System reliability analysis of the external stability of reinforced soil structures

Abstract

A probabilistic model is formulated in order to assess the reliability of the external stability of reinforced soil structures, in the presence of geotechnical and material uncertainty. Three modes of failure are taken into account (bearing capacity, sliding and excessive eccentricity) and overall stability is modelled as an in-series system. Geomaterial shear strength properties are modelled as beta-distributed random variables and system reliability is computed using a Monte Carlo simulation algorithm that does not require arbitrary assumptions of independence between potential failure modes. Dependency between failure modes is evaluated in terms of coefficients of linear correlation. As an illustration of the developed methodology, a case example is analysed and results are discussed in the context of system reliability analysis. For this example, the results demonstrate that the bearing capacity failure mode is subjected to a higher degree of uncertainty, compared to sliding and excessive eccentricity. The computed system probability of failure indicates that commonly accepted simplifications of complete independence or perfect correlation between the failure modes (expressed through first-order reliability bounds) may not be appropriate, particularly when the reliability bounds are not close to each other. On the other hand, the second-order reliability bounds seem to accurately reflect the computed system probability of failure. Finally, in the present study, the results indicate that ignoring the cross-correlation between the shear strength parameters has an impact on the computed system probability of failure.