Contribution of the bounding surface plasticity to the simulation of gallery excavation in plastic clays

Excavation of a gallery induces perturbations of the surrounding in-situ stress (σ) field and pore pressure (u_w) field, combined with the generation of displacements (u). In the framework of the underground storage for high level radioactive waste, an accurate assessment of these perturbations is important. In the past, analytical solutions (σ and u fields) have been derived, based on classical continuum mechanics (total stress analysis) and on simple constitutive models (elastic, Tresca, Mohr-Coulomb). More recently, these solutions have been extended to the undrained response (σ, u and u_w fields) of saturated porous media, still considering simple mechanical models. However, these solutions underestimate the displacements and the variations of pore pressure compared to the in-situ values monitored for the Boom clay, obtained during the excavation of the experimental Test-Drift (TD) gallery at Mol, Belgium. Here, we investigate whether such discrepancy could result from the sharp elastic-plastic transition on which classical plasticity models (of which the Cam-clay) are based. To take into account a smooth transition from elastic to plastic states, two bounding surface (BM) models based on the Cam-clay model are presented. The first one is a simplified version of the Dafalias-Kaliakin bounding surface model, and the second one is based on the Al Tabbaa's bubble model which has been generalised here. In the following, available in-situ data are briefly presented. Eventually, undrained hydromechanical simulations of the Test-Drift excavation are performed, considering the classical Cam-clay and the previously mentioned bounding surface models. As a main result, the radial convergence obtained with the bounding surface models is quantitatively much closer to in-situ measurements than results obtained with the Cam-clay model. Also, the extent of the hydraulic disturbance predicted from the retained bounding surface models is qualitatively closer to in-situ data than the value predicted from the Cam-clay one. In the future, these conclusions will be checked against new in-situ data which will be available for Boom clay in late 1999 within the CLIPEX (Clay Instrumentation Programme for the EXtension of an underground research laboratory) European project.