SKM 2023 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

MM: Fachverband Metall- und Materialphysik

MM 39: Phase Transformations: Simulation and Machine Learning

MM 39.3: Vortrag

Donnerstag, 30. März 2023, 12:15–12:30, SCH A 215

Ab initio thermodynamics and atomistic modeling of NiTi SMA with machine learning interatomic potentials — •Prashanth Srinivasan and Blazej Grabowski — Institute for Materials Science, University of Stuttgart, Germany

Equiatomic Nickel-Titanium (NiTi) possesses interesting properties such as pseudoelasticity and shape memory effect that arise from a reversible transformation between the austenite (B2) and the martensite phase (B19′). Other competing phases (B19 and B33) make modeling of NiTi challenging. Ab initio molecular dynamics (AIMD) calculations have shown that B19′ and B2 phases are entropically stabilized (Haskins et. al., 2016), but the calculations were restricted to a single DFT exchange correlational functional (XC). Studying the kinetics of phase transformation using only DFT is also severely expensive.

In this work, we address both challenges. Using a recently developed thermodynamic integration technique (Jung et. al., 2022) aided with machine-learning based moment tensor potentials (MTPs, Shapeev 2016), it is possible to efficiently compute high-temperature thermodynamic phase stability to DFT accuracy. We perform such calculations to analyze three different XCs (GGA, LDA and SCAN). We also perform large-scale molecular dynamics (MD) simulations using the MTPs to study the kinetically-driven phase transformation behavior in each of these cases. Preliminary results show the necessity of having a high k-point density in the underlying DFT calculations. When fitted to such a training set, the MTPs predict highly accurate thermodynamic properties and phase transformation behavior.