Comparison of numerical and analytical solutions for reinforced soil wall shaking table tests

Abstract

The paper describes a simple numerical FLAC model that was developed to simulate the dynamic response of two instrumented reduced-scale model reinforced soil walls constructed on a 1-g shaking table. The models were 1 m high by 1.4 m wide by 2.4 m long and were constructed with a uniform size sand backfill, a polymeric geogid reinforcement material with appropriately scaled stiffness, and a structural full-height rigid panel facing. The wall toe was constructed to simulate a perfectly hinged toe (i.e. toe allowed to rotate only) in one model and an idealized sliding toe (i.e. toe allowed to rotate and slide horizontally) in the other. Physical and numerical models were subjected to the same stepped amplitude sinusoidal base acceleration record. The material properties of the component materials (e.g. backfill and reinforcement) were determined from independent laboratory testing (reinforcement) and by back-fitting results of a numerical FLAC model for direct shear box testing to the corresponding physical test results. A simple elastic-plastic model with Mohr-Coulomb failure criterion for the sand was judged to give satisfactory agreement with measured wall results. The numerical results are also compared to closed-form solutions for reinforcement loads. In most cases predicted and closed-form solutions fall within the accuracy of measured loads based on I standard deviation applied to physical measurements. The paper summarizes important lessons learned and implications to the seismic design and performance of geosynthetic reinforced soil walls.