Hydro-elasto-viscoplastic modeling of a drift at the Meuse/Haute-Marne underground research laboratoratory (URL)

• Published in: Computers and geotechnics 2017-05, Vol.85, p.306-320
• Main Authors: Souley, M., Armand, G., Kazmierczak, J.-B.
• Format: Article
• Language: English
• Subjects: Clay; Computer simulation; Connectors; Couplings; Damage; Dredging; Drift; Earth Sciences; Excavation; Fracturing; Geomechanics; Hydraulic properties; Hydromechanical and viscoplastic behavior; In situ measurements; Laboratories; Mathematical models; Mechanical properties; Modelling; Numerical implementation; Numerical modeling; Parameter identification; Parameters; Permeability; Plastics; Properties; Radioactive wastes; Sciences of the Universe; Strain; Underground construction; Waste disposal
• ISSN: 0266-352X; 1873-7633
• DOI: 10.1016/j.compgeo.2016.12.012

An underground research laboratory (URL) is being constructed by Andra in eastern France, in Callovo-Oxfordian claystone (COx) in which various in situ geomechanical experiments are being undertaken or are to be carried out. The aim of this URL is to characterize the in situ properties of COx claystone and to test disposal technologies in a realistic way in order to assess the short- and long-term safety of a deep radioactive waste repository. In parallel, theoretical and numerical models able to reproduce the phenomena observed under different types of loading paths must be developed. The phenomenological elastic-visco-plastic model developed by Souley et al. (2011) has been enhanced to reflect recent advances in understanding of the mechanical and hydromechanical behavior of COx claystone and the modification of the mechanical and hydraulic properties in the EDZ (Excavation Damage Zone). In particular, the influence of induced damage and fracturing on the delayed strains and strain rates of the COx claystone and the permeability changes observed at the site scale, as well as hydro-mechanical couplings, are discussed and incorporated in a new model. This model is implemented into the commercial code FLAC3D. Short- and long-term test data (Armand et al., 2016) can be used to identify possible key parameters for the model. These tests were also used to identify certain parameters of our model. Some tests were simulated to verify the numerical implementation of the proposed model. Finally, the simulation of the GCS drift excavation (Seyedi et al., 2016) has been performed. Comparisons to in situ measurements are discussed and some accordance and discrepancies were observed.