ML-driven Models for Material Microstructure and Mechanical Behavior - DDPS

Explore machine learning-driven approaches for modeling material microstructures and mechanical behavior in this comprehensive webinar. Delve into two key challenges: digitally generating representative microstructure ensembles and leveraging limited physics-based analyses to train ML models capturing key localizations. Learn about transfer learning techniques for rapid 3D microstructure generation and the acceleration of analyses using ML models connecting microstructural images to properties and local stress/strain/damage contours. Discover how these accelerated representations support materials design optimization and real-time decision-making in high-throughput experimentation. Gain insights from Professor Lori Graham-Brady's expertise in computational mechanics, uncertainty quantification, and multiscale modeling of materials with random microstructure.