Verification of a stochastic dynamic analysis method by shaking table tests of a slope retaining system

Abstract

Various sources of uncertainty can greatly affect the seismic responses and seismic reliability of slope retaining systems. Many studies have observed and highlighted that ignoring the effect of uncertainties, especially the randomness of ground motions, would cause an inaccurate evaluation of seismic performance of slope retaining system. It is thus of great significance to employ the stochastic dynamic analysis method to quantify these uncertainties and corresponding effect. A recently developed stochastic dynamic analysis method, namely probability density evolution method (PDEM), have been proposed for the stochastic seismic responses analysis of geotechnical engineering. Although its efficiency and accuracy have already been proved by numerical cases, its feasibility and accuracy has not been verified from the experimental perspective, and its application in the stochastic seismic responses analysis using the field monitor data and experiment results in geotechnical engineering has not yet been studied. In the present study, a hard-strength rock slope model reinforced by pile-anchor structures is firstly constructed for the repeating shaking under a series of ground motions. A set of fully non-stationary stochastic ground motions are generated to quantify the randomness of input ground motions. Then, the stochastic seismic responses of the slope retaining system, including acceleration responses, amplification effect and earth pressure along pile-anchor structures, are analyzed and compared by the PDEM and various assumed probabilistic distribution models to illustrate the feasibility and accuracy of PDEM from the experimental perspective. The results demonstrate that the PDEM can obtain relatively more accurate stochastic seismic response results, and can also be used to analyze the experimental data from the probabilistic perspective.