5. Conclusions
Numerical analysis on the seismic response of a one-story and twospan underground structure is conducted in this study for three different soil profile conditions, including a layered liquefiable ground, a homogeneous liquefiable ground, and a homogeneous non-liquefiable ground. PEER database was used to select and scale seven linear matched ground motions to the UHS of a particular site, for the analyses. A comprehensive plasticity model for large post-liquefaction shear deformation of sand was used for the liquefiable soil, and an elastic perfect-plastic Mohr-Coulomb model was used for the non-liquefiable soil. A new combined structure element modelling technique was used to represent both the cross-section properties and the volume of the reinforced concrete, by tying together non-linear fiber beam elements with quadrilateral elements. Drift ratios, bending moments, and shear forces of the underground structure were used to evaluate its dynamic response.
For the shallow buried underground structures, the existence of the liquefiable soil around the structure could cause the structure to be more prone to failure compared with the same structure in non-liquefiable ground. In this study, for the structures in liquefiable ground, the bending moment in several parts of the structure reached or almost reached the moment capacity, and the shear force in the central pillar in the M-liq condition was close to its shear capacity. However, the drift ratio of the structure was well within the limit value, suggesting that the current limitation value requirement for underground structures may need to be reduced.
