|M.Sc Thesis||Department of Civil and Environmental Engineering|
|Supervisor:||Prof. Emeritus Baker Rafael|
Most practical slope stability calculations are based on the linear Mohr-Coulomb failure criterion. However, many experimental studies showed that failure criteria of most soils are not linear when large ranges of normal stresses are considered. This non linearity is particularly evident in the range of small normal stresses. Non linearity of the failure criterion in the small normal stress range is particularly significant for slope stability calculations since in many practical stability problems critical slip surfaces are shallow and normal stresses acting on these surfaces are therefore small.
Experimental evaluation of soil strength at very small normal stresses is not trivial, involving various technical difficulties, and common strength tests are usually performed at relatively large normal stresses. This practice results in a discrepancy between the normal stress range at which tests are done and the stress range relevant for the stability problems. The present work shows that in certain cases such a discrepancy may result in a significant over estimate of safety factors and critical heights.
The present work presents an experimental study of the strength of compacted Israeli clay at low normal stresses. Linear and non-linear strength laws fitted to the same experimental data set predict similar strength values in the range of normal stresses containing the most of the data points. Predictions of such models differ from each other mainly at very small and very large normal stresses (which do not include actual experimental data). In both of these ranges the non-linear strength function predicts smaller strength values than Mohr-Coulomb. In order to ensure a safe design in cases where the experimental and relevant normal stress ranges are not the same it is necessary to base the stability analysis on the non-linear strength law.