Progress in Seismic Isolation Technology Research in Soft Soil Sites: A Review

Abstract

Soft soil sites can amplify the peak acceleration by a factor of 1.5 to 3.5 and exhibit the filtering effect on seismic waves. This effect results in the attenuation of high frequencies, amplification of low frequencies, and extension of the predominant period of ground motion. Consequently, soft soil sites have a more pronounced impact on isolation buildings constructed on them. The seismic isolation structure design typically involves assuming rigid foundation for calculations. However, the soil properties can significantly impact the dynamic response of the structure, affecting factors such as input ground motion, changes in vibration characteristics, radiation energy dissipation, and material damping energy dissipation. Therefore, neglecting these influences and relying solely on the rigid foundation assumption for calculations can lead to significant errors in the final seismic response analysis of the structure. Currently, there are numerous LNG storage tanks, museums, and other isolation buildings constructed on soft soil sites. Therefore, research on seismic isolation measures for soft soil sites holds significant practical importance. In light of this, this paper, firstly, provides a systematic summary of seismic isolation strategies and engineering applications for soft soil sites. Secondly, it further discusses advancements in research on the dynamic interactions of soil–isolated structures, covering analytical methods, numerical investigations, and experimental studies on soft soil sites. Lastly, the paper concludes with insights on current research progress and prospects for further studies.

1. Introduction

Earthquakes frequently occur in various countries worldwide, exhibiting high intensity and a broad spectrum of activity. Strong seismic tremors can damage buildings and infrastructure, potentially leading to collapses, which result in significant human casualties and economic losses. According to relevant data from seismic damage investigations [1,2,3,4,5], traditional seismic structures exhibit significant issues related to structural safety, building damage, and loss of functionality, and as a result, fail to meet the increasing demands for enhanced building safety performance. Seismic isolation technology has been identified as an effective strategy for improving the seismic performance of structures. It maximizes the protecting of life and property during earthquakes and facilitates effective earthquake relief efforts. Consequently, this technology has been widely implemented in numerous practical engineering projects. The effectiveness of some seismic isolation engineering projects has been thoroughly validated during earthquake events. For instance, the seismic isolation test building at Tohoku University in Japan [6] exhibited no cracks during the earthquake on 11 March 2011. Similarly, the terminal building of Changshui International Airport in Kunming [7] demonstrated effective seismic isolation during the quake on 9 March 2015. Additionally, the Dertejörg State Hospital [8] showed strong seismic isolation performance during the earthquake on February six, 2023, with no damage to the isolation structure and only a few millimeters of residual displacement observed in the isolation layer. However, site soils consisting of soft soils characterized by high compressibility, low shear strength, low permeability, inhomogeneity, a natural moisture ratio greater than or equal to 35% or a liquid limit, a natural void ratio greater than or equal to 1.0, and a cross-sheet shear strength of less than 35 kPa [9,10,11,12] are widely distributed in regions including Tianjin, Shanghai, Ningbo, Zhejiang Province, Lianyungang, Jiangsu Province, and Qinhuangdao, Hebei Province, as well as other coastal areas. The physical and mechanical properties of these soils are detailed in

Table 1. Statistical table of physical and mechanical properties of soft soils in selected coastal regions of China.

Region	Soil Type	Natural Moisture Ratio $\mathit{\omega }$ (%)	Void Ratio e	$\mathbf{Liquidity}\mathbf{Index}{\mathit{I}}_{\mathit{L}}$	$\mathbf{Plasticity}\mathbf{Index}{\mathit{I}}_{\mathit{P}}$	$\mathbf{Coefficient}\mathbf{of}\mathbf{Compressibility}{\mathit{a}}_{1-2}$
Tianjin	Land-phase silt loam, soft peat layers	55.2~237	0.89~5.24	0.253~1.09	23.1~48.2	0.543~4.261
	Marine sedimentary silty soil, silt layer	46~59	1.26~1.62	1.2~1.33	19~28.4	0.94~1.16
Ningbo	Silty clay	34~58	>1	1.02~1.94	20~30	0.76 (the average value)
Wenzhou	Slit	>55	>1.5	>1.1	>17	>1.0
	Silty clay	37~56	1.0~1.5	1.0~1.7	17.7~25.3	0.57~1.5
	Silty chalky clay	35~50	1.0~1.4	1.09~1.85	>11.0~17.7	0.5~1.5
Lianyungang	Silt, silty soft soil subgrade	65.9~92.0	1.0~2.56	0.225~1.80	16~38	0.9~4.0

. Soft soils could filter out the high-frequency components of seismic waves, resulting in an increase in the period of ground motion input to the base of a structure, and causing them to experience seismic subsidence, making building sites on soft soils unfavorable for earthquake resistance during large earthquakes. The earthquake damage phenomenon mainly includes superstructure sinking and destruction due to foundation settlement and outdoor ground subsidence [13,14] (Figure 1a,b), cracks in the floor [1] (Figure 1c), and settlement and tilting of buildings caused by liquefaction of shallow foundations [15] (Figure 1d), as well as pile foundation occurred damage [16,17,18] (Figure 1e). However, seismic isolation structures built on soft ground exhibit a greater response in their fundamental mode compared to those on hard ground, leading to a rocking motion at the base of the pile foundation [19,20,21,22]. Hence, the codes [23,24,25,26] also have provisions stipulating that those buildings utilizing seismic isolation technology should be situated on sites classified as I, II, or III class. If an isolation building is constructed on the IV class site, appropriate measures must be implemented. According to the analysis provided above, it is of great practical significance to investigate the implementation of seismic isolation technology on soft soil sites.

Based on the preceding analysis, investigating the application of seismic isolation technology on soft soil sites holds significant practical importance. Accordingly, Section 2 of this paper will provide a summary of research concerning the application of commonly used seismic isolation bearings in China—specifically, rubber isolation bearings, elastic sliding isolation bearings, and friction pendulum isolation bearings (as illustrated in Figure 2) apply in soft soil structures, which have demonstrated effective seismic isolation performance. Additionally, this section will address the development of seismic isolation systems designed to mitigate the challenges associated with large horizontal displacements of the seismic isolation layer in soft soil environments. Section 3 will focus on the interaction between soil and isolation structures, covering various analysis methods and improvements, as well as numerical studies examining the SSI effects in soft soil sites and structures. Furthermore, this section will include experimental investigations of the SSI effects on seismic isolation structures situated in soft ground. Finally, Section 4 will offer a perspective on future research directions regarding the application of seismic isolation technology in soft ground conditions.

2. Seismic Isolation Scheme and Engineering Application for Soft Soil Sites

Based on extensive scientific research and engineering practice, the seismic isolation scheme for structures on soft soil sites typically falls into one of five main categories. The first category involves the utilization of lead-core rubber bearings (LRB) in combination with natural rubber bearings (NRB) [28,29]. LRB are placed around the isolation layer to enhance torsional stiffness, and NRB are placed at the center of the isolation layer, for they lack lateral stiffness. The seismic isolation program aims to regulate the horizontal displacement of the seismic isolation bearing by augmenting the lead core to enhance damping, and the horizontal stiffness of the seismic isolation layer increases. However, a large increase in the horizontal stiffness of the seismic isolation layer may prevent the achievement of the intended seismic isolation objectives, leading to a higher horizontal damping coefficient ($\beta$ > 0.4). Introducing rigid inclusions to strengthen the soft ground and make the peak acceleration response spectrum value of the foundation increase could potentially improve the seismic isolation effectiveness [30]. The second category involves the utilization of high-damping rubber isolation bearings (HDR) [31], which are arranged in a full-plane layout. However, due to their comparable function to lead-core rubber bearings, the horizontal displacement of the seismic isolation layer may exceed allowable limits. Consequently, it is necessary to supplement these systems with natural rubber bearings. The third category involves the utilization of friction isolation bearings (R-FBI) [32], which are designed for a full-plane layout, causing increased energy dissipation through hysteresis in the vibration isolation layer while minimizing energy consumption due to the deformation of the superstructure. However, in cases of rare earthquake, the friction pendulum bearing may experience a phenomenon known as lifting away. The fourth category entails the utilization of lead-core rubber bearings in conjunction with natural rubber bearings and viscous dampers (VFD) or U-shaped steel rod dampers (BZUD-I) [20,33,34]. This involves substituting a portion of lead-core rubber bearings with natural rubber bearings and incorporating dampers to dissipate energy, which could control the horizontal displacement of the seismic isolation layer and diminish the shear force exerted on the structural floor. The fifth category entails the utilization of lead-core rubber bearings or viscous dampers in combination with natural rubber bearings and elastic sliding plate bearings [35,36,37]. The elastic sliding plate bearings are strategically placed at the position of large axial force at the bottom of the column of the superstructure to withstand the gravity loads, dissipating seismic energy through frictional slippage. The rubber bearing does not need to bear the vertical pressure, supplemented with the additional damping ratio provided by the lead-core rubber bearing or viscous damper, which can effectively control the horizontal displacement of the isolation layer.

In summary, when utilizing commonly employed seismic isolation bearings at soft soil sites, it is necessary to consider a combination of seismic isolation methods, except for the friction pendulum isolation bearing. Attention must be given to controlling the horizontal displacement of the seismic isolation layer.

Table 2. Characteristics and effectiveness of five types of seismic isolation programs.

Isolation program	LRB and NRB [28,29]		HDR [31]	R-FBI [32]
Earthquake fortification intensity	7	8	7	8.5
Height of building (m)	(The paper [28] not specified)	13.8~22.2	23.8	14.1~19.5
Type of structure	Upper tower: RC frame-shear wall structure; lower podium: RC frame structure	RC frame structure	RC frame structure	RC frame structure
Principle of bearing arrangement	Structure perimeter: LRB Middle position of the structure: NRB		Full-plane layout	Full-plane layout
Selection of bearings	Large diameter rubber bearings ($d$ > 800 mm) and tensile devices should be installed at locations where the tensile stress exceeds 0.5 MPa	Large diameter rubber bearings ($d$ > 800 mm)	$d$ = 600 mm and 700 mm	The friction coefficient range from 0.01 to 0.05
Advantage	The vertical stiffness is comparatively low, while the vertical bearing capacity is high		Increased damping values contribute to a reduction in displacements of isolation layer under rare earthquakes	The vertical bearing capacity is greater than that of rubber bearings, and minimal vertical compression deformation, as well as significant horizontal displacement
Limitation	Large-diameter rubber bearings are required		If PGA exceeds 300 cm/s2, it is essential to utilize a large diameter bearing (d > 800 mm), and the structural model of the bearing follows the DHI model	In cases of rare earthquake, the friction pendulum bearing may experience a phenomenon known as lifting away
Natural period $T_1$	Equivalent period: 2.5 s; equivalent damping ratio: >15%	2.412	3.22	2.213
Horizontal damping coefficient $\beta$	0.3	0.27	(The paper not specified)	0.267
Horizontal displacement of the isolation	The average value < 0.5 $d$ = 400 mm	The maximum value < 0.55 $d$ = 440 mm	PGA = 220 cm/s2: The average value < 0.55 $d$ = 330 mm PGA = 300 cm/s2: The average value > 0.55 $d$ = 330 mm PGA > 300 cm/s2: The average value > 0.55 $d$ = 330 mm	The maximum value: <0.85*Horizontal ultimate displacement of the bearing = 425 mm
Isolation program	LRB, NRB, and VFD [20,33]		LRB, NRB, VFD and BZUD-I [34]	VFD, NRB, and ESB [35,36]		VFD, NRB, and ESB [37]
Earthquake fortification intensity	7	7	8.5	9	9		7
Height of building (m)	22.95	(The paper [33] not specified)	10.15~22.2	8.4	11		61.9
Type of structure	RC frame structure	Upper tower: RC frame-shear wall structure. Lower podium: RC frame	RC frame structure	RC frame structure	RC frame-shear wall structure		(The paper [37] not specified)
Principle of bearing arrangement	LRB is arranged around the structure, NRB is arranged in the middle of the structure, and VFD is arranged in the part with the large displacement of bearings after rare earthquake calculation		LRB is arranged around the structure, NRB is arranged in the middle of the structure, and VFD and BZUD-I is arranged in the part with the large displacement of bearings after rare earthquake calculation	Under frame columns: ESB Between columns: NRB. Under perimeter frame beams: VFD	The elastic sliding bearing is positioned at the location where the axial force at the base of the inner column is maximal
Selection of bearings	Rubber bearings: $d$ = 700 mm and 800 mm VFD: The damping index ($\partial$) ranges from 0.2 to 0.5	Rubber bearings: $d$ = 800 mm. VFD: The damping index ($\partial$) range from 0.2 to 0.5	Rubber bearings: $d$ = 700 mm, 800 mm and 900 mm. VFD: The damping index ($\partial$) was 0.3	Rubber bearings: $d$ = 600 mm VFD: The damping index, ($\partial$) range from 0.2 to 0.5 ESB: The friction coefficient, range from 0.01 to 0.03	ESB: The friction coefficient, range from 0.005 to 0.025		Rubber bearings: $d$ = 500 mm, 800 mm, 1000 mm and 1200 mm
Advantage	The incorporation of VFD contributes to a reduction in the shear force of the structure, the control of horizontal displacement, and the mitigation of tensile stresses in bearings		BZUD-I exhibits a significant deformation capacity, with deformations reaching approximately 500 mm; this characteristic enhances the yield force of the seismic isolation layer	The horizontal sliding stiffness of an elastic sliding bearing is zero when it has a sliding, rendering it insensitive to the spectral characteristics of ground vibrations and sliding displacements; in comparison, the horizontal stiffness is significantly smaller than rubber bearing
Limitation	BZUD-I and VFD are more costly			The ESB cannot be utilized independently; it must be used in conjunction with rubber bearings
Natural period $T_1$	(The paper [20] not specified)	Equivalent period: 2.5 s~3.0 s. Equivalent damping ratio: 15%	2.277~2.453	(The paper [35] not specified)	(The paper [36] not specified)		2.339
Horizontal damping coefficient $\beta$	0.31	0.35~0.42	0.36~0.37	0.27			0.51
Horizontal displacement of the isolation	The maximum value < 0.55 $d$ = 385 mm	The average value < 0.5 $d$ = 440 mm	The maximum value < 0.55 $d$ = 385 mm	The average value < 550 mm			The maximum value < 0.5 $d$ = 385 mm

provides a systematic overview of the advantages and disadvantages associated with the five types of seismic isolation systems discussed, as well as the seismic isolation effectiveness that can be achieved. The following key points can be highlighted: (1) For rubber seismic isolation systems, it is essential to select bearings with a larger diameter ($d$ > 800 mm) to control the horizontal displacement of the isolation layer effectively; (2) for friction pendulum isolation systems, the friction coefficient in the range of 0.01 to 0.05 is recommended to enhance self-recovery capabilities; and (3) elastic sliding plate bearings exhibit insensitivity to the spectral characteristics of ground vibrations, making them a preferable option for soft soil conditions. While LRB, HDR, and VFD can serve similar functions, VFD typically incurs higher costs.

To address the significant horizontal displacement of the seismic isolation layer in soft soil sites, Researchers have incorporated seismic isolation bearings, resulting in improved seismic isolation effectiveness. Zhang et al. [38,39] utilized the double concave friction pendulum isolation bearing (MFPS) along with a seismic isolation system comprising the annular damper reaction wall, viscous damper, and seismic isolation bearing installed at the top of the pile foundation (Figure 3a) in the seismic isolation design for large storage tanks situated in soft soil sites. They developed a multi-degree-of-freedom system for an oversized LNG tank with a capacity of 160,000 m³ as a case study. A simplified model was employed to compare and analyze the tank’s seismic response pre- and post-isolation. The findings indicate that both the MFPS isolation system and the investigated seismic isolation system can extend the tank’s natural period and decrease the horizontal shear force on the pile foundation, irrespective of the liquid level in the tank (at 1/3, 2/3, or design level). Liu et al. [40] introduced a combined sliding isolation system for LNG tanks (Figure 3b) to address the issue of significant displacement of the seismic isolation layer of LNG tanks on soft soil sites and developed a combined seismic isolation device, incorporating a damper reaction wall, natural rubber bearing (NRB), viscous-hysteresis damper (DPR), and inverted sliding seismic isolation bearing (SLB). Among these components, the SLB bearing positions the stainless-steel panel inverted on the lower surface of the tank, resulting in a surface area that can be exponentially larger than that of the PTFE plate. This configuration allows for greater slip of the PTFE plate, facilitating its adaptation to the larger horizontal displacements experienced by the seismic isolation layer. Additionally, viscous dampers are installed diagonally between the base plate and the reaction wall, enabling the dissipation of seismic energy through the substantial horizontal displacements of the seismic isolation layer. To evaluate the efficacy of this new combined sliding isolation system, the seismic responses of the pre-seismic isolation tank, the lead-core rubber isolation system tank, the double-concave friction pendulum isolation tank, and the combined sliding isolation tank were compared and analyzed based on key performance indicators including basal shear at the outer wall of the tank, impulsive basal shear, monopile shear, shaking wave height, and horizontal displacement of the isolation layer. The findings indicate that the combined sliding isolation system exhibits favorable adaptability on soft soil sites with significant displacement of the isolation layer and could ensuring the structural integrity of the upper tank under strong earthquakes. Therefore, the combined sliding isolation system represents a promising option for implementing seismic isolation technology in soft soil sites. Li et al. [41] developed a seismic isolation system utilizing elastic sliding bearings along with a small number of thick-layer rubber bearings and viscous dampers to enhance the seismic isolation of timber structures situated on soft soil sites. Specifically, the thick rubber bearings feature significantly greater thickness in a single rubber layer compared to natural rubber bearings, resulting in lower vertical and horizontal stiffness to accommodate increased vertical and horizontal deformations. The mechanical properties of this isolation system were evaluated, demonstrating its favorable seismic isolation performance, making it well-suited for long-period isolation structures in soft soil sites. These findings can serve as a valuable reference for similar seismic isolation reinforcement projects. Jangid [42] employed the numerical search technique to determine the optimal damping and tuning frequency ratio of a tuned inertial damper (TID) subjected to smooth white noise and filtered white noise seismic excitations. The study then utilized curve fitting to establish explicit formulas for the damping and tuning frequency of the TID under white noise excitation. It was observed that the error in these formulas could be negligible, indicating that the expressions are highly effective. The optimized design of the TID was found to provide an enhanced seismic isolation effect for base isolation structures on soft soil sites compared to those on hard soil. Wang et al. [43] introduced the key technologies for developing underground spaces directly beneath a heritage building (Figure 4), a combined project that was conducted as part of a renovation and repair project for a heritage building located on soft soil in Shanghai. These technologies involved integrating pile buttressing, foundation support, and building jacking, and adding the seismic isolation layer beneath the building. The specific process involved temporarily reinforcing the building, firstly, followed by the application of deformation-controlled foundation replacement technology to transfer the load from the superstructure to the replacement pile foundation. Additionally, a scheme combining micro-disturbance pit support with the foundation replacement structure was used for earth excavation and underground structure construction beneath the heritage building. Subsequently, jacking equipment was employed to reposition the heritage building onto the new foundation base plate. Finally, a seismic isolation bearing was installed, and the jacking equipment was removed to establish a new load-bearing structural system. Throughout the project, monitoring of building deformation was conducted, and the results indicated that both total deformation and differential deformation of the building remained within acceptable limits and met the preservation standards for the heritage building.

The above studies on seismic isolation strategies for soft soil sites and engineering applications all basically overlooked the influence of ground and foundations on the structure. Specifically, they relied on rigid foundation assumptions in conducting structural dynamic response analysis, disregarding potential alterations in the propagation of seismic waves within the soil medium. But, as previously mentioned, in the case of structures located on soft soil sites, the soil will impact the dynamic response outcomes of the structure in four ways, that is, via input ground vibration, alteration of system vibration characteristics, dissipation of radiant energy, and material damping energy. Neglecting the factors mentioned above and employing the rigid foundation assumption method for calculations will lead to seismic response results that do not accurately reflect the true seismic response of the structure, but instead exhibit significant errors. Therefore, to address the issues mentioned above and achieve more accurate computational outcomes in the assessment of structural seismic response, a seismic design approach considering soil–structure dynamic interaction (SSI effect) has been developed, known as the soil–structure dynamic interaction analysis method. The following sections will summarize the advancements in the study of soil–structure dynamic interaction, focusing on analysis methods, numerical investigations of SSI effects on seismic isolation structures in soft soil conditions, and experimental studies of SSI effects on seismic isolation structures in soft soil conditions. These findings helped to establish a fundamental groundwork for further research on the seismic performance of seismic isolation structures in soft soil environments and to enhance their seismic reliability effectively.

3. Investigation of the Dynamic Interaction between Isolation Structures and Soil

The dynamic interaction of isolated structures and soil involves not only the simulation of ground, seismic isolation structures, and base, but also requires solving dynamic equations, considering soil boundary effects, addressing nonlinearities in pile–soil dynamic interactions, material nonlinearities, and other challenges. Additionally, the characteristics of ground shaking during strong earthquakes must be considered. The integration of these various factors makes research in this area highly complex [44]. Generally speaking, research methods for the dynamic interaction of soil–isolated structures can be categorized into simplified and holistic analysis methods. On one hand, the simplified analysis method involves the creation of the simplified model where the soil and pile foundation are assumed to exhibit linear elasticity, neglecting their nonlinear properties. On the other hand, the holistic analysis method considers the nonlinear properties of the soil and pile foundation, as well as the heterogeneity of the ground, leading to the establishment of a more accurate computational analysis model [45]. Numerical and experimental studies are conducted based on above analytical methods to investigate the impact of soil–structure interaction on seismic isolation structures in soft soil sites.

3.1. Methods for Analyzing the Dynamic Interaction of Soil–Isolated Structures

3.1.1. Simplified Analysis Method

The simplified analytical approach comprises the set-total parameter method and the substructure method. The lumped-parameter method utilizes the impedance function to characterize the impact of pile–soil interaction, simplifies the elastic half-space foundation, and treats the pile foundation as a spring-damped-concentrated mass system. Common computational models employed in this approach include the SR model and the multi-mass system model, depicted schematically in Figure 5a,b. The substructure method can be broadly divided into two steps. The initial step involves determining the individual responses of the seismic isolation structure, pile foundation, and foundation, or the pile–soil coupled structure and the superstructure. The subsequent step involves combining these individual responses in a manner that satisfies the conditions of interaction, thereby obtaining the dynamic response of the overall system.

The SR model is a computational model in which horizontal and rotational springs related to horizontal displacement and rotation of the foundation are set up at the pile foundation of the structure. The structure is treated as a shear or bending-shear multiple-mass system, with a shear-type mass system utilized for a free site, and the acceleration response at the site surface is considered the input ground vibration at the pile foundation of the model. Since the bottom surface of the bearing platform is generally in contact with the foundation soil, an approximate method is usually employed to combine the impedance of the foundation soil with the horizontal and rotational stiffnesses of the pile foundation to determine the stiffness of the springs [46], as described in Equations (1) to (4).

$$K_x=\sqrt{{(K_x^F)}^2+{(K_x^p)}^2}$$

(1)

$$K_{\theta }=K_{\theta }^F+K_{\theta }^p$$

(2)

$$C_x=c_x{\displaystyle \frac{K_{s}R}{v_s}}$$

(3)

$$C_{\theta }=c_{\theta }{\displaystyle \frac{K_{s\theta }R}{v_s}}$$

(4)

where $K_x$ and $K_{\theta }$ denote the horizontal and rotational stiffness of the pile–soil coupled system, respectively. $K_x^F$ and $K_{\theta }^F$ represent the horizontal and rotational stiffness of the soil foundation without the pile. $K_x^P$ and $K_{\theta }^P$ represent the horizontal and rotational stiffness of the pile foundation. $K_s$ and $K_{s\theta }$ denote the static stiffness of the foundation. R is the radius of the foundation footing. $v_s$ represents the shear wave velocity of the soil. $C_x$ and $C_{\theta }$ represent the horizontal and rotational damping coefficients of the pile–soil system, respectively.

Figure 5. Computational model for the lumped-parameter method: (a) SR model [47]; (b) multi-mass system model (adapted from [48]).

Figure 5. Computational model for the lumped-parameter method: (a) SR model [47]; (b) multi-mass system model (adapted from [48]).

The multi-mass system model comprises the structural system, which includes the upper seismic isolation structure, pile foundation, and additional foundation soil, and the multi-mass free field system, which is independent of the structure. The free field system is regarded as a single soil column area, taking into account the varying soil properties of the site. It is segmented into multiple horizontal soil layers based on the actual soil conditions, with the mass of each layer concentrated at the interface of the soil layers. The superstructure is analyzed as a multiple-mass system subjected to shear or bending-shear forces, with the mass of the piles concentrated at the interfaces of different horizontal soil layers in the foundation soil, forming a bending-shear mass system. Springs and dampers are placed between the free-field system and the structural system to transfer the seismic response from the free-field to the structural system.

Researchers both domestically and internationally have primarily conducted theoretical investigations based on the SR model by developing a simplified model of seismic isolation structures that accounts for the SSI effect. Li et al. [49,50] utilized the simplified analysis method from the SR model to study seismically isolated structures, considering the SSI effects. They developed a simplified model for large aspect ratio seismically isolated structures considering SSI effects. Their findings indicate that (1) the SSI effect on high-rise seismic isolation structures is more pronounced than on multistory structures. Specifically, this effect results in an extension of the natural period of the seismic isolation structure and alters the damping ratio of the fundamental mode, either increasing or decreasing it; and (2) even in soft soil conditions with a shear wave velocity $v_{s(30)}$ of 91.5 m/s, the fundamental vibration period and damping ratio of the seismic isolation structure align well with the literature [51], demonstrating strong applicability. Ashiquzzaman and Hong [52] improved the SR model and introduced a simplified SSI model that incorporates the rotational inertia of the seismic isolation bearing. The findings demonstrate that the SSI impact influences the modal properties and damping ratio of the structure. Neglecting the moment of inertia of the seismic isolation bearing can lead to significant errors in floor displacement of the seismically isolated structure and eccentricity of the superstructure, posing a safety risk to the nuclear power plant building. Zhang et al. [53] developed a simplified model of a story isolation system that incorporates SSI effects, building on the systematic SR model proposed in previous research [54]. By analyzing the structural response transfer function and variance, the study investigated the impact of aspect ratio on the structural response while considering SSI effects. The findings indicate that when the structural aspect ratio $\eta$ is greater than or equal to 3.0, the peak displacement responses of both the upper and lower substructures increase as the structural aspect ratio increases. Conversely, when $\eta$ is less than or equal to 3.0, the displacement responses gradually increase as the structural aspect ratio decreases. According to the SR model, Liu et al. [55] initially developed a 4-degree-of-freedom (4-DOF) simplified model of the interlayer seismic structure incorporating SSI effects. However, due to the complexity of calculations involved in this model, they proposed a 2-DOF simplified model with mass, mode of vibration, and damping ratio equivalent to the original 4-DOF model. Subsequently, they presented a methodology for predicting the seismic response of the interlayer seismic structure while considering SSI effects, based on the equivalent simplified model, and validated the accuracy of the simplified analysis method through a case study, demonstrating that the equivalent model can effectively replace the original model, as well the proposed method is straightforward, efficient, and meets the required level of accuracy.

3.1.2. Holistic Analysis Method

The integral analysis method involves calculating the foundation soil, pile foundation, and upper seismic isolation structure as a unified system. It aims to accurately and effectively account for the interactions among these components and to provide a comprehensive and precise assessment of the seismic response of the entire system. Commonly used techniques for integral analysis include the finite element method and the boundary element method. The finite element method is an effective approach for addressing the consideration of SSI effects through discretization. Commonly used software includes ABAQUS, ANSYS, PLAXIS, and Midas GTS NX. However, the finite element method is limited in its ability to simulate the radiation-damping of infinite foundations, necessitating the introduction of artificial boundaries. These artificial boundaries typically include bounded boundaries [56], transmission boundaries [57], viscous boundaries [58], and viscoelastic boundaries [59,60,61,62]. The boundary element method involves constructing a regional integral equation on the boundary as the governing equation, converting it to a boundary integral equation using differential equations and Green’s formula, and then discretizing the boundary into a series of cells, which are then transformed into a set of algebraic equations for solution. Other methods include the infinite element method, the finite element-boundary element method, and the finite element-infinite element method.

3.2. Numerical Study of SSI Effect on Seismic Isolation Structure in Soft Soil Sites

The finite element method offers several advantages. Firstly, it enables the establishment of simplified or complex models for soil–isolated seismic structure interaction. Secondly, it allows for the precise control and adjustment of various parameters, such as geometric properties of the structure and pile foundation, material characteristics, soil stiffness, and damping, facilitating design scheme optimization. Thirdly, it provides high-precision numerical solutions. Lastly, it features advanced visualization and post-processing capabilities, enabling an intuitive display of dynamic responses, force distribution, and deformations of structures and foundations. Consequently, numerical simulations using the finite element method have become a crucial tool for investigating dynamic interaction issues in soil–isolated seismic structures. Based on this, this section summarizes numerical studies on the impact of SSI effects on seismic isolation structures in soft soil sites, employing simplified and comprehensive analysis methods to offer insights for designing seismic isolation structures in soft soil sites, respectively.

3.2.1. Simplified Model

In numerical investigations utilizing a simplified model for seismic isolation structures on soft soil sites with consideration of SSI effects, some researchers used continuously distributed nonlinear springs and dampers to equate the interaction between piles and soil as a beam on nonlinear Winkler foundation (BNWF) for the analysis of interactions between pile–soil systems and bridge seismic isolation structures. Soneji and Jangid [63] examine the impact of pile–soil–structure interaction (PSSI) on a diagonally tensioned seismically isolated bridge. The study aimed to achieve the following objectives: (1) assessing the influence of layered soil flexibility on the seismic response of cable-stayed isolated bridges; (2) comparing the PSSI effect with linear and nonlinear soil considerations; and (3) identifying scenarios where the PSSI effect is crucial in the seismic isolation design of cable-stayed bridges. The findings revealed that the PSSI effect significantly affects the lateral shear force at the tower base in soft soil conditions, with the effect diminishing as the soil layer stiffness increases. Analysis using a linear soil model inaccurately predicts the tower base shear force, while the nonlinear soil model provides a more accurate response to the dynamic behavior of the pile–soil system. Fosoul and Tait [64] initially performed seismic isolation reinforcement of a bridge on the soft soil site by implementing the fiber–rubber reinforced isolation (FREI) system. The seismic performance of the bridge was then evaluated both pre- and post-isolation, considering the PSSI effect. Vulnerability curves for the bridges were derived through vulnerability analysis. The findings indicate that the implementation of seismic isolation techniques enables the bridge deck to endure greater lateral displacements, leading to reduced acceleration response of the superstructure and minimizing the risk of extensive damage and collapse of the abutment piles. Almansa et al. [65] analyzed the seismic performance of a 22.95-meter-tall RC frame building with an isolated foundation in Shanghai, taking into account the SSI effect. The results indicated that mid- to high-rise isolated buildings demonstrate improved suitability for soft soil sites. Tena-Colunga et al. [66] conducted a study to assess the viability of implementing the friction pendulum foundation isolation system in six SF6 power stations located on soft soil sites in Mexico. The findings of the research demonstrated that the friction pendulum seismic isolation system is suitable for soft soil sites in Mexico when accounting for the SSI effect. Yanik and Ulus [67] examined the influence of the SSI effect on frame-base-isolated structures. They constructed and analyzed structures with 5, 10, and 40 stories to evaluate their seismic response under varying soil conditions. The findings indicate that the SSI effect is more pronounced for all three structures situated on soft soil sites. Abdeddaim et al. [68] incorporated magnetorheological (MR) dampers into the lead-core rubber isolation system to improve the seismic performance of the base isolation structure. Initially, they developed a simplified model of the base isolation system considering the SSI effect (Figure 6), and then optimized the parameters of the MR dampers using the particle swarm optimization (PSO) algorithm. Subsequently, they compared the seismic response of the foundation seismic isolation system, accounting for the SSI effect in firm soil, medium-soft soil, and soft soil before and after optimization, as well as the seismic response without the addition of MR dampers, were compared. The results indicate that the introduction of the MR damper and parameter optimization reduced the seismic response of the structure, with the greatest reduction observed in medium-soft soil and soft soil conditions. Baidya and Roy [69] examined the impact of the SSI effect on multistory base isolation structures utilizing shape memory alloy rubber isolation bearings (SMRAB). Their findings indicated that, when accounting for the SSI effect, the peak acceleration was reduced by 3.1%, 27.8%, and 35.8% for hard, medium-soft, and soft soil conditions, respectively. Additionally, the isolation displacements decreased by 15.2%, 24.9%, and 32% under the same soil conditions. Regarding the investigation of the SSI effects in story isolation structures, Gao et al. [70] developed a double-story isolation model (Figure 7a), a story isolation model (Figure 7b), and a fixed foundation model (Figure 7c). The study considered the SSI effects for both soft and hard soil sites (Figure 7d). Elastic-plastic time-history analyses were conducted under rare-earthquake conditions. The results indicated that the double-story seismic isolation structure exhibited superior seismic performance on hard soil sites compared to soft soil sites, while the layer-spaced seismic isolation structure performed better on soft soil sites than the double-story seismic isolation structure.

3.2.2. Comprehensive Model

In numerical investigations utilizing comprehensive model for seismic isolation structures on soft soil sites with consideration of SSI effects, the primary focus is on examining the influence of SSI effect on the performance of base seismic isolation systems under various site conditions. Radkia et al. [71] analyzed the seismic response of irregular steel framed seismically isolated structures under different site conditions considering the SSI effect under three-dimensional earthquake. A total of 24 models with different seismic isolation systems (pure sliding friction isolation system, friction pendulum isolation system, and lead-core rubber isolation system) were developed for fixed-foundation (40% irregularity), five-floor (regularity, 20%, 40%, and 60% irregularity), and 10-floor (40% irregularity) structures. The results showed that the soft soil site has a significant effect on the seismic response of the structure, but the regularity of the structural system does not affect the seismic response of the structure. The implementation of LRB effectively mitigates floor displacements and accelerations in multi-story buildings, whereas the use of R-FBI is effective in reducing floor displacements and accelerations in high-rise buildings. Zhang et al. [72] compared the overturning resistance of friction pendulum isolation structures and fixed foundation structures with and without SSI effects in soft, medium-soft, and hard soil sites. The findings indicate that the friction pendulum isolation system with the SSI effect exhibits higher resistance to overturning in soft soil compared to the system without the SSI effect. However, the overturning resistance is comparable to that of fixed foundation structures in medium-soft and hard soil sites. Liu et al. [73] developed a dynamic analysis model for the base-isolated nuclear power plant considering the SSI effect based on the theory of exogenous fluctuation. They utilized the finite element software ANSYS to establish a refined analysis model of the ground–foundation–nuclear island seismic isolation structure (Figure 8). A self-programmed time-domain computation program was employed for solving and post-processing tasks. The study compared the SSI effect on base-isolated buildings of the nuclear power plant at different sites. The findings indicated that as the soil shear wave velocity decreases, there is a stronger amplification of the low-frequency component of the seismic wave, resulting in a greater SSI effect on the input seismic wave. Therefore, it is crucial for the seismic isolation design of the nuclear island building to fully consider the influence of the SSI effect.

In summary, there is a substantial body of research on base isolation, encompassing both simplified and comprehensive models. The findings of the study (

Table 3. The main conclusions of the numerical study on the SSI effect on seismic isolation structures.

Research methods	Numerical studies—practical engineering [65,66]		Numerical studies—base isolation structure [69,71]		Numerical studies—base isolation structure [70]
Research purpose	Investigation of the applicability of base isolation	Seismic isolation of buildings for power stations considering soil–structure interaction effects	Feasibility study of implementing SMRAB in structures	Seismic response analysis of irregular steel-framed structures considering soil–structure interaction effects under three-dimensional Earthquakes	Investigation of the seismic response of double-story isolated structures considering soil–structure interaction effects
Site type	Soft soil	Different sites	Different sites	Different sites	Different sites
Isolation system	LRB, NRB, and VFD	R-FBI	SMRAB	P-F, R-FBI and LRB	LRB
layers	6	(The paper [66] not specified)	5	1, 5 and 10	12
Type of superstructure	RC frame structure	Steel frame structure	RC frame structure	Steel frame structure	RC frame structure
Main conclusions	Base isolation represents an effective solution for mid-to-high-rise reinforced concrete buildings located in regions with soft soil characteristics, such as Shanghai.	FPS is effective in soft soil sites.	After considering the SSI effect, the peak acceleration and the isolation displacements have a maximum reduction amplitude under soft soil sites.	Structural irregularities did not affect the dynamic response of the structure.	The double-layer seismic isolation structure exhibits superior seismic performance on hard soil sites. However, its performance is inferior to that of the laminated isolation structure on soft soil sites.
Effectiveness of the study	Following the Chinese code [25] and European regulations [74,75,76].	Refer to the Mexican Technical Code [77,78,79,80].	Initially, the peak acceleration and displacement responses of the seismic isolation bearings at the top of a five-story seismically isolated building with a rigid foundation are compared to those reported in the literature [81]. This comparison aims to validate the proposed numerical model of the seismically isolated building, which does not account for SSI effects. Following this, to assess the influence of SSI, soil parameters analogous to those described in the literature [82] were employed.	Following the American code [83].	The paper could inform the future development of high-performance building structures.
Limitations of the study	As an example of an RC frame structure only, the structure is not representative.	The swaying effects that arise when considering the SSI effect for structures on soft soil sites are not addressed.	Considering solely unidirectional seismic effects.	The analysis focused solely on the impact of structural regularity, without taking into account the influence of aspect ratio on the structure.	The double-layer seismic isolation structures exhibit an increased susceptibility to resonance when constructed on soft ground. This aspect warrants greater consideration during the design phase.

) indicate the following conclusions. (1) Structural regularity does not significantly influence the performance of base-isolated structures when considering SSI effects. (2) For the base-isolated structures situated on soft soil, the differences in seismic response before and after accounting for SSI effects are considerable, highlighting the necessity of incorporating SSI effects. Conversely, for structures on hard soil, the impact of considering SSI effects on seismic response is minimal, suggesting that SSI effects may not need to be included. (3) As to multi-story buildings (ranging from three to seven stories), engineering examples [65,66] demonstrate that isolation systems such as LRB+NRB+VFD or FPS can be effectively utilized for seismic isolation. For high-rise buildings (exceeding seven stories), the R-FBI seismic isolation system is recommended. However, it is important to note that this system should not be subjected to tensile forces.

3.3. Experimental Study of SSI Effect on Seismic Isolation Structure in Soft Soil Sites

Due to the sudden nature of earthquakes, obtaining measured data is challenging. As a result, simulating the impact of earthquakes on structures through model tests is a common practice. These tests help refine and enhance theoretical and numerical methods. One widely used approach is the shaking table test, which involves scaling the prototype structure based on a specific similarity ratio. This method can replicate the dynamic response of larger-scale models subjected to multidimensional and complex ground shaking. However, challenges arise in designing similarity ratios for acceleration and predicting the influence of infinite soil boundaries on test outcomes, potentially leading to inaccuracies in the results [47].

Analogous to numerical studies, some researchers have examined the SSI effect on seismic isolation systems in soft soil sites. Yamashita et al. [84] analyzed the seismic response of a 12-story reinforced concrete frame base isolation structure situated on soft soil in Tokyo during the 2011 Great Japan Earthquake by monitoring the soil–foundation-structure system. The findings indicated minimal change in foundation settlement and pile-raft load sharing before and after the earthquake. Additionally, the horizontal acceleration of the superstructure was reduced to 30% of the horizontal acceleration of the ground near the surface. Zhao et al. [85] conducted a study on the impact of SSI effect on the lead-core base seismic isolation system of the AP1000 nuclear power plant situated on soft soil through a series of shaking table tests. The findings revealed that the edges of the foundation experienced the highest soil pressure. At the seismic margin earthquake (SME) input level, the interface between the foundation of the seismically isolated structure and the soil started to separate, but the presence of the seismic isolation layer led to a decrease in the structural acceleration response. Conversely, for non-seismically isolated structures, the SSI effect led to an increase in seismic response of the subgrade in the short term, accompanied by rocking behavior at the operational base earthquake (OBE) input level. Most researchers have examined the impact of the SSI effect on seismically isolated structures under various site conditions using shaking table tests. Xu et al. [86,87] conducted a study to examine the impact of SSI effect on the dynamic properties of base isolation system in soft soil conditions utilizing shaking table tests (Figure 9). Initially, the feasibility of the testing approach was assessed, followed by a comparison of the dynamic characteristics of structures in both hard and soft soil conditions. Subsequently, the seismic responses of the pile foundation, the lead-core rubber seismic isolation bearings, and the seismic isolation structure were analyzed. The findings indicated that the rotational effects on the pile cap and the seismic isolation layer were more pronounced in soft soil conditions and were further amplified by the presence of the seismic isolation layer. This amplification effect would be more significant with an increase in the peak ground acceleration of the input ground shaking. Chen et al. [88] introduced a performance-evaluation framework for sliding isolation structures that takes into account the interaction between underground structures, soil, and surface structures (USSI) (Figure 10). Initially, a reference database for the performance-evaluation framework is established using shaking table tests and numerical simulations. Subsequently, the framework examines the influence of soil conditions, sliding isolation bearing parameters, and seismic excitation on slip-isolated structures considering the USSI effect. The findings indicate that this seismic isolation system not only reduces floor acceleration, floor displacements, and interstory displacements of the structure but also effectively mitigates the adverse effects of nearby subway stations. Furthermore, the parametric analysis results reveal that the seismic performance of structures within the USSI system is particularly sensitive to soil conditions, friction coefficient, spring stiffness, and input frequency, all of which can be incorporated into the proposed performance-evaluation framework. Yu et al. [89] investigated the impact of considering SSI effect on seismic isolation structures with small-aspect-ratio foundations on soft and hard soil sites through shaking table tests. Initially, they proposed an energy balance equation for the seismic isolation system, taking into account the SSI effect. Subsequently, they analyzed the influence of SSI on the energy dissipation of small-aspect-ratio seismic isolation structures on soft soil foundations. The findings indicate that the SSI effect notably amplifies the seismic response of seismic isolation structures on soft soil sites, and the ratios of kinetic energy, damping dissipation, and hysteretic deformation energy dissipation of the isolation structures on soft soil foundations differ significantly from those on rigid foundations. The impact of the SSI effect is more pronounced during large earthquakes, leading to an increase in the ratio of kinetic energy to damping dissipation of the isolation structures and a decrease in the ratio of hysteresis deformation energy dissipation of the isolation layers, with the decrease being linked to the seismic motion input characteristics. Xu et al. [90] conducted a study utilizing a combination of shaking table experiments and numerical simulations to investigate the dynamic characteristics and seismic response behaviors of layer-spaced seismic structures subjected to long-period ground shaking, taking into account the SSI effect. The study examined four distinct types of ground shaking: near-field ordinary, far-field ordinary, near-field impulse and far-field harmonic (long-period) ground shaking. Two structural models were developed for the analysis: one representing a rigid foundation and the other representing a soft soil ground spaced vibration structural model. The findings indicate that the acceleration responses of both structural models are diminished when the SSI effect is incorporated. However, the displacement ratio in the soft soil model exceeds that of the rigid foundation, and the horizontal deformation of the seismic isolation layer is greater under long-period ground shaking compared to ordinary ground shaking.

In general, the majority of experimental studies concentrate on foundation isolation, utilizing simpler experimental models that typically feature more regular frame structures. Additionally, the properties of the foundation soil are often more homogeneous, and there is a lack of systematic examination regarding the changes in flexible foundation stiffness. For more complex structures, such as underground constructions, further numerical studies are required. Consequently, the test results can only provide a qualitative analysis of the influence of the SSI effects on seismic isolation systems.

4. Visions for Future Research Endeavors

This paper provides a comprehensive overview of the measures associated with implementing seismic isolation technology for soft soil sites. It discusses both the seismic isolation strategies employed when the SSI effects are not considered, as well as the methodologies applied when the SSI effects are considered, alongside relevant research findings. The aim is to furnish a scientific foundation and technical support for engineering design, thereby ensuring the stability and safety of structures situated on soft soil sites. When SSI effects are disregarded, the most effective seismic isolation solution for soft soil sites involves using LRB, VFD, and HDR in conjunction with NRB and ESB. However, upon considering SSI effects, existing studies reveal a significant gap in the selection of seismic isolation bearings. These bearings are often designed solely based on code requirements, without assessing whether the seismic isolation structures meet the relevant criteria before considering SSI effects. Key considerations include whether the horizontal damping coefficient of the superstructure is less than 0.4 to achieve the desired “degree reduction”, and whether the eccentricity of the bearings complies with specified requirements. However, it is important to note that the insights presented in this paper are a synthesis of the reviewed studies; therefore, there may be inherent biases in the interpretation of studies not included in this analysis.

(1)

Based on the analysis presented in Section 2, it is evident that the scheme integrating LRB/VFD with NRB and ESB is a more advantageous option for the implementation of seismic isolation technology on soft soil sites. This approach effectively reduces the horizontal displacement of the seismic isolation layer and minimizes the deformation of the superstructure. Furthermore, it enhances the tensile capacity of the seismic isolation structure and improves the overall efficiency of seismic isolation. The seismic isolation system designed specifically for the characteristics of soft soil sites also demonstrates commendable seismic performance. One study [41] showed that incorporating thick rubber bearings into the rubber bearing and elastic skateboard bearing system and arranging them in the periphery of the isolation layer can mitigate torsional effects and residual deformation. Thick rubber bearings, a type of three-dimensional seismic isolation bearing, excel at isolating vertical ground vibrations while dissipating horizontal seismic energy, making them suitable for safeguarding critical equipment and major construction projects. Other effective options include three-dimensional vibration isolation double-control bearings [91], which consist of horizontal and vertical vibration isolation bearings connected in series, and three-dimensional overturning-resistant seismic isolation bearings [92]. Therefore, similar 3D seismic isolation bearings can also be considered to be introduced into the generalized scheme to study the anti-seismic performance of the structure.

(2)

In the current numerical study on the SSI effect of seismic isolation structures in soft soil sites, the majority of the research is focused on base isolation. However, for structures like offshore buildings, historical buildings, and large-span connecting corridor-type structures where base isolation is not feasible due to various constraints, the form of story isolation may be employed [93]. Nevertheless, there is a lack of research on the consideration of SSI effects in interstory isolation structures in soft soil sites. Therefore, further investigation into the impact of SSI effects on interstory isolation structures is warranted.

(3)

Under the long-term environmental effects, the changing rule of the performance of seismic isolation structures in soft soil sites is also worth studying. For example, excessive uniform settlement can lower the building’s elevation, affecting its functionality, while uneven settlement can lead to additional stresses and potential structural damage, compromising the building’s safety. Therefore, it is valuable to investigate the impact of foundation settlement on seismic isolation buildings in different sites, both considering the SSI effect and without considering it. Additionally, research [93] has shown that bearing settlement resulting from the creep of rubber vibration isolation bearings or construction irregularities during normal use can affect the performance of vibration isolation structures under wind loading. However, there is currently a lack of studies on the effect of bearing settlement on seismic isolation structures in soft soil sites. Therefore, it is necessary to explore the influence of bearing settlement on seismic isolation structures in soft soil sites, considering and excluding the SSI effect.