Effects of Burial Depth on the Seismic Response of Subway Station Structure Embedded in Saturated Soft Soil

For the aim to study the effects of burial depth on seismic response in the subway station’s structure embedded in saturated soft soil, on the basis of the OpenSEES program, numerical models with four cases of burial depth for the nonlinear interaction system including subway station’s structure and saturated soft soil are developed with the u-p formulations of Biot’s theory as it seeks to describe the saturated two-phase media. )e effects of burial depth on the seismic response and the failure pattern of the subway station structure are investigated. )e numerical results show that (1) the burial depth has a significant effect on the nonuniform uplift of station’s structure due to the movement of soil from far field towards baseplate; (2) with the increase of burial depth, the inertial effect of overlaying soil of the subway station structure becomesmore obvious, and the peak zone of the pore pressure ratio contracts from the outer area to the baseplate bottom with decreasing amplitude; (3) the dynamic index R decreases with the increasing burial depth of the station’s structure; and (4) in terms of the middle column, the burial depth of the station’s structure has the most crucial effect on the additional dynamic shear force and bending moment at the top end.


Introduction
e current antiseismic code design of the subway station structure system in the soft soil site is rougher than that of liquefiable sandy soil foundation.ere are only a number of qualitative and general provisions, and there is still a lack of more in-depth and systematic research [1][2][3][4].In this way, the consideration of the influence of the soft site on seismic response of the subway station structure dynamic system has an important theoretical value and practical significance to deepen the antiseismic understanding of the subway station structure and promotes the development of geotechnical earthquake engineering.As we all know, burial depth is one of the important factors that affect the seismic response of subway station structures [5,6].All previous earthquake disasters show that the subway station structure with a different thickness of a soft overlying soil layer has different site effects; that is, the dynamic damage degree and model of the subway station structure in the saturated soft soil site under different burial depth conditions has a significant difference [7,8].On the one hand, the shear effect and the soil dynamic pressure of the overlaying soil layer on the subway station structure are increased with the increase of the burial depth, which is unfavorable to the seismic response of the subway station structure.On the other hand, the constraint effect of the site on the subway station structure is strengthened with the increase of the burial depth, restricting the deformation reaction of the subway station structure, which is beneficial to the seismic response of the subway station structure.erefore, it is significantly necessary and urgent to study the seismic response characteristics of subway stations under different burial depth conditions in the soft site.
Many scholars have conducted many researches on the seismic response of the subway station structure under different burial depth conditions.Liu et al. [9,10] applied the nonlinear liquid-solid two-phase body dynamic finite element method to study the response of the burial depth of the subway station structure in the saturated liquefiable sand site under the horizontal and vertical seismic actions.Bilotta et al. [11] based on the total stress finite element method studied the dynamic response of a circular tunnel of 6 m diameter, with the axis at a depth of 15 m in a 30 m thick layer of medium dense gravel, sand, or soft clay, overlying a relatively stiff bedrock.Chen et al. [12] studied the response of the burial depth to the seismic displacement of the subway station structure based on the dynamic total stress finite element method and compared and analyzed the internal forces of the subway station structure under the seismic action.Zhuang et al. [13] made numerical simulation analysis of the seismic response of the subway station structure in 5 soft sites of the two-storey three-span island subway station with soft soil layers of different burial depth conditions in the side and bottom foundations and 1 general site based on the dynamic total stress finite element method.Amorosi et al. [14] adopted two different approaches, both accessible in the engineering practice to evaluate the increments of seismic-induced loads in the transverse direction of the tunnel lining in terms of hoop force and bending moment to study the seismic transversal response of a shallow tunnel built in two ideal clayey deposits of different consistencies.Chen et al. [15] based on the dynamic total stress finite element method carried out the numerical analysis of rock mass and tunnel lining and modeled the incident waves as harmonic S and P waves to investigate the influence of the depth of a tunnel on its seismic damage.Wang et al. [16] used the ultimate shear stress as the failure criterion based on the dynamic total stress finite element method, simulated the seismic damage process of the subway station structure by the birth-death element method, and considered the impact of the initial ground stress on the antiseismic property of the subway station structure by the burial depth of the structure.Li et al. [17] established two analysis models of subway station structure' burial depth based on the dynamic total stress finite element method to study the influence of the distribution of different soil layers of layered foundation on the seismic response of the structure.Tsinidis [18] conducted a numerical parametric study on diverse soil-rectangular tunnel systems, aiming at shedding light on critical response characteristics of rectangular tunnels subjected to transversal ground shaking based on the dynamic total stress finite element method.Cui et al. [19] adopted the orthogonal design scheme of the "four factors and three levels" numerical test and two-phase medium dynamic analysis method to analyze the dynamic response characteristics and the catastrophic mechanism of the subway station structure in the soft site and combined with the multiindex comprehensive balance method to carry out the sensitivity analysis of material parameters in view of the changes of the indexes such as the dynamic pore pressure ratio at the soil typical point and the dynamic response indexes including the internal force at the characteristic positions of the subway station structure caused by the change of soft foundation soil parameter factors.
Due to the complexity of the research problem of the antiseismic subway station structure in the saturated soft soil site and the limitation of calculation theory and hardware calculation, most of the existing research results in the subject of the seismic response of the subway station structure under the conditions of different burial depth conditions in the soft soil site are based on the total stress dynamic analysis method without considering the influence of pore pressure, in which the influence of dynamic pore pressure distribution change cannot be considered [20].However, the decoupled effective stress dynamic analysis method based on the pore pressure model has some limitations in the complex stress state deformation characteristics of saturated soil [21,22].It is rare in the existing research results to comprehensively consider the complex factors such as the multiphase medium property of saturated soft soil, the dynamic nonlinearity of the foundationstructure system, and so on to conduct the related research on the effect of burial depth on the underground structure.
On this base and based on the characteristics and limitations of the existing domestic and foreign research results, the dynamic Biot' two-phase medium coupling theory is adopted to describe the dynamic constitutive nonlinearity of saturated soft soil with the elastoplastic multiyield surface plastic kinematic model, and the nonlinearity dynamic performance of the station structure is described with the fiber section element and the modified Kent-Park concrete model.On this basis, the effective stress coupling nonlinear dynamic numerical solving model is established to analyze the effect of burial depth on the seismic response of the subway station structure dynamic system in the saturated soft soil layer, so as to provide guidance and reference for the antiseismic design and analysis of the subway station structure in the soft soil site.

Numerical Calculation Model
e numerical example is based on the typical single-storey double-span subway station in the saturated soft clay site, as shown in Figure 1, in which the cross section size of the subway station structure is 17 m (width), 7.17 m (height), and 3.5 m (center column spacing).Figure 2 shows the 2dimensional plane strain finite element model mesh of the established dynamic system of the subway station structure in saturated soft soil foundation, the size of which is 170 m × 30 m. e more unfavorable cases between soil and structure of complete bond and coordinated deformation are considered without considering the separation and sliding between them.Four different burial depth working conditions (2 m, 5 m, 8 m, and 11 m, resp.) are established to make a related analysis of the coupled effective stress numerical model for the saturated soft soil site-subway station structure dynamic interaction system.
In the analyses, the boundary between the soil deposit and the bedrock was assumed fixed in both horizontal and vertical directions and serving as the bottom boundary of the analyzed domain.e ground surface was assumed flat and free of loadings.e test simulated the free-field boundary conditions using laminar boxes.In the OpenSEES program, the behavior of the laminar box can be modeled by adopting the tied node feature. is feature allows the horizontal and vertical displacements at the two boundaries with the identical value.
e stress-strain behavior of the material model for saturated soil could simulate the hysteresis damping to a certain extent.e bottom and both sides of the site are set as undrained boundaries, the surface is set as 2 Advances in Civil Engineering a drainage boundary, and both sides of the model are set as the tied boundary [23].e more unfavorable cases are considered without considering the separation and sliding between soil and structure, that is, complete bond and coordinated deformation between them [24,25].Energy dissipation e ect is stimulated by the Rayleigh damping adopted in the numerical model, and the speci c damping ratio is selected as 5% [26].e HHT step-wise time integration method is used in the dynamic calculation, which can consider the energy dissipation and the second-order accuracy.
e numerical formula of the saturated two-phase medium matrix used in the present paper is as follows [27]: where M is the total mass matrix; u is the displacement vector; B is the strain-displacement matrix, B ≡ LN u , which is related to the strain and displacement increment dε Bdu; σ ′ is the e ective stress tensor; Q is the discrete gradient operator for soil-water coupling; p is the pore pressure vector; S is the compression coe cient matrix; H is the permeability coe cient matrix; and vectors f u and f p , respectively, indicate the given boundary conditions of the volume force in the soil-water mixture and the liquid phase.e u-p form plastic constitutive relation of the elasticplastic multiyield surface motion based on Biot's theory as shown in Figure 3 is adopted for clay.
e yield surface formula of the clay multiyield surface model is where τ is the partial stress tensor; m is the number of the mth yield surface, m ∈ (1, 2, . . ., n); n is the total quantity of the yield surface; the parameters α (m) and K, respectively, indicate the center and radius of the mth yield surface; and K (m) is equal to 3/2 √ times of the radius of the yield surface. e double dot product of tensors A and B is e plastic constitution of the clay multiyield surface adopts the law of partial kinematic hardening, and the movement direction tensor of the yield surface is de ned as where s T is the second-order deviatoric stress tensor, which represents the deviatoric stress tensor of the intersection point of the yield surfaces f m+1 and f m , and α m and (p ′ + p ′ 0 )α m+1 are the center of the yield surfaces f m and f m+1 , respectively.Figure 4 shows a representation of the nonlinear soil behavior in terms of normalized shear modulus and damping curves vs shear strain amplitude and a comparison   Advances in Civil Engineering with standard literature behaviors [28].It can be seen that the nonlinear soil behavior in terms of normalized shear modulus and damping curves vs shear strain amplitude is in line with standard literature behaviors.erefore, the numerical model in the present paper can be used to simulate the dynamic properties of clay.
e site soil is simulated by the quadUP element.e soft soil constitutive calculation parameters are obtained by referring to the recommended values of OpenSEES clay constitution, speci cally as listed in Table 1 [29].e designed strength of the concrete with established embedded depth is C40 and the density is 2500 kg/m 3 [30,31].e adoption of the ber section unit of the underground structure considers its nonlinear dynamic performance.e schematic diagram of the ber section is shown in Figure 5 [32].Speci cally, the Concrete02 constitutive model (modi ed Kent-Park concrete model) is used in the structural concrete, and the stress-strain relationship is shown in Figure 6(a).e steel bar uses Steel02 dynamic isotropic hardening bilinear material model, and the stress-strain relation is shown in Figure 6(b), in which the elastic modulus of the steel bar is 200 GPa and the yield strength is 400 MPa.
e north-south component of the 1995 Kobe earthquake is scaled 0.2 g and used as the horizontal excitation in the analyses.
e magnitude is 6.9.North-south component was recorded at the Kobe Japanese Meteorological Agency (JMA) station during the Hyogo-ken Nanbu (Kobe) earthquake (Rjb (km): 0.94, Rrup (km): 0.96, and VS30 (m/s): 312).e Kobe seismic wave with duration of 30 s is selected and input horizontally from the base rigid boundary, and the acceleration input of the seismic ground motion and corresponding Fourier spectrum are shown as Figure 7. Initial ground stress balance (the elastic + plastic stage) is rst carried out, and then the seismic action is applied.e numerical calculation is completed by a 64-bit 32-core CPU-64G memory DELLT7600 high-performance desktop workstation, and all the numerical results are extracted and visualized by the self-compiled MATLAB interface program.

Analysis of Seismic Response of Burial
Depth to the Subway Station Structure in Saturated Soft Foundation

In uence of Burial Depth on Displacement of the Subway
Station Structure.Figure 8 shows the residual displacement of the soft site after the earthquake under di erent burial depth conditions.It can be seen from Figure 8 that, with the increase of burial depth, the uneven uplift e ect of the subway station structure caused by the movement and extrusion of soil from the far eld to the bottom of the oor is weakened gradually.e displacement time history curve of the middle point of the oor of the subway station structure in the soft soil site under di erent burial depth conditions is shown in Figure 9.It can be seen from Figure 9 that the peak of the horizontal and vertical displacements time history curve of the subway station structure gradually decreases with the increase of the burial depth, and the uplift amount of the typical point of the structure of 11 m (burial depth) is reduced by 50% than that of 5 m (with burial depth).It can be seen that the burial depth of the subway station structure in the saturated soft soil site has a signi cant in uence on the uneven uplift e ect.

In uence of Burial Depth on the Peak of Soil Dynamic Pore
Pressure Ratio.According to the principle of e ective stress, the dynamic pore pressure ratio (representative symbol: DPPR) is computed as the ratio of excess pore pressure to initial e ective vertical stress.Figure 10 shows the peak envelope diagram of the dynamic pore pressure ratio of the soft soil site under di erent burial depth conditions.It can be seen from the diagram that the predominance area of the dynamic pore pressure ratio peak of the soft soil around the shallow-buried subway station shows a typical ring-shaped distribution characteristic, and the maximum dynamic pore pressure ratio (0.5) appears near the oor of the subway station and the shear wall.With the increase of the burial depth of the station structure, the vertical e ective stress level of the deep soil near the structure becomes higher, and the inertial dynamic extrusion e ect of the overlaying soil in the station structure is increased accordingly.erefore, the range of the dynamic pore pressure ratio peak area of soft   Advances in Civil Engineering soil around the subway station structure is gradually close to the structure from outside, and the peak also decreases.
Figure 11 shows the time-history curves of the dynamic pore pressure ratio at the locations of characteristics points of soft soil sites under di erent burial depth conditions.It can be seen from Figure 11 that, with the increase of the burial depth of the station structure, the dynamic pore pressure ratio peak of the soil typical points F 1 and F 2 around    the structure gradually decreases, which is in accordance with the variation of the peak range and peak level of the dynamic pore pressure ratio of the soft soil around the subway station structure in Figure 9.

Influence of Burial Depth on Internal Force Response of the Subway Station
Structure.e position of the structural column-slab joint as shown in Figure 12 is selected as the typical section of the structure for analysis.For example, Tables 2-4 show the peak of the internal force response at the typical section of the subway station structure under different burial depth conditions, and the figure in the brackets in the tables is the peak increase percentage of the dynamic internal force with the burial depth of 11 m compared with that of 2 m.It can be seen from the tables that, with the increase of burial depth, the peak position of the maximum internal force in the whole structure will also be adjusted under the seismic action.Specifically, the maximum axis force peak of the whole structure appears at the left end of the floor (position of P 1 ) when the burial depth is 2 m, while the maximum axial peak position of the whole structure appears at the lower end of the middle column (position of P 7 ) when the burial depth is 11 m; the maximum shear peak of the structure appears at the lower end of the left wall (position of P 5 ) when the burial depth is 2 m, and when the burial depth is 11 m, the maximum shear peak position of the whole structure appears at the lower end of the right wall (position of P 9 ); and the maximum bending moment peak of the whole structure under the condition that the burial depth is 2 m appears at the left end of the floor (position of P 1 ), while the maximum bending moment peak position of the overall structure appears at the middle end of the roof (position of P 12 ) when the burial depth is 11 m.In addition, with the increase of burial depth, the peak of internal forces at typical sections of structural members also increases in varying degrees.In particular, for the typical section of the same structure, the maximum bending moment peak of the middle end (position of P 2 and P 3 ) of the structure floor when the burial depth is 11 m has an increase rate of about 750% compared with that of 2 m (burial depth), which is far greater than the increase rate of the maximum internal force peak of the other typical sections in the whole structure.e above analysis is aimed at the maximum internal force peak of the key point of the structure and the increase with the change of the burial depth.In order to further understand the effect of different burial depth conditions on the dynamic internal force response of the structure under the seismic action, the nondimensional dynamic response evaluation index R is defined, which is the ratio of additional dynamic internal force to total internal force (seismic action + gravity action) caused by seismic action.
Figure 13 shows the R distribution of the dynamic evaluation index at the typical section of the structure when the burial depth is 2 m.According to the order of the value of R in each dynamic internal force evaluation index in the figure, the positions of the first 2 typical sections are selected as the most unfavorable antiseismic component position in the subsequent analysis.Specifically, the dynamic axial force and dynamic bending moment of P 14 , the dynamic shear and dynamic bending moment of P 10 , the dynamic axial force of P 11 , and the dynamic shear force of P 6 are selected as the most unfavorable antiseismic component positions and indexes, and the change of corresponding dynamic evaluation indexes with burial depth is analyzed.
Figure 14 shows that the dynamic evaluation index of the most unfavorable typical section above varies with the burial depth of the structure.It can be seen from Figure 14 that, with the increase of the structural burial depth, the value of the dynamic response evaluation index R for each corresponding position has a decreasing trend, which illustrates from the other perspective that the dynamic failure of the shallow-buried subway station in the saturated soft soil site may be more serious.In fact, the internal force increase of the underground structure caused by the gravity of the system is larger than that of the additional internal force, that is, the R ratio decreases.e dynamic evaluation index of the dynamic bending moment at the upper end of the right wall (position of P 10 ) and the right end of the roof (position of P 14 ) is the highest, and the changes with the burial depth of the structure are also the most significant.
In particular, the shear and bending moment of the middle column are close to zero only under the gravity, there is R ⟶ 1 at each burial depth condition for the middle column, and the nondimensional dynamic response evaluation index R is not applicable.In particular, the increase of the shear and bending moment of the middle column is close to zero only under the system gravity; that is, for the middle column, there is R ⟶ 1 at each burial depth condition, and the nondimensional dynamic response evaluation index R is not applicable.As we know, the roof collapse caused by dynamic failure of the middle column is one of the important forms of seismic damage of the subway station structure.On this basis, in addition to the analysis of the most unfavorable seismic position according to the R value increase, the middle column is further selected as the most unfavorable seismic component position, and the change of additional dynamic internal force which increases with the burial depth is adopted for evaluation and analysis.Figure 14 shows that the additional internal forces of the upper and the lower ends of the middle column vary with the burial depth.As shown in Figure 15, the additional dynamic shear and dynamic bending moment at the upper end of the middle column (position of P 8 ) are most significantly influenced by the burial depth (the variation coefficients are 0.468 and 0.464, resp.), second to which is the additional dynamic shear and the dynamic bending moment at the lower end of the middle column (position of P 7 ) (the variation coefficients are 0.445 and 0.448, resp.), and the additional dynamic axial force of the upper and lower ends of the middle column are most slightly influenced by the burial depth (the variation coefficient is 0.199).In addition, Tables 2-4 also show that the maximum internal force peak and the change and increase of the middle column are lower than those of the other key characteristic positions under all burial depth conditions.

Conclusions
Based on the u-p form of Biot's dynamic consolidation equation and the e ective stress dynamic solution method of saturated two-phase medium, a numerical model of e ective stress coupled dynamic analysis for the saturated soft soil sitesubway station structure interaction system is established in the present paper, and the seismic dynamic response law of the saturated soft soil site-subway station structure interaction system under di erent burial depth conditions is analyzed.
e calculation and analysis results show the following: (1) With the increase of burial depth of the station, the uneven uplift e ect of the subway station structure caused by the movement and extrusion of soil from the far eld to the bottom of the oor is weakened    Advances in Civil Engineering gradually, and the burial depth in the saturated soft soil site has a significant influence on the uneven uplift effect.(2) With the increase of the burial depth of the station structure, the soil's vertical effective stress level at the depth of the structure becomes higher, and the inertial dynamic extrusion effect of the overlying soil layer is enhanced simultaneously.e range of the dynamic pore pressure ratio peak of the soft soil around the station structure is gradually close to the structure from the outside, and the peak level also decreases.(3) With the increase of the burial depth of the station structure, the value of nondimensional dynamic response evaluation index R for all positions shows a decreasing trend, in which the dynamic moment evaluation index of the upper end of the right wall and the right end of the roof is the largest, and the change with the structural burial depth is the most significant.(4) For the internal force index of the middle column, the additional dynamic shear and dynamic bending moment at the upper end of the middle column are most significantly affected by the burial depth of the structure, followed by the lower end of the middle column.Compared with the corresponding value of the other key characteristic positions of the structure, the maximum internal force peak of the middle column and the increase with the burial depth are lower.

Figure 1 :Figure 3 :
Figure 1: e cross section of the subway station structure (unit: m).

Figure 4 :
Figure 4: Comparison of the (G)/(G) max vs c and λ vs c curves.

Figure 6 :
Figure 6: Constitutive model of reinforced concrete: (a) stress-strain curve of concrete; (b) stress-strain curve of the steel bar.

Figure 10 :Figure 11 :
Figure 10: Contours of peak response of the dynamic pore pressure ratio of soft soil ground with di erent burial depths: (a) 2 m; (b) 5 m; (c) 8 m; (d) 11 m.

Figure 12 :
Figure 12: Schematic diagram of key parts of dynamic force of the subway station structure.

Figure 14 :Figure 15 :
Figure 14: Variations of the dynamic index with burial depths at typical sections.

Figure 13 :
Figure 13: Comparison of the dynamic index at typical sections of the structure with 2 m burial depth.

Table 1 :
Model parameters of soft clay.

Table 2 :
e maximum axial forces at typical sections of the structure with di erent burial depths (unit: kN).