Reliability Assessment for Time-Dependent Behaviour of Soft Soils Considering Cross Correlation between Visco-Plastic Model Parameters

Abstract

ABSTRACT An elastic visco-plastic creep model was combined with the Monte-Carlo probabilistic method incorporating multivariate copula and nonlinear analysis to investigate the effects of uncertainties in the elastic visco-plastic model parameters on time-dependent settlement and the distribution of excess pore water pressure in soft soils under applied loads. The elastic-plastic model parameter ( λ / V ) and creep coefficient ( ψ 0 / V ) were considered as random variables with lognormal distribution while considering the cross correlation between these two random variables. When λ / V and ψ 0 / V were used as random variables, the coefficient of variation of time-dependent deformation gradually decreased approximately 25% over time until reaching an asymptote. By adopting over 50 years of monitoring data from the case study of Vasby test fill and results from the settlement ratio, the most appropriate cross correlation coefficient between selected random variables was introduced. The results revealed that increasing the cross correlation coefficient between λ / V and ψ 0 / V increased the standard deviation and the coefficient of variation of settlement up to 40%. Meanwhile, the corresponding statistical features for the predicted excess pore water pressure decreased as the cross correlation coefficient increased. This study also provides a practical insight into selecting the most suitable cross correlation coefficient between elastic visco-plastic model parameters, while adopting reliability-based design approach that captures the time-dependent deformation of embankments and structures built on soft soils.