Data-driven Design of New Materials and Structures

Prof. Miguel A. Bessa form the TU Delft, The Netherlands, will hold a guest lecture in the framework of SFB 837 about 'Data-driven Design of New Materials and Structures'.

A data-driven computational framework combining Bayesian machine learning for imperfection sensitive quantities of interest, uncertainty quantification and multi-objective optimization is developed to analyze and design new materials and structures. This talk intends to demonstrate the generality of the proposed framework, highlighting key challenges and possible solutions illustrated by three different design problems: toughening composite materials by tuning the plastic behavior of the constituents, improving collapse behavior of ultra-thin satellite structures with uncertain ultimate buckling strength, and finding unprecedented properties by exploring a new material concept.

SFB837_Gastvortrag_Bessa.pdf