Probabilistic Analysis of Unreinforced and Reinforced Soil Slopes Using the Random Finite Element Method
The focus of this thesis is probabilistic analysis of unreinforced and reinforced soil slopes using the random finite element method (RFEM). RFEM combines the finite element method (FEM) with soil properties that are described by either random (homogeneous) or spatial variability. The RFEM approach is used for the first time to investigate the influence of spatially variable cohesive soil strength on probabilistic margins of safety for bearing capacity failure of a footing located close to the crest of a slope. Analyses showed that the coefficient of variation and the spatial correlation length of soil cohesion can have a large influence on footing bearing capacity, particularly for slopes with large height to footing width ratios. Probabilistic slope stability analyses of simple geosynthetic reinforced soil slopes were carried out for the first time using the shear strength reduction method in combination with the RFEM approach. The FEM code was modified to include bar elements to model horizontal layers of geosynthetic reinforcement. Analysis results demonstrate that large reductions in probability of failure can be realized by adding geosynthetic reinforcement layers to constructed slopes. The same modified RFEM code was used to carry out probabilistic stability analyses of constructed wrapped-face reinforced slopes (or embankments) using frictional soils. The influence of fill placed at different layer thickness and compacted to different levels was simulated by adjusting the soil strength and unit weight, and the vertical spatial correlation length in the anisotropic spatially variable strength field used in each slope realization. Numerical results showed that vertical spatial correlation lengths approaching the magnitude of fill lift heights can control the probability of failure for reinforced slopes constructed with weak fills placed in lift heights close to but less than the wrapped reinforcement spacing used in the study. The RFEM codes developed in this study are used next to carry out numerical simulations of physical full-scale unreinforced and reinforced embankments loaded to failure using a strip footing. The results of analyses showed that the assumption of vertical spatial variability in soil properties due to compaction was able to capture the observed failure mechanisms in these structures.