|About this Abstract
||2019 TMS Annual Meeting & Exhibition
||Computational Approaches for Big Data, Artificial Intelligence and Uncertainty Quantification in Computational Materials Science
||Machine Learning for High-Temperature Alloy Design: High-Quality Data, Scientific Descriptors and Curve Fitting
||Dongwon Shin, Bruce Pint, Govindarajan Muralidharan, Yukinori Yamamoto, Michael Brady, Jiheon Jun, Sangkeun Lee, J. Allen Haynes
|On-Site Speaker (Planned)
We will discuss three critical aspects of applying a machine learning approach to accelerate the design of high-temperature alloys for energy applications: 1) consistently measured, relevant experimental data with known pedigree, 2) in-depth features that can be correlated with target properties, and 3) training of surrogate machine learning models. We will present examples of predicting creep properties of alumina-forming austenitic (AFA) alloys and oxidation behavior of Ni-Cr alloys, using data which have been measured at Oak Ridge National Laboratory over decades.
The research was sponsored by the LDRD Program of Oak Ridge National Laboratory and the Department of Energy, Vehicle Technologies Office, Propulsion Materials Program.
||Planned: Supplemental Proceedings volume