|About this Abstract
||2019 TMS Annual Meeting & Exhibition
||Computational Approaches for Big Data, Artificial Intelligence and Uncertainty Quantification in Computational Materials Science
||Max Phase Thermo-mechanical Approximation via Machine Learning
||Daniel Sauceda, Raymundo Arroyave
|On-Site Speaker (Planned)
In previous work by colleagues, thermo-mechanical and functional properties of MAX phases (i.e structural, electronic, and mechanical) were modeled using were modeled using special quasirandom structures (SQS) and calculated using Density Functional Theory (DFT) via the Vienna Ab initio Simulation Package (VASP). In other work conducted by The Kolpak Group, a computation package, PROPhet, utilizes neural networks to map arbitrary relationships between set(s) of material properties and other material properties by training on known VASP calculations. This presentation investigates the utility of PROPhet and other machine learning methods for interpolating, approximating, and scaling MAX phase unit cell mechanical properties via analytical methods rather than first-principles calculations. Thus lowering computational costs and increasing search space possibilities.
||Planned: Supplemental Proceedings volume