| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Environmental Degradation of Multiple Principal Component Materials
|
| Presentation Title |
Predicting Alloy Oxidation Resistance Using Physics Informed Machine Learning |
| Author(s) |
Richard P. Oleksak, William Trehern, Aditya Sundar, Leebyn Chong, Madison Wenzlick, Kyle A. Rozman, Martin Detrois, Paul D. Jablonski, Michael C. Gao |
| On-Site Speaker (Planned) |
Richard P. Oleksak |
| Abstract Scope |
Alloy oxidation resistance and, more specifically, the ability of an alloy to form and maintain a protective oxide layer, is a complex question. In particular, various thermodynamic and kinetic properties of the alloy with oxygen play a key role. Here we show that machine learning methods which incorporate these physically meaningful properties offer a powerful approach to predict alloy oxidation performance. Thoughtful data collection strategies are coupled with high throughput CALPHAD calculations to develop machine learning models capable of accurately predicting alloy mass change behavior during high temperature oxidation experiments. The models succeed in capturing a diversity of behavior ranging from rapid oxidation and severe spallation, to slow parabolic kinetics representative of stable and protective oxide layers. Model interpretation sheds light onto the key factors affecting alloy oxidation, while prospects for using the developed approaches to design new oxidation resistant alloys with co-optimized properties are described. |
| Proceedings Inclusion? |
Planned: |