Research On Methods For Seismic Analysis Of Underground Structures In Saturated Soft Soil

This dissertation conducts a systematical study on the seismic responses of underground structures in saturated soft soils.The primary objectives of this study are to(1)properly model the saturated soil in order to capture its dynamic characteristics,(2)study the features of the seismic soil pressure that acting on the underground structures and(3)investigate the seismic responses of the structures embedded in saturated soft soils.To achieve these goals,this dissertation proposes an indirect boundary element method and a finite element-indirect boundary element(FE-IBE)coupling method for seismic analysis of underground structures in saturated soft soils.The saturated soils are modeled by the Biot’s dynamic theory of poroelasticity thus rigorously accounting for the effects of the soil skeleton-pore water coupling.Using the indirect boundary element method and the FE-IBE coupling method,this dissertation firstly studies the seismic soil pressure of a circular tunnel,in terms of its magnitude and distribution patterns.The effects of tunnel mass,tunnel stiffness as well as tunnel buried depth,and the effects of the frequency contents and the incident angles of input seismic waves are considered.Based on the parametric study,a practical simplified method for the calculation of the seismic pressure is proposed.This simplified method is more accurate than traditional quasi-static simplified methods,since it can account for the amplification effect due to dynamic soil-tunnel interaction.Then,this dissertation investigates the seismic responses of underground tunnels and subway stations,with emphasis on the influence of soil nonlinearity on the structural seismic responses and emphasis on the wave passage effect on the longitudinal responses of the tunnel.Moreover,two widely used analytical solutions for seismic analysis of tunnels are assessed,with the results obtained by the FE-IBE coupling method being the benchmark.In addition,of particular interest are the effects of the soil skeleton-pore water coupling on the structural seismic responses.Hence,the differences of the results between two-phase models and single-phase models are investigated.Furthermore,the reliability of using simplified single-phase models(e.g.,water & soil combining method)instead of two-phase models in engineering practice is critically examined.The study shows that the tunnel stiffness is a key factor that influences the seismic soil pressure and pore pressure.The interaction of soil and stiff tunnels is dominated by the site dynamic characteristics.The seismic soil pressure of the stiff tunnel is significantly amplified compared to the free-field soil pressure,while their distributions are similar.For the examples in this dissertation,the seismic soil pressure of the stiff tunnel is 1.2～2 times larger than the free-field soil pressure.The interaction of soil and flexible tunnels is affected not only by the site dynamic characteristics but also the dynamic characteristics of the tunnel itself,and the tunnel stiffness plays dominant role in this matter.Moreover,the distribution of the seismic soil pressure of the flexible tunnel is evidently different from that of the free-field soil pressure.In addition,the nonlinear soil behavior has a great influence on the seismic responses of the underground structures.It is shown that the soil nonlinearity may lead to 27%～37% increases in the tunnel lining internal forces and 31% increases in the tunnel lining deformations,and it may also cause 12% increases in the deformation of the subway station.Moreover,the wave passage effect induces spatially varying earthquake ground motions,which cause significant longitudinal deformations of the tunnel and associated internal forces.In addition,the wave passage effect also has a great influence on the circumferential internal forces and ovaling deformation of the tunnel lining.The study also shows that the coupling the solid skeleton and the pore water alters the dynamic characteristics of saturate soil compared with a single-phase medium.For low permeability,the saturated soil behaves like a single-phase medium,making it possible to approximate the dynamic behavior of the saturated soil via an equivalent single-phase model.But this is not valid for high permeability or high frequencies.Using a single-phase model to approximate a two-phase model may cause noticeable errors in structural seismic responses.For the examples in this dissertation,the maximum difference of the tunnel lining longitudinal internal forces between a two-phase model and a single-phase model can be 35.5%,while the difference of the tunnel lining deformation is 12.5%.And the maximum difference of the internal forces of the subway station between the two models is up to 42.5%,while the difference of the lateral deformation is about 11%.