| About this Abstract |
| Meeting |
MS&T23: Materials Science & Technology
|
| Symposium
|
Leveraging Integrated Computational Materials Engineering for High-fidelity Physics-based and Machine Learning Models
|
| Presentation Title |
New Refractory High Entropy Alloys Discovery by Physics Discovery |
| Author(s) |
Lele Luan, Chen Shen, Scott Oppenheimer, Feng Zhang, Ryan Jacobs, Liping Wang |
| On-Site Speaker (Planned) |
Lele Luan |
| Abstract Scope |
Refractory high-entropy alloys (RHEAs) have attracted extensive interest in aerospace application after its first development a decade ago and various machine learning (ML) algorithms have been applied to accelerate new material discovery. In this work, we proposed the application of a new ML concept, physics discovery, to the modeling and design of RHEAs. Physics discovery distilled a human-readable equation from data. Prior knowledge of material characteristics and mechanism were utilized as modeling constraints to ensure physical interpretability. The physics discovery model was built by taking a new RHEA dataset, 86 samples with different element compositions, published very recently. The discovered human-readable equations not only help to understand the mechanism of high-hardness of RHEAs, but also to predict the new materials (element compositions) with higher hardness. This works can be treated a preliminary attempt to build interpretable ML model for better understanding of RHEAs. |