Probabilistic Analysis of Cantilever Sheet Pile Walls Using Copula Models

Abstract

Stability check of cantilever sheet pile walls by considering the variation in the soil properties shows that a minor change in value of properties can cause of failure of the whole structure. Factor of safety is used in deterministic analysis to account for the variability, but the probabilistic design is more practical compare to deterministic design, as the inherent variability is considered in probabilistic method. The objective of this paper is that the probabilistic analysis of cantilever sheet pile walls against overturning failure mode considering variability of soil parameters and water table position. To depict the real condition on cantilever sheet pile walls, the soil properties (including cohesion, friction angle, unit weight) and water table depth ratio (ratio between water table depth from the top of the wall and wall height above dredge line) are considered as random variables. The correlation of soil parameters likes cohesion, friction angle and unit weight is measured by multivariate copula models. Normal copula under elliptical class and Studentst, Gumbel, Frank copula under Archimedean class are used. The parameters of the random variables are collected from the existing literature. The probability of failure against overturning mode is calculated by Monte Carlo simulation technique. In this paper a probabilistic model accounting correlation is proposed based on copula theory for designing cantilever sheet pile walls. The proposed method can be very effective in performing a site specific probabilistic analysis of it embedded in different soil conditions and variable water table position.