In his doctoral dissertation, MSc (Tech) Mika Sarvilahti demonstrated how modern machine learning and optimization methods can be applied to material design in the field of computational materials physics. Convolutional neural networks proved to be effective for predicting the plastic response of crystalline materials, whereas the Bayesian optimization method was demonstrated to perform well in micro-structural optimization problems when optimization constraints are handled appropriately. The study also validated a new, versatile precipitate model that both improves the reliability of dislocation dynamics simulations and makes it possible to study more advanced optimization problems in the future. The findings pave the way for future materials research where intelligent methods are employed to improve material design processes, potentially resulting in the discovery of novel materials with tailor-made mechanical properties that outperform those of currently known materials.
The doctoral dissertation of MSc (Tech) Mika Sarvilahti in the field of Physics titled Computational Studies of Crystal Plasticity and Intelligent Material Design will be publicly examined at the Faculty of Engineering and Natural Sciences at Tampere University.
The Opponent will be Docent, University Researcher Fredric Granberg from the University of Helsinki. The Custos will be Professor Lasse Laurson from the Faculty of Engineering and Natural Sciences, Tampere University.