System reliability analysis of slopes based on the method of slices using sequential compounding method

Reliability analysis of slope stability using limit equilibrium approaches without considering the correlation between slices can lead to obtaining unrealistic reliability indices. In this paper, considering the method of slices and by selecting stochastic soil parameters, the most critical failure surface is determined by the particle swarm optimization algorithm. The procedure is continued up to 2,000 iterations for an arbitrary slope. Then, for each critical failure surface, the reliability index is determined considering the slices as parallel components of a system with cross-correlation. Through this procedure, the effects of safety margin against the instability of each slice can be implemented to analyze the slope stability. The system reliability index of the slope is finally determined by combining the reliability indices of the critical surfaces based on their corresponding correlation by the Sequential Compounding Method (SCM) as a powerful technique for reducing the time and computational cost. The results are verified by Monte Carlo Simulation (MCS); however, this method is quite time-consuming for this purpose.