Evaluation of various seismic response analysis methods for underground structures in saturated sand

Abstract

Underground structures in liquefiable saturated sand are prone to earthquake damage. Various simplified pseudo-static seismic response analysis methods for underground structures have been recommended by current design codes and guidelines, but lack adequate systematic validation in liquefiable ground. In this study, the effectiveness of four categories of simplified pseudo-static methods in saturated sand are evaluated. These methods include a displacement-based method, a force-based method, a F(flexibility ratio)-R(racking ratio) method, and a detailed equivalent static method. High-fidelity dynamic analysis of circular and rectangular cross-section underground structures in non-liquefiable and liquefiable ground under different input ground motions are conducted, using a validated plasticity constitutive model for large post-liquefaction deformation of sand, to serve as a basis for the evaluation. Although the pseudo-static methods can generally provide adequate analysis results for underground structures in non-liquefiable ground, especially the displacement-based method, their performance in saturated sand are generally poor. The deviation of pseudo-static analysis results from reality stems from their inadequacies in considering the effects of dynamic soil-structure interaction in liquefiable ground. Temporal and spatial variability of soil-structure interaction in saturated sand, caused by the accumulation of excess pore pressure and subsequent decrease in soil effective stress and constraint on underground structures, are not considered in pseudo-static analysis methods. Modifications to the pseudo-static analysis methods and high-fidelity dynamic analysis methods are suggested to be adopted for the evaluation of the seismic response of underground structures in liquefiable ground.