Séminaire LGF – Anirudh Raju Natarajan – 03 octobre 2025

Titre

Elucidating the properties of multicomponent alloys with first-principles statistical mechanics

Résumé

Modern engineering applications rely on complex multicomponent materials, requiring the careful tuning of chemistry and processing to achieve desired properties. However, this chemical complexity and the large processing parameter space present a significant challenge to rational material design. Conventional computational screening techniques are often impractical, largely because accurate atomistic models require large training datasets and complex model architectures. The lack of predictive models with sufficient accuracy significantly limits the development of new alloys and processing techniques. At the atomistic scale, computational models must capture a range of structural and chemical disorder. Atomistic models can then be coupled with statistical mechanics techniques to derive thermodynamic, kinetic and chemo-mechanical descriptions of materials. These quantities serve as essential ingredients to meso and continuum scale theories that describe phase transformations and microstructure evolution. The continuum models ultimately enable the careful tuning of alloy chemistry and processing techniques. In this talk, I will highlight recent theoretical advances that enable the rigorous coarse-graining of electronic structure calculations through statistical mechanics techniques. I will demonstrate the utility of these techniques in gaining insights into lightweight metallic alloys and multicomponent refractory alloys.