Seismic performance of reinforced earth walls (1989)
AuthorsFairless, G. J.show all
Previous seismic testing of reinforced earth walls, and present design methods, are outlined. Six one metre tall model reinforced earth walls were tested under normal gravity on a shaking table. Accelerations, displacements and reinforcing strip forces were measured. A series of simple sinusoidal single-pulse wave forms, and the El Centro 1940 north-south recorded earthquake motion were used as input to the shaking table. Permanent outward displacement of a block consisting of the facing, part of the reinforced block, and a wedge of retained soil behind the reinforced block occurred when a limiting or yield acceleration was exceeded. The wall facing remained essentially vertical. The critical (minimum) yield acceleration was found to occur when and after the failure surface outcropped at the fill surface. This critical yield acceleration was calculable using a limiting equilibrium formulation, within the variability of the observed results. A sensitivity analysis of the formulation is presented. The apparent soil friction angle is found to reduce during repeated seismic shaking, while the apparent soil-strip friction coefficient is found to increase probably to a limiting value. In design, the peak value of the friction coefficient found from direct shear tests between the soil and reinforcing can be used. To calculate design strip forces, an upper bound seismic earth pressure coefficient based on the Mononobe-Okabe dynamic earth pressure coefficient KAE is proposed. Measured displacements are plotted on charts for three sliding block displacement prediction methods. One approach, using random vibration and probability theory, is more rational and complete than the others, and provides an estimate of the prohability of exceedence of the calculated displacement. The method is quite complicated, however, and for everyday design a simple upper bound on a chart derived from an equivalent pulse technique is recommended.