Probabilistic Slope Stability Analysis: Subset Simulation versus Monte Carlo approach

Understanding and accurately predicting slope failure probabilities is a major challenge in geotechnical engineering. This study presents a comparative evaluation of Monte Carlo Simulation (MCS) and Subset Simulation (SS) methods for slope stability analysis using the Limit Equilibrium Method (LEM). Recognizing the limitations of deterministic approaches, particularly under uncertainty, probabilistic analysis was adopted with input parameters modeled as lognormally distributed random variables. MCS, though intuitive and widely accepted, demands extensive computational effort, especially for low-probability failure events. Alternatively, SS improves efficiency by progressively guiding sampling toward the critical failure domain. The study also accounts for the correlation between cohesion and friction angle, demonstrating its measurable impact on failure probability and result variability. Findings revealed that SS achieves comparable accuracy to MCS while dramatically reducing the number of required simulations and computational time. Probability density functions and cumulative probability curves for both methods were analyzed, indicating close agreement. These results demonstrate the potential of Subset Simulation as a computationally efficient and robust alternative for probabilistic slope stability assessments, especially for rare-event analysis.

کلیدواژه ها:

Slope stability ، Limit equilibrium ، Stochastic ، Monte Carlo simulation ، Subset simulation