MuORoDe

MuORoDe: Multi-Objective Robust DesignAxes: Simulation & ModelingLeaders: Koji SHIMOYAMA (IFS, TU) & Frédéric GILLOT (LTDS, ECL)Participants: Sébastien BESSET (LTDS, ECL), Kazuki OZAWA (IFS, TU)Double degree PhD student: Pradeep MOHANASUNDARAM (LTDS, ECL & IFS, TU)In this project, we focus on robust shape optimization of systems with friction interfaces. Main objectives are to provide efficient global optimization loops for problems exhibiting uncertainties in the design space. Such loops will have to rely on specific finite element models adapted to structural dynamic issues, as well as metamodels able to deal with the dimensional curse when dealing with parameters uncertainties. The optimization algorithms will be studied too, as problems considered are usually based on large number of parameters (n>10). As a result, it will provide robust designs candidates within a reasonable computational time. The current work includes a double diploma Ph.D. student, P. Mohanasundaram, who began his thesis in November 2017 under the direction of K. Shimoyama, F. Gillot and S. Besset. He has been spending one year at IFS/TU since october 2018, before coming back at ECL for his final Ph.D. year. The subject of the thesis is untitled “Shape optimization of complex structures under vibroacoustic criteria – application to brake squeal”. Specific work on batch Expected Improvment for efficent optimization loop is under development with M2 student Kazuki Ozawa from Pr. Shimoyama's lab, who is currently hosted at ECL under the supervision of F. Gillot and S.Besset.