Probabilistic assessment of seismic stability of a rock slope by combining the simulation of stochastic ground motion with permanent displacement analysis

Abstract

A probabilistic framework that considers the variability and uncertainty of the ground motions caused by earthquakes and the parameters of the slope material is proposed to assess the seismic stability of rock slopes. A stochastic ground motion simulation method, which reflects certain earthquake scenarios and site characteristics, was used to generate accelerograms as seismic inputs. Then, the probabilities of failure (Pf) of the slopes when subjected to earthquake ground motions were computed via Monte-Carlo simulations based on the simulated ground motions and the decoupled Newmark permanent-displacement analysis. An example rock slope was used to illustrate the proposed procedure and verify its rationality. The results showed that the magnitudes of the earthquakes and the source-to-site distances had significant influences on the seismic stability of the slopes. For an earthquake of the same magnitude, the Pf of the slope decreased rapidly as the source-to-site distance increased. In addition, given the same source-to-site distance, the Pf of the slope increased steadily as the magnitudes of the earthquakes increased. Also, it was found that the influence of coefficient of variation of the slope's shear strength parameters (i.e., c, φ) was not negligible when the uncertainty of earthquakes was not considered. However, it may be overshadowed by the significant influence of a great earthquake, especially when the variability of ground motions is taken into account. The results that were obtained agreed well with observed earthquake-induced landslide phenomena and the statistical data of two typical earthquakes that occurred in Japan and China, which implies that the proposed method is accurate.