No abstract

No abstract

The traditional source-site-structure model for the calculation of seismic response of underground structures at near-source sites is restricted by the grid scale and the size of the structure. As a result, an excessive number of elements in the model make the numerical solving process difficult. To solve problems such as an inefficient computation and challenging nonlinear simulation, a multiscale analysis method for the calculation of the seismic response of underground structures at near-source sites is developed. The generalized free-field seismic response of the near-source region is obtained by establishing a large-scale calculation model of the source site and is used to simulate the fracture mechanism of faults and the process of seismic wave propagation. Then, using the method of seismic wave input based on artificial boundary substructures, the free-field motion of the wave is transformed into the equivalent seismic load, which is the seismic wave input data for the small-scale region of interest. Finally, with the help of local elements with special shapes to realize the grid transition of different scales, a small-scale model with reasonable soil-underground structure interaction is established, and the seismic response of the overall model can be effectively solved. The calculation and analysis of the seismic response of underground structures in irregular terrain are carried out. Compared with the results obtained directly from the source-site-structure model, the multiscale method has satisfactory accuracy and meets the needs of engineering design. Since the number of elements is fewer and the calculation time is much shorter than those required by the traditional model, the advantages in computational efficiency of the new method are highlighted. In addition, the reflected waves are too weak to have a considerable impact on structures because of the great energy loss at the reflection interface, which further proves the feasibility of the closed artificial boundary substructure method.

In this study, a soil–tunnel model for clay under earthquake loading is analyzed, using finite element methods and a kinematic hardening model with the Von Mises failure criterion. The results are compared with those from the linear elastic–perfectly plastic Mohr–Coulomb model. The latter model does not consider the stiffness degradation caused by imposing cyclic loading and unloading to the soil, whereas the kinematic hardening model can simulate this stiffness degradation. The parameters of the kinematic hardening model are calibrated based on the results of experimental cyclic tests and finite element simulation. Here, two methods—one using data from cyclic shear tests, and the other a new method using undrained cyclic triaxial tests—are used to calibrate the parameters. The parameters investigated are the peak ground acceleration (PGA), tunnel lining thickness, tunnel shape, and tunnel embedment depth, all of which have an effect on the resistance of the shallow tunnel to the stresses and deformations caused by the surrounding clay soils. The results show that unlike traditional models, the nonlinear kinematic hardening model can predict the response reasonably well, and it is able to create the hysteresis loops and consider the soil stiffness degradation under the seismic loads.

No abstract