Determining the Factors of Safety of Spatially Variable Slopes Modeled by Random Fields

The merits of incorporating the spatial variability of soil properties into slope reliability analyses have been demonstrated by many researchers. However, the suitability of the use of limit equilibrium methods (LEMs) for calculating the minimum factor of safety (FS) of a spatially variable slope modeled by a random field is not well understood. The present paper investigates this topic by performing a series of analyses using circular and noncircular LEMs and a strength reduction method based on finite-element analysis (FE SRM). The LEMs are found to encounter three possible issues in random fields, including (1) noncircular slip surfaces, (2) a significant increase in the complexity of an optimization problem, and (3) failure to converge. The LEM analysis results in random fields can only be reliable when these three issues are adequately addressed. It is found that a noncircular rigorous LEM with a carefully selected interslice force function may work adequately in random fields. Moreover, the FE SRM is suitable for random field cases because of its robustness, accuracy, and versatility.