Discrete-Element Method Study of the Seismic Response of Gravity Retaining Walls

Abstract

A three-dimensional microscale framework utilizing the discrete-element method is utilized herein to analyze the seismic response of a soil-retaining wall system in the time domain. The granular soil deposit is idealized as a set of spherical particles while the retaining wall is modeled as a rigid body composed of glued particles to yield the physical characteristics of a real-life gravity-type retaining structure. The model inherently accounts for the influence of material nonlinearity, potential phase shift between the wall and soil deposit, and viable failure of the wall by overturning or slippage. The effect of the amplitude and frequency of the input dynamic excitation on the response of the system is studied. A comparison is made between simulation results and those of the popular pseudostatic Mononobe–Okabe (M–O) method. This comparison showed that, in general, the M–O method is conservative. However, soil thrust and deformation of the structure showed dependence on the frequency of the input motion at the same amplitude and peaking at frequencies of input motion close to the natural frequency of the deposit. The simulations also showed that soil pressure distribution on the sides of the wall changed with movement and displacement of the wall. Maximum soil thrust occurs when the wall moves toward the backfill, and minimum soil thrust when the wall moves away from the backfill. Additionally, residual earth pressure may increase after shaking. Finally, the phase difference between local maxima time instances in wall acceleration and corresponding local minima in total dynamic soil thrust was also observed.