Effects of soil spatial variability on soil-structure interaction

Nobahar, Arash (2003) Effects of soil spatial variability on soil-structure interaction. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (9MB)

Abstract

Many physical systems in general, and soil materials in particular, exhibit relatively large spatial variability in their properties, even within so-called homogeneous layers. Physical descriptions of this spatial variability are not feasible owing to the prohibitive cost of sampling and uncertainty induced by measurement errors. This variability is widely dealt with as uncertainty in soil properties. Probabilistic methods currently used to represent this uncertainty often suffer from many limitations. For instance, they often only account for uncertainty in estimating the average soil properties. A probabilistic approach was developed here to investigate the effects of soil heterogeneity and provide practical recommendations and guidelines to account for these effects in routine engineering design. -- There are still many unknown consequences of spatial variability. It is shown here that natural variability of soil properties within geologically distinct and so-called uniform layers affects soil behaviour. This study found that the phenomena governed by highly nonlinear constitutive relations are the most affected by spatial variability of soil properties. The bearing capacity of shallow foundations and lateral interaction loads of buried pipelines are functions of soil shear strength and, therefore, are governed by highly nonlinear stress-strain relationships. -- The effects of soil heterogeneity were investigated for a strip foundation placed on elastic perfectly plastic soil and subjected to vertical loads. From a comparison of Monte Carlo simulations, accounting for the spatial variability of soil strength, and deterministic analyses assuming uniform soil properties, it was found that the soil heterogeneity changes the mechanical behaviour of foundations. A parametric study was performed to quantify the effects of soil heterogeneity parameters on foundation response; the studied cases were pre-designed using statistical methods (Design of Experiments, DOE). It was observed that soil strength's degree of variation and probability distribution, which characterize the amount of weak pockets of soil, have the most effects on the foundation behaviour for the range of parameters considered. Correlation distances also affected the variability of foundation responses owing to local averaging effects. -- The results of the parametric study are presented as simple regression equations (response surfaces) to estimate probabilistic characteristics of foundation responses - namely mean and coefficient of variation of bearing capacity and bearing pressures at damage criteria. They were used to calibrate partial design factors for limit state design methods, LSD, and estimate characteristic values for routine engineering design. The results, in terms of regression equations, can also be employed directly in level II & III reliability analysis methods. -- A similar study with a limited scope was performed for lateral loading of a buried pipeline. Only one burial depth (geometrical configuration) was taken for the pipeline. Among the probabilistic characteristics of soil considered here, the degree of variability of soil strength was found to be the most significant factor affecting pipeline response. The response and failure mechanism of a laterally loaded buried pipeline is complicated and is dependent on several deterministic factors such as burial depth, pipe-soil interaction coefficients, and soil weight. The study could be further developed to account for other probabilistic characteristics and deterministic parameters, and their corresponding interactions.

Actions (login required)

View Item