Earthquake response of retaining walls

Abstract

This study investigates the background and the earthquake response of earth retaining walls by applying the pseudostatic Mononobe-Okabe method and the finite elements method. For the finite elements method the equivalent linear approach was used. The methods were compared to each other against and to simple cases where the behavior is known. Equivalent linear parameters were determined by free field response analysis, and analyses were done to investigate how a finite elements model should be set up so equivalent linear parameters are valid. The methods were compared to cases where analytical solutions exists and to cases reported in scientific literature. The equivalent linear method was found to reasonably reproduce the results of a more complex nonlinear model. Validation of using the finite elements method for analysis of retaining walls was attempted by reproducing a centrifuge model test with a shaking table. However similar behavior could not be obtained.

The established method was applied on a real case; a 16 meter tall gravity retaining wall, set on firm rock and supporting a rock fill for a road. The safety was found to be sufficient. The effects of EPS geofoam were also modeled for the case, using a linear elastic model for the EPS. The seismic thrust was significantly reduced; however the internal stability of the retaining wall was lowered. Different types of EPS and different shapes were analyzed, and it was found that the least dense EPS was the most efficient at reducing seismic thrust. Low density EPS implies soft EPS with low Poisson’s ratio. For the different shapes of EPS tested, the constant width, spanning the entire height of the wall was found to be the most efficient.