3D Modelling of Excavation-Induced Anisotropic Responses of Deep Drifts at the Meuse/Haute-Marne URL
Résumé
Understanding the excavation-induced fractured zone (EFZ) around drifts is paramount in the context of the deep geological disposal for nuclear waste since fractures can introduce pathways for the migration of radionuclides. Drifts in the Meuse/Haute-Marne Underground Research Laboratory (URL) have been essentially excavated following the two main directions of major and minor horizontal stresses. Field observations on the two drifts GCS (parallel to major horizontal stress direction) and GED (parallel to minor horizontal stress direction) in the URL show anisotropic shapes of EFZ around drifts through both orientations and anisotropic convergences. These anisotropic responses resulted from the inherent and/or induced anisotropies of the host rock as well as the anisotropic stress field. This study focuses on 3D numerical modelling of excavation-induced anisotropic responses including shape and extent of EFZ, and short-term convergences of drifts. The main assumption is that the failure of claystone material is due to fracturing along weakness planes (ubiquitous joints) and the failure of the rock matrix. The ubiquitous joint failure is represented by perfectly plastic models for both tensile and shear yield functions. Their orientation is determined from the stress state based on the fracture mechanics, which includes tensile, longitudinal splitting and shear (conjugate planes) cracks. The rock matrix is assumed to be elastoplastic with hardening, softening and residual behaviours. Confining pressure dependency for the post-peak behaviour with a brittle–ductile transition is taken into account for the rock matrix. The proposed model is implemented into a commercial numerical software FLAC3D. The main features of the implemented model are shown by the simulation of laboratory triaxial compression tests, as well as field observation within the URL. In particular, comparisons between 3D simulations of GCS and GED drifts with in situ observations shows promising results, which demonstrates advances of present model with respect to existing models.
Origine : Fichiers produits par l'(les) auteur(s)