Application of risk analysis to the prediction of slope instability

Publisher Website
Google Scholar
Abstract
The problem of the prediction of stability or instability of natural clay slopes is examined in view of the random intrinsic nature of both soil properties and external actions. The probabilistic method of analysis appears to be a useful tool, which not only could account for these random properties but also could consider uncertainties derived from incomplete knowledge of pertinent model parameters and conditions of stability.Using the familiar method of slices, the different sources of error have been incorporated into a first-order probability analysis of the simplified Bishop model in order to arrive at quantitative information concerning the probability of failure. Field and laboratory data from an instrumented test valley slope in the Ottawa region have been considered to arrive at an instability risk prediction of the test slope. The mean functions of the strength parameters have been made explicitly dependent on a number of statistical parameters to emphasize dependence on available data. The analysis distinguishes between statistical uncertainty, which comes from insufficiency of available data, and probabilistic uncertainty, which is a measure of random heterogeneity of the soil.

This publication has 0 references indexed in Scilit: