Probabilistic evaluation of the seismic stability of infinite submarine slopes integrating the enhanced Newmark method and random field

Abstract

For evaluation of the seismic stability of submarine slopes, traditional deterministic methods may not reflect the actual situation due to the uncertainties in input values such as seismic acceleration, soil properties, and hydraulic conditions. In this study, probabilistic analyses were conducted based on the use of an enhanced Newmark method to evaluate the seismic stability of submarine slopes. In the probabilistic workflow, the infinite slope model was used, whereby the spatially varied strengths of marine sediments were simulated by non-stationary random fields discretized by the Karhunen-Loeve expansion. The positions of the potential slip surfaces were searched automatically, rather than being predefined as in the traditional limit equilibrium method. Artificial earthquake accelerations, exhibiting the same spectral characteristics, were simulated as the input ground motions in the probabilistic analysis. The pore water pressure generation and dissipation models were incorporated into the enhanced Newmark method when calculating the permanent displacements of the submarine slopes. Monte Carlo simulations were conducted for statistical characterization of the slope displacements and their failure probabilities. The results of the analysis indicate the superiority of the probabilistic framework and demonstrate the significant effects of pore water pressure and the spatial variability of the soil strength on the displacements and failure probabilities of submarine slopes.