Search

Realistic prediction of seismic landslides is critical for performance-based design in the seismically active regions. To date, analytical methods for estimating seismic landslides have been based on simplified models. Many gaps still remain in the scientific understanding of earthquake-induced landslides, especially the landslide triggering process and post-failure behavior. In this study, a strain-softening fully nonlinear dynamic soil model is formulated using the material point method (MPM) to simulate the soil slopes failure process under dynamic loading. The study demonstrated that MPM can capture entire slope failure process, including slide triggering, shear band formation, runoff and final deposition. The complicated interaction between sliding masses can also be captured. Numerical simulation also demonstrated that residual soil strength is an important factor in determining landslide runoff and deposition. The study demonstrated the great promise of MPM method in improving our understanding of the coseismic landslide process and these underlying influential factors.