AI-powered approach accelerates design of complex materials for aerospace and batteries

Researchers from MIT, utilizing the Expanse supercomputer at the San Diego Supercomputer Center (SDSC), have developed a new artificial intelligence approach to model complex metal alloys at the atomic scale. This advance, published in Science Advances, is expected to significantly speed up the design process for materials used in jet engines, computer chips, medical implants, and batteries, reducing reliance on costly and time-consuming lab experiments. The AI models describe atomic interactions within alloys and are used in atom-by-atom simulations to predict material behavior, including melting points and thermal expansion.