84 × 10⁴ kN·m (+15.2%) as the AACR increases from 5% to 15% under non-NFGMs. While the mean value of the pier bottom bending moment increases from 1.90 × 10⁴ to 2.26 × 10⁴ kN·m (+18.6%) under NFGMs. The curvatures of the joint seam and UE-GS show no evident change with an increase in prestress under non-NFGMs. However, the curvatures of the joint seam and UE-GS increase significantly by 11.9% and 137.3%, respectively, with an increase in the prestress under NFGMs. In other words, the degree of plastic damage development in the joint seam and UE-GS increases with an increase in the AACR under NFGMs. This is because when the spectral value of the ground motion is relatively large (NFGM), the plastic damage to the concrete caused by compression increases, and the bond-slip effect between the confined concrete and rebar becomes severe with a significant increase in the AACR.

4.2 Longitudinal reinforcement ratio

When the AACR was set as 10%, Fig.19 shows the variation curves of the main seismic response of the PSP-CB with different longitudinal reinforcement ratios (LRRs).

As shown in Fig.19, the effect of LRR on the main seismic response of the PSP-CB under non-NFGMs is very similar to that under NFGMs. The pier-top displacement decreases as the LRR increases. The mean value of the pier top displacement decreases by −13.4% and −20.4% under non-NFGM and NFGM conditions, respectively, with the LRR increasing from 1% to 2%. The mean value of the pier bottom bending moment increases by 42.1% and 49.8% under non-NFGM and NFGM conditions, respectively, with an increase in LRR from 1% to 2%. With an increase in the LRR, the curvature of the joint seam increases, whereas the curvature of the UE-GS decreases. Under the non-NFGM and NFGM conditions, the mean value of the curvature of the joint seam increases by 1146.8% and 2148.5% when the LRR increases from 1% to 2%, respectively. However, in this case, the mean values of the curvature of the UE-GS decrease by 63.8% and 69.1%, respectively. This is because, when the LRR increases, the initial stiffness of the pier increases. Hence, the pier top displacement decreases, and the pier bottom bending moment increases. However, an increase in the LRR leads to a weakened ductility. The plastic deformation is more concentrated at the joint seam. Consequently, the curvature of the joint seam increases, whereas that of the UE-GS decreases.

4.3 Slenderness ratio

When the AACR is 10% and the LRR is 1.5%, Fig.20 shows the displacement response of the pier top for the PSP-CB with different slenderness ratios (SRs). Fig.21 shows the variation curves of the main seismic response of the PSP-CB with different SRs.

As Fig.21 shows, the influence laws of SRs on the main seismic response of the PSP-CB under non-NFGMs are similar to those under NFGMs. The pier top displacement increases with an increase in the SR. The mean value of the pier top displacement increases by 330.1% and 319.1% under non-NFGM and NFGM conditions, respectively, when the SR increases from 4 to 10. The pier bottom bending moment, and the curvatures of the joint seam and UE-GS decrease significantly with an increase in SR. The mean value of the pier bottom bending moment decreases by 52.2% and 33.6%. The mean value of the curvature of the joint seam decreases by 98.2% and 84.6% In addition, the mean value of the curvature of the UE-GS decreases by 63.6% and 39.9% under non-NFGM and NFGM conditions, respectively, as the SR increases from 4 to 10. This is owing to the stiffness of the pier, which decreases when the SR increases. The PSP-CB period is also extended (Tab.10). The earthquake action suffered by the pier was reduced. Therefore, the pier-top displacement increases, whereas the other dynamic responses decrease. It should be noted that the NFGMs have a substantially greater influence on the pier-top displacement of the PSP-CB.

4.4 Span number

When the AACR is 10%, the LRR is 1.5%, and the SR is 7, Fig.22 shows the variation curves of the main seismic response of PSP-CB for different span numbers (SNs). It should be pointed out that when the SN is 3, pier #2 of the middle span is a fixed pier, whereas when the SN is 4 and 5, pier #3 is a fixed pier.

As Fig.22 shows, the influence laws of the SN on the main seismic response of the PSP-CB under non-NFGMs are the same as those under NFGMs. The pier top displacement, pier bottom bending moment, and curvatures of the joint seam and UE-GS increase with the SN. This is because the mass of the superstructure increases with an increase in SN, whereas the lateral resistance capacity and initial stiffness decrease. In this case, the PSP-CB period is extended (Tab.11). However, there is little difference among the acceleration responses of the superstructures of the PSP-CBs with different SNs (as shown in Fig.23). Consequently, the inertial force of the superstructure acting on the fixed pier increases with the SN, which leads to an increase in the seismic response of the PSP-CB.

In summary, the seismic performance of the PSP-CB is significantly affected by its design parameters, including the AACR provided by the prestress, LRR, SR, and SN. The spectral characteristics of the ground motions have a significant influence on the seismic response of the PSP-CB with different design parameters. The PSP-CB with the optimal design parameters exhibit good seismic performance under both non-NFGM and NFGM conditions. Therefore, the PSP-CB can be used in mid- and high-intensity seismic areas.

5 Influence of pile−soil interaction

5.1 Soil parameters

The length of the pile was 30 m and its diameter was 1.5 m. The concrete and steel materials of the pile were the same as those of the pier. The pile remained elastic during the finite element analysis. The plastic properties of the pile material are not considered when investigating the effect of PSI on the FEMs. The horizontal effective stiffness of the soil acting on each pile node was first calculated [45–47]. The soil parameters are listed in Tab.12.

5.2 Fiber element model verification

Based on Tab.12, the FEMs of RC-CB, SP-CB, and PSP-CB considering the PSI were built. In the FEMs, the PSI was simulated by an equivalent node spring with different horizontal effective stiffness [48,49], as shown in Tab.12. The dynamic characteristics of the three different bridges calculated by the Midas and OpenSees platforms considering PSI are listed in Tab.13.

Tab.7 and Tab.13 show that the first three vibration frequencies of the RC-CB, SP-CB, and PSP-CB all decrease after considering the PSI. The peak decreasing ratios of the first three vibration frequencies of the RC-CB, SP-CB, and PSP-CB are approximately 17.2%, 18.1%, and 20.3%, respectively. Therefore, the PSI has a greater influence on the vibration frequencies of the fabricated bridges. However, the dynamic characteristics of the three different bridges calculated by the OpenSees platform still consistent with the results calculated by the Midas platform and existing findings [43]. Moreover, the main modes of the Midas and OpenSees models considering PSI are the same. Hence, the OpenSees model that considers the PSI can be used for further seismic analysis. It should be noted that the local vibration modes of the FEMs become much more common when PSI is considered.

5.3 Pile-soil interaction effect

5.3.1 Seismic performance

Fig.24 and Fig.25 show the hysteresis curves of the fixed pier for RC-CB, SP-CB, and PSP-CB under typical non-NFGM and NFGM conditions after considering the PSI, respectively.

As Fig.24 and Fig.25 show, after considering PSI, the initial stiffnesses of the fixed SP-CB and PSP-CB piers are still higher than that of the RC-CB pier owing to the introduction of GS and PTs, which is similar to the numerical results without considering PSI; the pier-bottom shear forces of the PSP-CB clearly increase, and the residual displacements of the fixed piers of the RC-CB and SP-CB are significantly higher than those of the PSP-CB. This indicates that the fabricated pier with GS-PT composite connections still has a certain self-reset capability when PSI is considered. Moreover, when PSI is considered, the plastic deformation at the plastic hinge zone of the pier bottom and at the UE-GS both decrease. The curvatures of the joint seam and UE-GS for the fabricated bridges are significantly higher than that of the RC-CB, particularly for SP-CB. Therefore, considering PSI, the concentrated damage at the joint seam is much more serious because of the introduction of the GS. Fig.26 and Fig.27 compare the displacement response of the pier tops of the three bridges along the longitudinal direction under typical non-NFGM and NFGM conditions before and after considering PSI.

As Fig.26 and Fig.27 show, after considering PSI, the peak displacements of the fixed pier of the three bridges significantly increase under typical non-NFGM and NFGM conditions. The peak increasing ratios of the peak displacements of the RC-CB, SP-CB, and PSP-CB fixed piers are 26.0%, 33.9%, and 46.5%, respectively. Thus, the influence of PSI on the peak displacements of the fixed pier of the fabricated bridges is much higher than that of the RC-CB. Moreover, the residual displacements of the SP-CB and PSP-CB fixed piers are still lower than those of the RC-CB after considering PSI. Therefore, the PSP-CB still has a good self-reset capability when PSI is considered.

5.3.2 Influence of the design parameters

Fig.28–Fig.31 show the mean influence of the prestress, LRR, SR, and SN on the seismic performance of the RC-CB, SP-CB, and PSP-CB before and after considering PSI.

As Fig.28 shows, before and after considering PSI, the pier top displacement of the PSP-CB slightly decreases with increasing prestress, whereas the bending moment increases as the prestress increases, and the curvatures of the joint seam and UE-GS show no evident change with an increase in the prestress.

As Fig.29 shows, before and after considering the PSI, the influence laws of the LRR on the main seismic response of the PSP-CB under non-NFGMs are very similar to those under NFGMs. After considering the PSI, the pier top displacement decreases as the LRRs increase, whereas the pier bottom moment and curvature increase. It should be pointed out that after considering the PSI investigation, the influence of the LRR on the curvatures of the joint seam and UE-GS varied significantly. The damage states of the joint seam and UE-GS both need to be considered when investigating the PSI.

As Fig.30 shows, before and after considering PSI, the pier top displacement of the PSP-CB clearly increases with increasing SR, whereas the pier bottom bending moment and curvature both decrease. After considering PSI, the influence of the SR on the seismic performance of the substructure of the PSP-CB decreases. This is because the PSI causes an increase in the vibration periods of the PSP-CB, which to some extent is similar to the influence of the SR.

As Fig.31 shows, before and after considering PSI, the main seismic responses of PSP-CB increases with increasing SN. The reason for this is covered in the detailed analysis in Subsection 4.4. After considering the PSI, the influence of the SN on the seismic performance of the substructure of the PSP-CB also decreases. This is owing to the vibration periods of PSP-CB increasing slightly with increasing SR, which was also balanced when PSI was considered.

In summary, the PSI can extend the vibration periods of the PSP-CB, increase the displacement response, and decrease the pier bottom bending moment. This is because the PSP-CB is more prone to joint damage and more sensitive to long-period ground motion when the joint seam cracks. PSI may affect the failure probability and failure mode of the PSP-CB. Accordingly, the influence of the PSI on the seismic performance and design of the PSP-CB should not be ignored.

6 Conclusions

In this study, an FEM of a fabricated concrete pier with GS-PT composite connections was built and verified. The seismic performances of RC-CB, SP-CB, and PSP-CB models were compared through numerical analysis. Furthermore, the seismic performance of the PSP-CB was systematically investigated. The main conclusions are as follows.

(1) The plastic damage of PSP occurs mainly at the position of the joint seam and UE-GS. The stiffness, bearing capacity, hysteresis behavior, and failure mode of the PSP specimen simulated using the FEM was consistent with the experimental results. The proposed numerical analysis method is, therefore, suitable for analyzing the seismic performance of PSP, as well as the seismic responses of the RC-CB, SP-CB, and PSP-CB.

(2) Shifting of the plastic hinge zone and concentrated damage of SP-CB and PSP-CB occur under both non-NFGM and NFGM conditions. Under NFGMs, the plastic deformation at the joint seam and UE-GS and the concentrated damage at the pier bottom are very clear. NFGMs with a higher spectrum over a long period have a more substantial impact on the seismic response of the RC-CB than those of the SP-CB and PSP-CB. The PSP-CB exhibits better seismic performance under both non-NFGM and NFGM conditions.

(3) Within a certain range (0%–10%) of the AACR, the seismic performance of the PSP-CB improves as the prestress increases. The plastic deformation is more concentrated at the joint seam with an increase in the LRR. The pier bottom bending moment, curvatures of the joint seam and UE-GS of the PSP-CB decrease significantly with an increase in the SR. The main seismic response of the PSP-CB increases slightly with an increase in SN. A PSP-CB with optimal design parameters can be used in high-intensity seismic areas.

(4) The PSI can extend the vibration periods of the PSP-CB, increase the displacement response, and decrease the pier bottom bending moment. Overall, the PSI may affect the failure probability and failure mode of the PSP-CB. The influence of the PSI on the peak displacements of the fixed pier of the fabricated bridges is much higher than that of the RC-CB. Thus, the influence of the PSI on the seismic performance and design of the PSP-CB should not be ignored.

References

[1] Ameli M J, Parks J E, Brown D N, Pantelides C P. Seismic evaluation of grouted splice sleeve connections for reinforced precast concrete column-to-cap beam joints in accelerated bridge construction. PCI Journal, 2015, 60(2): 80–103

[2] Kurama Y C, Sritharan S, Fleischman R B, Restrepo J I, Henry R S, Cleland N M, Ghosh S K, Bonelli P. Seismic-resistant precast concrete structures: State of the art. Journal of Structural Engineering, 2018, 144(4): 03118001

[3] Al-Rousan R. Behavior of prefabricated full-depth precast concrete bridge deck panel system: Optimum prestress level. Procedia Manufacturing, 2020, 44: 607–614

[4] Qu H Y, Li T T, Wang Z Q, Wei H, Shen J, Wang H. Investigation and verification on seismic behavior of precast concrete frame piers used in real bridge structures: Experimental and numerical study. Engineering Structures, 2018, 154: 1–9

[5] Li T T, Qu H Y, Wang Z Q, Wei H, Jiang S. Seismic performance of precast concrete bridge columns with quasi-static cyclic shear test for high seismic zones. Engineering Structures, 2018, 166: 441–453

[6] WangZ QZhangY BJiangS CWeiHYanX. Experimental study of shear performance of precast segmental bridge piers with grouted splice sleeve. Journal of Tongji University (Natural Science), 2018, 46(6): 767–775

[7] Xin G T, Xu W B, Wang J, Yan X, Chen Y, Yan W, Li J. Seismic performance of fabricated concrete piers with grouted sleeve joints and bearing-capacity estimation method. Structures, 2021, 33: 169–186

[8] Haber Z B, Mackie K R, Al-Jelawy H M. Testing and analysis of precast columns with grouted sleeve connections and shifted plastic hinging. Journal of Bridge Engineering, 2017, 22(10): 04017078

[9] Al-Jelawy H M, Mackie K R, Haber Z B. Shifted plastic hinging for GS column connections. ACI Structural Journal, 2018, 115(4): 1101–1114

[10] Nishiyama M, Wei Y. Effect of post-tensioning steel Anchorage location on seismic performance of exterior beam-to-column joints for precast, prestressed concrete members. PCI Journal, 2007, 52(2): 18–30

[11] Li C, Hao H, Zhang X H, Bi K. Experimental study of precast segmental columns with unbonded tendons under cyclic loading. Advances in Structural Engineering, 2018, 21(3): 319–334

[12] Wang S J, Xu W B, Wang J, Xin G, Chen Y, Zhao Z, Bai Y. Experimental research on anti-seismic reinforcement of fabricated concrete pier connected by grouting sleeve based on CFRP and PET materials. Engineering Structures, 2021, 245: 112838

[13] Ou Y C, Chiewanichakorn M, Aref A J, Lee G C. Seismic performance of segmental precast unbonded posttensioned concrete bridge columns. Journal of Structural Engineering, 2007, 133(11): 1636–1647

[14] Palermo A, Pampanin S, Marriott D. Design, modeling, and experimental response of seismic resistant bridge piers with posttensioned dissipating connections. Journal of Structural Engineering, 2007, 133(11): 1648–1661

[15] HewesJ T. Seismic design and performance of precast concrete segmental bridge columns. Dissertation for the Doctoral Degree. San Diego: University of California, 2002

[16] Chou C C, Chang H J, Hewes J T. Two-plastic-hinge and two dimensional finite element models for post-tensioned precast concrete segmental bridge columns. Engineering Structures, 2013, 46(Jan): 205–217

[17] Han W L, Zhao Z Z, Qian J R, Cui Y, Liu S. Seismic behavior of precast columns with large-spacing and high-strength longitudinal rebars spliced by epoxy mortar-filled threaded couplers. Engineering Structures, 2018, 176: 349–360

[18] Yang C C, Okumus P, Ren R L. A hysteretic model for self-centering precast concrete piers with varying shear-slip between segments. Engineering Structures, 2019, 188: 350–361

[19] Liu H T, Chen J N, Xu C S, Du X. Seismic performance of precast column connected with grouted sleeve connectors. Journal of Building Engineering, 2020, 31: 101410

[20] AbdelkarimO IGheniA AAnumoluSElgawadyM A. Hollow-core FRP-concrete-steel tubular columns subjected to seismic loading. In: Symposium on Towards Sustainable Infrastructure with Fiber Reinforced Polymer Composites. Kansas: American Concrete Institute, 2014, 185–200

[21] Elgawady M A, Dawood H M. Analysis of segmental piers consisted of concrete filled FRP tubes. Engineering Structures, 2012, 38: 142–152

[22] Dawood H, Elgawady M, Hewes J. Behavior of segmental precast posttensioned bridge piers under lateral loads. Journal of Bridge Engineering, 2012, 17(5): 735–746

[23] Zhang Q, Alam M S. Evaluating the seismic behavior of segmental unbounded posttensioned concrete bridge piers using factorial analysis. Journal of Bridge Engineering, 2016, 21(4): 04015073

[24] Wang Z Q, Ge J P, Wei H Y. Seismic performance of precast hollow bridge piers with different construction details. Frontiers of Structural and Civil Engineering, 2014, 8(4): 399–413

[25] OuY C. Precast segmental post-tensioned concrete bridge columns for seismic regions. Dissertation for the Doctoral Degree. Buffalo: State University of New York at Buffalo, 2007

[26] Tazarv M, Saiidi M S. UHPC-filled duct connections for accelerated bridge construction of RC columns in high seismic zones. Engineering Structures, 2015, 99(sep.15): 413–422

[27] Zhuo W D, Tong T, Liu Z. Analytical pushover method and hysteretic modeling of precast segmental bridge piers with high-strength bars based on cyclic loading test. Journal of Structural Engineering, 2019, 145(7): 04019050

[28] GeJ P. Seismic performance analysis of a continuous girder bridge with precast segmental pier. In: 1st International Conference on Civil Engineering, Architecture and Building Materials (CEABM 2011). Clausthal-Zellerfeld: Trans Tech Publications, 2011, 1966–1970

[29] Zhao L F, Bi K M, Hao H, Li X. Numerical studies on the seismic responses of bridge structures with precast segmental columns. Engineering Structures, 2017, 151: 568–583

[30] Royero J B, Saiidi M S, Itani A. Seismic performance analysis and assessment of a precast bridge computational model. Dyna (Medellín, Colombia), 2020, 87(212): 80–89

[31] Liu H T, Wang Z Y, Xu C S, Du X. Influence of axial compression ratio on the seismic performance of precast columns with grouted sleeve connections. Journal of Structural Engineering, 2021, 147(12): 04021194

[32] Jia J, Zhang K, Saiidi M S, Guo Y, Wu S, Bi K, Du X. Seismic evaluation of precast bridge columns with built-in elastomeric pads. Soil Dynamics and Earthquake Engineering, 2020, 128: 105868

[33] Bu Z Y, Ou Y C, Song J W, Zhang N S, Lee G C. Cyclic loading test of unbonded and bonded posttensioned precast segmental bridge columns with circular section. Journal of Bridge Engineering, 2016, 21(2): 04015043

[34] WangX LYeL PLuX Z. Numerical simulation for the hysteresis behavior of prestressed concrete structures under cyclic loads. Earthquake Resistant Engineering and Retrofitting, 2006, 6: 28–32 (in Chinese)

[35] Song L L, Guo T, Chen C. Experimental and numerical study of a self-centering prestressed concrete moment resisting frame connection with bolted web friction devices. Earthquake Engineering & Structural Dynamics, 2014, 43(4): 529–545

[36] Hua L J, Rahman A B A, Ibrahim I S. Feasibility study of grouted splice connector under tensile load. Construction & Building Materials, 2014, 50: 530–539

[37] Fan J J, Feng D C, Wu G, Hou S, Lu Y. Experimental study of prefabricated RC column-foundation assemblies with two different connection methods and using large-diameter reinforcing bars. Engineering Structures, 2020, 205: 110071–110075

[38] Liu H T, Han Q, Bai Y L, Xu C, Du X. Connection performance of restrained deformed grouted sleeve splice. Advances in Structural Engineering, 2018, 21(3): 488–499

[39] Zhao J, Sritharan S. Modeling of strain penetration effects in fiber-based analysis of reinforced concrete structures. ACI Structural Journal, 2007, 104(2): 133–141

[40] FIBTask Group on Bond Models. Bond of Reinforcement in Concrete. Lausanne: International Federation for Structural Concrete (fib), 2000

[41] JG/T408-2013. Cementitious Grout for Coupler of Rebar Splicing. China: Ministry of Housing and Urban-Rural Development of the People’s Republic of China, 2013 (in Chinese)

[42] Ding Y N, Xu W B, Chen Y J, Wang J, Yan W. Experimental research on seismic performance of precast cogging high-strength bolt composite joint and influence of its arrangement location. Engineering Structures, 2020, 225: 111294

[43] Mohebbi A, Saiidi M S, Itani A M. Shake table studies and analysis of a precast two-column bent with advanced materials and pocket connections. Journal of Bridge Engineering, 2018, 23(7): 04018041–04018046

[44] JTG/T2231-01-2020. Specifications for Seismic Design of Highway Bridges. China: Ministry of Transport of the People’s Republic of China, 2019 (in Chinese)

[45] JTG3363-2019. Specifications for Design of Foundation of Highway Bridges and Culverts. China: Ministry of Transport of the People’s Republic of China, 2019 (in Chinese)

[46] Xie W, Sun L M. Assessment and mitigation on near-fault earthquake wave effects on seismic responses and pile-soil interactions of soil-pile-bridge model. Soil Dynamics and Earthquake Engineering, 2021, 143: 106596

[47] Li S, Zhang F, Wang J Q, Alam M S, Zhang J. Seismic responses of super-span cable-stayed bridges induced by ground motions in different sites relative to fault rupture considering soil−structure interaction. Soil Dynamics and Earthquake Engineering, 2017, 101: 295–310

[48] Zheng J Z, Takeda T. Effects of soil−structure interaction on seismic response of PC cable-stayed bridge. Soil Dynamics and Earthquake Engineering, 1995, 14(6): 427–437

[49] Khan R A, Ahmad S, Datta T K. Effect of soil−structure interaction on seismic risk of FAN type cable stayed bridges. Journal of Seismology and Earthquake Engineering, 2004, 6(2): 47–56

RIGHTS & PERMISSIONS

Higher Education Press