CMS-9613906, Ronaldo Borja, Stanford U. The objective of this research program is to develop a mathematical model for analyzing the problem of lateral flow and liquefaction-induced large ground movement during and following an earthquake. The model is based on a two-phase mixture theory, and involves the development of a finite element code capable of representing the behavior of saturated soils prior to and during liquefaction. Prior to liquefaction, the soil is represented by an already- validated effective-stress phenomenological constitutive model based on an elastoplastic formulation with nonlinear kinematic hardening. During liquefaction, the liquefied soil is represented by a total-stress viscoplastic model whose viscosity is determined from extensive numerical simulations with random suspensions of spheres. The two constitutive models are developed within the framework of a finite deformation theory based on multiplicative plasticity. The resulting finite element model is used to study the mechanism of liquefaction-related large ground displacement, including the lateral flow and prediction of the free ground surface resulting from such flow. Two case studies are considered in order to validate the model. The first involves a loose saturated sample of Toyoura sand submerged in water and sloped inside a test subjected to dynamic excitation, causing the soil to liquefy and flow like a viscous fluid. The second study involves a re-analysis of the nonlinear ground response recorded by a downhole array in Port Island (reclaimed ground near downtown Kobe City, Japan) during the earthquake of January 17, 1995, focusing on the impact of this changed soil behavior during liquefaction. The ground response is simulated assuming a condition of vertically propagating seismic waves, with a full kinematical coupling of the three components of motion.
