Machine Learning Advances in Mechanics of Materials

Speaker: Miguel A. Bessa, Faculty of Mechanical, Maritime and Materials Engineering at Delft University of Technology Abstract: This event is part of the Hariri Institute's Distinguished Speaker Series: Machine Learning for Model-Rich Problems. In the last decade Machine Learning (ML) has become an essential tool in Science and Engineering. There is however a large rift separating ML developments in Computer Science from applications in Solid Mechanics. Materials have complex microstructures from which uncertain, irreversible, and history-dependent phenomena emerge. This creates difficulties in using ML because the learning process requires high throughput yet physically accurate data. These challenges are addressed by novel data-driven methods that lead to materials and structures with unprecedented properties.