About this Abstract |
Meeting |
2nd World Congress on High Entropy Alloys (HEA 2021)
|
Symposium
|
2nd World Congress on High Entropy Alloys (HEA 2021)
|
Presentation Title |
New Physics-based Features for Machine Learning Oxidation Properties of Refractory Complex Concentrated Alloys |
Author(s) |
Logan G. Ware, Haydn Schroader, Tinuade Daboiku, Emily Cheng, Todd Butler, Andrew Detor, Michael Titus |
On-Site Speaker (Planned) |
Logan G. Ware |
Abstract Scope |
Predicting and improving the high temperature oxidation performance of refractory complex concentrated alloys (RCCAs) remains a barrier to their adoption for ultra-high temperature structural components. Recent work has shown many internal and external oxides form at ultra-high temperatures in these alloys, including multi-component complex oxides. The formation of some complex oxides has been correlated with improved oxidation performance, but the conditions for favorable oxide formation are still unknown. In this presentation we report a high-throughput methodology utilizing ab-initio calculations to predict the thermodynamic driving force for oxide formation as a function of oxygen activity. This method is easily extensible and performs comparably to commercial thermodynamic modeling software at predicting oxide phases in RCCAs. From these predictions, we then extract physics-based features for use in machine learning models to estimate the oxidation properties of each alloy, and find these features to be better equipped to predict oxidation performance than rule-of-mixture features. |
Proceedings Inclusion? |
Undecided |