Many mechanical, tribological and geophysical problems bring difficulties such as large deformations, fracture, multiple localisations or wear. To fill the lack of knowledge and to bring information in fields where experimental approaches are limited (for example, contact interface measures), numerical tools have been developed and adapted. But if for some applications efficient numerical methods are now well identified, it still lacks tools able to cope with a class of multiproblems: multibodies, multicontacts, multiscales and multiphysics. Considering this kind of problems, needs to widen the purpose of classical modelling approaches, mixing continuous medium and discrete medium descriptions. The relevance of continuous models at the structure level and the discrete models at the microstructure level shows the necessity to converge to hybrid models able to switch between continuous and discrete descriptions if necessary. Moreover, it needs to improve theoretical and numerical related tools.

This part of my research work is to build an unified modelling framework able to cope with extended discrete element method: hybrid DEM/FEM models, multiphysics applications, constrained/unconstrained modelling of contact, various bulk behaviours and interaction laws, etc. Various applications are involved behind such kind of modeling such as fracture mechanics, geophysics and tribology.

A focus is performed on multiphysics application involving heat transfers and heat generation in discrete assemblies but also electrical transfert, physico-chemical aspects coupled or not with other physics.