A Probabilistic Modeling of Soil Variability: Effects of Random Fields on Reliability Index and Failure Probability

This study investigates the probabilistic modeling of soil variability using random field methods and its influence on the reliability index and failure probability of geotechnical structures. Soil variability is modeled through random fields, accounting for the spatial distribution of properties such as cohesion, internal friction angle, and unit weight, while incorporating uncertainties, spatial correlations, and threshold effects. Statistical and numerical analyses reveal that threshold effects significantly impact system stability when critical parameter values are reached. Notably, at coefficients of variation (COV) of ۲۰% and ۳۰%, the probability of failure at slope angles of ۶۰ and ۷۰ degrees was lower than at ۴۵ degrees, attributed to soil cohesion characteristics and threshold effects. Furthermore, an increase in correlation length initially reduces failure probability before stabilizing it. The results indicate that for a ۴۵-degree slope angle and ۳۰% COV, increasing the slope angle leads to unsafe conditions. These findings enhance geotechnical design and evaluation methods, contributing to improved risk assessment methodologies and fostering safer, more resilient geotechnical structures.

کلیدواژه ها:

Uncertainty ، Random field ، probability of failure ، Reliability Index ، threshold effects