Probabilistic seismic slope stability analysis adopting limiting profile approach

This paper communicates a futuristic limiting slope profile (LSP) guided probabilistic seismic stability assessment of soil slopes within an integrated framework of the Latin hypercube sampling with dependence (LHSD) and the slip line theory (SLT). The LHSD improves computing efficiency by employing fewer model simulations, whereas the SLT provides a robust slope stability estimate with an adaptive critical slip surface (CSS). Besides, the SLT directly derives the LSP, which preserves the state of limiting equilibrium with a factor of safety (FS) of 1.0. Consequently, the probability density function (PDF) of the overall slope angle (OSA) of such derived LSPs is obtained as the stochastic model response. Such model response eventually offers a more rational approach to compute the failure probability (Pf) of any given slope angle (alpha L), evading the need for separate probabilistic analyses for each alpha L, as observed in the traditional practice. Instead, the Pf for a given alpha L is calculated as the proportion of area under the obtained PDF curve corresponding to the OSA less than equal to alpha L. The feasibility of such an approach is demonstrated by investigating the Pf of a reported planar slope example and comparing the results with the literature. The potential applications of the approach are illustrated through a real-slope case study and a reliability-based slope design problem. Further, design charts are also provided for ease of practising engineers.