Refining 3D nonlinear finite element settlement predictions around a deep urban excavation with the help of a Bayesian approach

Abstract

This contribution presents a probabilistic and Bayesian approach for the finite element modelling of a deep excavation in a lacustrine environment in Geneva, Switzerland. The project is composed of a new 25-storey tower and two new activities buildings leading to a 15 to 19 m deep excavation. A 100 cm thick diaphragm wall has been designed, with three to four strut levels. Different finite element models have been constructed with ZSOIL in order to help the design team predicting surface settlements around the project as well as bending moments in the wall and normal forces in the struts: a full 3D model, and several 2D cuts and 3D slices. The mechanical behavior of the soil is simulated by the Hardening Soil with Small Strain extension constitutive law allowing a proper prediction of soil displacements.

A deterministic approach has been used for the first design steps. Then, a priori sensitivity and reliability analyses have been conducted on the 3D finite element slices with UQLab, in order to identify which geotechnical parameters are the most significant, and evaluate the probability that the neighboring sensitive structures’ (a water pipe and the Geneva tramway) settlements would exceed acceptable values, leading to failure.

Coupling non-linear 3D finite element modeling with highly accurate meta-modeling (polynomial chaos expansion and kriging techniques) techniques is shown to be necessary in order to run Sobol’ sensitivity and Monte-Carlo reliability analyses in a reasonable amount of time.

A priori input parameters’ probability density functions were based on geotechnical site tests and literature. Displacement and pore pressure measurements from on-site monitoring were subsequently used during the first excavation phases in order to conduct an inverse Bayesian analysis. This helped us refine our prior knowledge for final settlements’ probability density functions and associated failure probabilities. In particular, the need to install the fourth level of bracing has been evaluated in the light of this risk analysis.