Seismic Isolation Technology for Prestressed Segmental Precast Piers

Due to its poor energy dissipation capacity, the application of prestressed segmental precast piers in bridge engineering in areas with medium- and high-intensity seismic activity is limited. This study proposes the use of a functional separation seismic isolation device (FSID) that acts as an external replaceable energy dissipation device to increase the energy dissipation capacity of prestressed segmental precast piers. Full-scale in-situ load tests on an actual prestressed segmental precast pier were carried out to clarify the seismic performance of this kind of pier. Based on the results of the tests and further analysis, an optimal design of the functional separation seismic isolation device is obtained by incorporating the optimal mechanical performance parameters of the device. The FSID proposed in this study can meet the seismic performance requirements of prestressed segmental precast piers under different levels of earthquakes.


Introduction
A pressed segmental precast pier is constructed by applying post-tensioned prestress to focus the nonlinear rotation of the precast pier under seismic action mainly on the rocking joint ; Wang et al. 2018) to ensure that the prefabricated components remain elastic and that the residual deformation is small after loading (Li et  To solve the problem of both the poor energy dissipation capacity and the large displacement demand, external replaceable energy dissipation devices are used in prestressed segmental precast piers (Marriott et al. 2009;Marriott et al. 2011). Chou (Chou and Chen 2006) proposed a method of energy dissipation in which concrete segments are restrained by circular tubes and connected by post-tensioned tendons. The low cycle reciprocating test shows that the proposed energy dissipation method can ensure that the specimen has good energy dissipation ability. Marriott (Marriott et  . However, the energy dissipation effect of energy dissipation devices installed at the base of columns is limited. There are two reasons for this. First, the use of energy dissipation devices does not reduce the total energy input. Second, the energy dissipation device needs a certain amount of deformation (displacement, angle) to occur at the bottom of the column to properly dissipate energy.
To address the above problems, this paper proposes the placing of a functional separation seismic isolation device (FSID) between the pier cap beam and the upper beam body. Through deformation, the FSID can reduce and consume the total seismic energy input into the bridge pier, and adopts the form of function separation, which makes it easy to design and control. To study the seismic response control effect of the FSID on a prestressed segmental precast pier, a full-scale load test of the prestressed segmental precast pier was performed. Based on the results of this test and the seismic performance requirements of prestressed segmental precast piers under different levels of earthquake magnitudes, an optimized FSID design is proposed.

Specimen design and construction
Full-scale loading tests were carried out on a prestressed segmental precast pier modeled after that used in the Hetian Ruoqiang railway. The Hetian Ruoqiang railway is located in Hetian city in the south of the Xinjiang Uygur Autonomous Region. Figure 1 shows the geometry and section details of the prestressed segmental precast pier. The mechanical properties of the materials used to form the prestressed segmental precast pier specimens are listed in Table 1. The prestressed segmental precast pier is divided into three parts: the cap beam, the pier body segments, and the bearing platform. Each segment of the pier body is made separately, and then they are assembled to create the bridge pier when the concrete strength reaches the design strength. 12 high strength and low relaxation self-anchored prestressed tendons with a diameter of 15.2 mm were used to form the prestressed system. The standard tensile strength of each prestressed tendon is f pk = 1,860 MPa. The tension control stress at the anchor is 1,116 MPa.
The construction procedure of the prestressed segmental precast pier is as follows: (1) The metal bellows were embedded in the bearing platform, the column, and the cap beam, forming the tendon duct during the prefabricating process.
(4) The tendons pass through the metal bellows from the top of the cap beam to the bottom of the column. After self-anchoring the prestressed tendons at the bottom of the column, the prestressed tendons were tensioned at the top of the bent cap. The tension stress of each prestressed tendon was controlled at 1,116 MPa. After the tension stress became stable, the top of the prestressed tendons was anchored ( Fig. 2 (c)).
(5) Back ll concrete mortar was poured into the reserved grooves of the bearing platform. Concrete mortar and grout were poured in from the top of the bent cap to the bottom, sealing the metal bellows, preventing the corrosion of the tendons, and ensuring the synchronous deformation of the tendons and each segment of the pier (Fig. 2 (d)).  Fig. 3, the steel beam of the reaction frame, placed above the cap beam, is connected to the cushion cap with 32 D32 mm thread steel bars using anchors embedded in the cushion cap. Thus, the loading steel beam, steel bars, and cushion cap form the vertical reaction frame. Four jacks were set between the loading steel beam and the pier bent cap. The axial load, applied at the top of the bent cap, was maintained throughout the experiment to simulate the permanent load acting on the top of the pier. To reduce the in uence of the vertical reaction frame on the horizontal displacement of the pier, two metal friction bearings with very small friction coe cients were set between the jacks and the loading steel beam, as shown in Fig. 3.
A horizontal load was applied at the bent cap through a horizontal reaction frame. As shown in Fig. 3, this frame is a 15 m high steel frame with very high stiffness. Su cient concrete blocks were placed at the bottom of the horizontal reaction frame to ensure that it did not overturn during the loading procedure.
The bent cap was connected to the horizontal reaction frame with a prestressed tendon, which passed through a 15 cm diameter hole at the center of the bent cap and was anchored at the side of the cap, while its other end passed through the jack, which was installed on the horizontal reaction frame. The horizontal displacement was imposed at the bent cap through the jack.

Load programs
The horizontal displacement was applied step by step through the jack installed on the horizontal reaction frame, as shown in Fig. 3. The imposed displacement at the cap beam was gradually increased from 5 mm to 130 mm at an increment of 5 mm. During the test, a laser displacement sensor measured the relative displacement between the cap beam and the top surface of the bearing platform.

Load-displacement curves
The relationship between the horizontal load and the horizontal displacement at the top of the pier is plotted in Fig. 4. Based on the relationship curve, the parameters of the main characteristic points (yield point, peak point, and limit point) were obtained as shown in Table 3. The yield point was determined by the park method (Shim et al. 2008). It can be noted that the maximum bearing capacity is 1,479 kN, and the equivalent yield stiffness is 22,020 kN/m.

Overview Of Fsid
The prestressed segmental precast pier has a characteristic of restorability. However, its hysteretic energy dissipation capacity is weak, and it may undergo brittle failure. The purpose of this study is to increase the energy consumption capacity of a bridge system with prestressed segmental precast piers and improve its seismic performance using a combination of different energy dissipation devices.

Structure of the FSID
A functional separation seismic isolation device is proposed to improve the energy dissipation and deformation capacity of bridges with prestressed segmental precast piers. The isolation device (Fig. 5) is composed of two vertical friction sliding bearings placed vertically and two high damping laminated rubber bearings placed horizontally. The friction sliding bearing (Fig. 5) placed vertically bears all the vertical load and maintains stability through shear keys and friction. Through the friction sliding of the sliding interface, the seismic energy is isolated and input to the upper beam body, and the motion period of the beam body is prolonged to achieve the isolation of the seismic energy. The high damping laminated rubber bearing is placed horizontally so that it does not bear the vertical load and consumes earthquake energy through shear deformation.

Working mechanism
Based on whether the dowel in the friction sliding bearing is sheared or not during earthquake activity, the isolation device can be divided into three working stages.
Step 1: normal use stage. Frequent earthquakes with smaller magnitudes are more likely to occur resulting in the bridge being in a normal use stage. The horizontal seismic action does not cause the shear key to reach its ultimate bearing capacity in the friction sliding bearing, and the shear key remains elastic at this stage; this is to improve the horizontal bearing capacity of the bearing system and ensure that longitudinal sliding of the main beam does not occur. In this stage, the FSID is not in active use.
Under the action of a one-way earthquake, the restoring force model of the FSID at this stage is shown in Fig. 7 (a), and P is the ultimate horizontal bearing capacity of the shear key.
Step 2: seismic isolation stage. The magnitude of forti cation intensity earthquakes is usually larger.
During their occurrence, it is necessary to ensure that the structure has clear energy-consuming characteristics. In this stage, the shear key will be destroyed to ensure the proper sliding of the friction bearing and the shear deformation of the high damping rubber bearing to fully dissipate the seismic energy and reduce excessive damage to the superstructure. Under the action of a unidirectional earthquake, the restoring force model of the FSID at this stage is shown in Fig. 7 (b), in which µ 1 is the friction coe cient; M is the vertical load; K 1 is the initial stiffness of the high damping laminated rubber bearing; K 2 is the post yield stiffness; f y is the yield load.
Stage 3: displacement limiting stage. Rarely, very high magnitude earthquakes occur. When one does occur, the bridge needs a limiting device to control the excessive horizontal displacement, and collisions between the upper beams must be avoided. At this stage, the friction coe cient will be adjusted to balance the friction force of the friction bearing and the restoring force of the high damping rubber bearing to optimize the displacement control function. Under the action of a unidirectional earthquake, the restoring force model of the FSID at this stage is shown in Fig. 7 (c), in which µ 2 is the friction coe cient.

Advantages of the FSID
The functional separation seismic isolation device separates the vertical bearing function from the horizontal displacement control function, and the normal use function from the seismic isolation function. Some of the advantages of the FSID include that the friction sliding bearing has a large vertical bearing capacity, the FSID shows good hysteretic energy dissipation, and the high damping laminated rubber bearing has a strong self-reset capability. Since the friction sliding bearing and the high damping laminated rubber bearing are uncoupled, the energy dissipation device has the characteristics of a clear force transmission path, a single function for each component, and it is easy to design and control. The FSID is a multi-stage working mechanism to achieve a performance-based seismic design of prestressed segmental precast piers (Zou et al. 2019).

Characteristics of the analysis model
The elastic design method is usually adopted in the seismic design of prestressed segmental precast piers to prevent plastic behavior (Li et al. 2008; . Moreover, all the deformation and energy dissipation is mainly concentrated in the isolation devices after they are added. Therefore, the pier of the isolation device in this study is simpli ed into a single particle model as shown in Fig. 8. M is the vertical load on the isolation bearing. Table 3 shows the input ground motions. The ground motion in the table is divided into ve levels for consideration. In the following discussion, g is the acceleration of gravity, g = 9.8m/s 2 . The Newmark β (β = 0.25) is used to conduct a time history analysis for all the input ground motions, and an estimation is made for the maximum response of each level of the input ground motion.

Design variables and design method
Based on the analysis model shown in Fig. 8, we performed an investigation to identify the optimal mechanical parameters of the functional separation isolation device. In this experiment, the FSID is required to play different control roles according to different input ground motion levels. By analyzing the seismic performance requirements of piers under different levels of earthquakes, we obtained the optimal mechanical parameters of the FSID. In this study, the sequential quadratic programming (SQP) method was used to optimize the damper parameters (Zou et al. 2021;Zou et al. 2020).

1st optimum design
In the 1st optimum design, the FSID was used as an energy dissipation device under Lv.1 ~ Lv.4 level ground motions. The optimization problem is shown in Table 5. (2) (2) D c ≤0.05m Optimization parameters The elastic stiffness K 1 , yield load F y , and post yield stiffness K 2 of the high damping laminated rubber bearing and friction coe cient µ 1 were considered as the parameters to be optimized.
Objective function Under the Lv.1-Lv.4 ground motions, to keep the pier elastic at all times, the optimal mechanical parameters of the FSID were obtained to minimize the maximum seismic response acceleration of the capping beam.
(3) Constraints (a) Considering the safety and stability of the bridge and the limit of the expansion joint, the maximum displacement of the pier cannot exceed the limit value. In this project, the limit value was set as 0.15 m. (b) Under a Lv. 1 horizontal earthquake, the FSID is in working stage 1, which can give full play to the seismic performance of the pier itself. When a Lv. 2 earthquake occurs, the FSID enters the second working stage and performs seismic isolation. In addition, under a horizontal earthquake action of Lv.2 ~ Lv.4, the pier columns always remain in an elastic working state, and the maximum horizontal force of the FSID should not exceed the horizontal yield shear force of the pier column. Therefore, the ultimate horizontal bearing capacity F f can be calculated by the following formula. (1) Where: F Lv.1 and F Lv.2 are the maximum horizontal force on the cap beam under an earthquake action of Lv. 1 and Lv. 2, respectively; is the total bearing capacity of the dowels; F f is the total horizontal force of the FSID; F y is the yield shear force of the pier column. According to the full-scale test results, the F y is 1,090 KN (Fig. 4).
(c) It is necessary to control the residual deformation of the FSID to improve the repairability of the bridge after an earthquake and avoid collision damage to the beam caused by aftershocks.

2nd optimum design
The FSID is considered as a limiting device to control the excessive displacement of the FSID under Lv.5 earthquakes and prevent collision damage to the cap beam. The optimization problem is shown in Table  6.
(1) Optimization parameters We performed analysis to obtain the optimal value of the friction coe cient µ 2 of the friction sliding bearing.
(2) Objective function Under Lv.5 ground motions, to ensure that the pier does not reach the ultimate bearing capacity, the optimal FSID mechanical parameters were obtained to minimize the maximum seismic response displacement of the cap beam. (

3) Constraints
The maximum horizontal shear force on the cap beam must not exceed the ultimate bearing capacity of the pier column under Lv. 5 earthquakes. The horizontal ultimate bearing capacity F can be calculated by the following formula, where: F u is the ultimate bearing capacity of the pier column.
2  Table 7 shows the optimal FSID design variables obtained in this study.  Figure 9 shows the maximum responses of the beam. FSID (only µ 1 ) are the maximum responses obtained when the high damping laminated rubber bearing and the friction sliding bearing with only one friction coe cient µ 1 are adopted. FSID (µ 1 + µ 2 ) are the maximum responses obtained when the high damping laminated rubber bearing and friction sliding bearing with two friction coe cients are adopted. The without FSID values are the maximum responses when the pier and beam are consolidated.

Analysis results
In the without FSID case, the maximum horizontal shear force on the cap beam reaches the ultimate bearing capacity of the column, and the pier is damaged. Therefore, the maximum seismic response under Lv. 3 and Lv. 4 earthquakes is not drawn in Fig. 9.
In the 1st optimization, under Lv. 1 ~ Lv.4 ground motions, in the case of with FSID (only µ 1 ) and with FSID (µ 1 + µ 2 ), the maximum seismic response acceleration and the maximum horizontal shear force of the cap beam are controlled within the elastic working requirements. The maximum horizontal displacement of the upper beam is ampli ed, but it is still controlled within the constraint range. This shows that the FSID between the pier and the upper beam acts as an effective seismic isolation device and ensures that the fabricated pier always remains in the elastic working state.
In the 2nd optimal design, under Lv. 5 ground motions, in the case of with FSID (µ 1 + µ 2 ), due to the use of the second friction coe cient µ 2 , the horizontal displacement of the beam is further reduced (as compared with the case of with FSID (only µ 1 )) to ensure that the maximum horizontal force on the pier does not exceed the ultimate bearing capacity. This shows that under rare high-magnitude earthquakes, the FSID plays a controlling role as a limiting device between the pier and the upper beam.  Fig. 10 is fuller, and it shows obvious three-stage working characteristics. Figure 11 shows the time history around the maximum responses for the Lv.5 input ground motions of the TAF-EW earthquakes. At 3.4 s, the pier specimen without the FSID became damaged. In the with FSID case, the maximum displacement of the FSID is less than 0.15 m before 3.8 s, and the relative slip of the friction damper is always at the area of the friction coe cient µ 1 . Therefore, the seismic response of the pier is the same in both cases (with FSID (only µ 1 ) and with FSID (µ 1 + µ 2 )). After 3.8 s, the maximum displacement of the FSID exceeds 0.15 m several times, and the relative slip of the friction damper reaches the area of the friction coe cient µ 2 . It can be seen that when the relative sliding reaches the area of the friction coe cient µ 2 , the friction force of the friction bearing decreases, and the restoring force of the high damping rubber bearing comes into play so that the excessive displacement on one side is restored. Therefore, the maximum displacement and residual displacement in the with FSID (µ 1 + µ 2 ) case is reduced as compared to the FSID (only µ 1 ) case.

Conclusions
In this study, full-scale in-situ load tests were conducted on a prestressed segmental precast pier to determine the seismic performance of the pier. Based on the test results, a functional separation seismic isolation device is proposed to improve the energy dissipation and deformation capacity of prestressed segmental precast piers. The optimal performance parameters of the functional separation seismic isolation device are identi ed through an optimization design method. The following conclusions are obtained.
(1) Through in-situ load tests, the load and the corresponding displacement of the key points in the loaddisplacement curve were obtained.
(2) From the analytical results, it is clear that the proposed design method can provide an optimal solution that satis es the multiple design criteria corresponding to the each level of the input ground motion.
(3) The FSID was effectively activated according to the input ground motions, and it shows obvious threestage working characteristics. The prestressed segmental precast pier with mixed connection using the FSID always remains in an elastic working state. The FSID serves as an effective seismic isolation device between the pier column and the upper beam during moderate and large earthquakes. The FSID also acts as an effective limiting device to control the excessive displacement between the bridge pier and the upper beam under rare high-magnitude earthquakes. The multi-stage working mechanism of the FSID can help to achieve the performance-based seismic design of prestressed segmental precast piers.
Declarations Figure 1 Overview of the prestressed segmental precast pier prototype (cm) Figure 2 Construction process of the prestressed segmental precast pier specimen Load-displacement curve Figure 5 Schematic diagram of the prestressed segmental precast pier bridge with the FSID Figure 6 Structural drawing of the friction sliding bearing  Maximum responses Figure 10 Hysteresis loops of the FSID under TAF-EW earthquakes Figure 11