Generative Machine Learning for Microstructure Design: An Atomic- Scale-Informed Approach to Interfacial Engineering

PhD position on using generative AI to model material interfaces. The project focuses on predicting plasticity in interface-rich materials by generating defect density fields (Nye tensors) from atomistic simulation data via generative models like GANs. Aluminum is used as a reference material, but the approach may be extended to complex alloys. The research combines multiscale mechanics, physics-informed ML, and dislocation theory. Supervised by experts in AI for materials science, the PhD student will have access to national HPC resources (GENCI, EXPLOR, CASSIOPÉE, ENACT). The goal is to develop a GAN-based method that bridges atomic and continuum scales, enabling better interface engineering in crystalline materials. This position is funded by the ENACT AI Cluster.

Postuler en ligne