Advancing 304L Stainless Steel Modeling: From DFT to Machine-Learned Interatomic Potentials

Stainless steel alloys are well-known for their versatility, durability, and reliability, making them essential for a wide range of applications. In the low-carbon 304L alloy, radiation damage significantly influences the evolution of martensitic phase transformations—diffusionless phase changes in the material's crystal structure—and the resulting magnetic character of 304L. However, understanding the dynamic nature of martensitic transformations remains challenging due to modeling limitations. Density functional theory (DFT) lacks the time and length scales necessary to reliably and effectively model such systems. Molecular dynamics (MD) codes, such as LAMMPS, are now capable of simulating appropriate length and timescales but lack potentials capable of accurately representing the complex behavior of these alloys. This talk will explore the methods used to reliably calculate the properties of 304L stainless steel alloys using DFT and the generation of high-quality data for training machine-learned interatomic potentials (ML-IAPs), building on recent successes in generating accurate magneto-elastic ML-IAPs for alpha-Fe, a key component of 304L stainless steel.