Ever-growing computational power allows to simulate physical phenomena of increasing complexity. In particular, it can be interesting to take into account the lack of knowledge about the parameters of the model concerning geometry, material behavior, boundary conditions or about the model itself. Different numerical techniques for solving these types of problem have been developed recently.
In this presentation, the efficiency of numerical methods developed in this framework is presented. In particular, we are interested in studying errors due to approximations done. The goal is to develop techniques that improve the computation quality and preserve the computational cost. An optimal choice of the simulation parameters can be done using an error indicator. The objective is to design a tool which can be efficient on industrial structures.
- Professor at INSA Centre Val de Loire (France) since 2013
- Associate professor at ENS de Cachan : 2002-2013
- Ph.D., Mechanical Engineering, ENS de Cachan, 2002
- Postgraduate Degree, ENS de Cachan, 1998