Shaking Table Experiments of Reinforced Soil Retaining Walls

By: Saman Zarnani

ABSTRACT

The superior seismic performance of relatively ductile geosynthetic reinforced soil (GRS) retaining walls has been demonstrated in the literature by comparison to the poor performance of relatively rigid conventional gravity-type retaining walls subjected to the same earthquake event. Nevertheless, there are deficiencies in the current design of these systems and fundamental lack of understanding of the mechanics of these complex systems during earthquake loading. Shaking table experiments on reduced-scale models are the most practical approach to gain further qualitative and quantitative understanding of the behaviour of GRS walls under seismic loading. Most experimental tests on seismic behaviour of GRS walls have been performed on very small-scale models (0.5 ~ 1 m) where scale effects are expected to have a major influence on measured response.

This presentation describes details of a large shaking table-GRS model wall design, instrumentation and monitoring program developed at the Royal Military College of Canada (RMC) and two example test results. The testing program is unique in the literature because of the large number of different instruments deployed and the use of Particle Image Velocimetry (PIV) analysis of imagery captured using a high speed camera. The models are instrumented with strain gauges and extensometers attached to the geogrid reinforcing layers, LVDTs attached to the facing panel, load cells at the wall toe, reinforcement-facing load measurement, and accelerometers in the backfill and along the facing. The example model walls described here are 1.42 m in height and are part of a series of experiments that are underway to investigate seismic performance of GRS walls. The two models that are described here are identical except the boundary condition at the toe of the wall. In one model (hinged wall) the wall facing is allowed to only rotate around its base (without any horizontal sliding) and in the other model (sliding wall) the wall facing is allowed to slide horizontally in addition to rotation around its base. The effect of this toe boundary condition is illustrated through experimental results and it is shown that this parameter has significant effect on the overall behaviour of GRS model wall under dynamic loading conditions.