A new framework for probabilistic analysis of the performance of a supported excavation in clay considering spatial variability

This technical note presents a new framework for probabilistic analysis of the performance of a geotechnical system such as a diaphragm wall-supported excavation. The new framework consists of two components: a sampling algorithm and an integration method. With the new sampling algorithm, samples of uncertain variables are ‘uniformly’ generated in the possible domain of uncertain variables; the new integration method is then used to construct the cumulative distribution function of the system performance, based upon the integration weights of the generated samples of uncertain variables. To demonstrate this new framework, an illustrative example focusing on a diaphragm wall-supported excavation in clay is analysed; the system performance of the diaphragm wall-supported excavation is studied with Flac (a finite-difference computer code) and the spatial variability of soil properties, in the context of numerical modelling, is modelled with randomfield theory. Finally, a parametric analysis is conducted to examine the influence of the statistical characterisation of soil properties on the results of the probabilistic analysis. The innovation of the new framework is demonstrated with the results presented.