About this Abstract |
| Meeting |
MS&T26: Materials Science & Technology
|
| Symposium
|
Progress in High Entropy Materials: Integrating Experiments, Computation, and Machine Learning
|
| Presentation Title |
Accelerated Discovery of High Entropy High-Temperature Materials by Data-Driven Methodology |
| Author(s) |
Kun Wang |
| On-Site Speaker (Planned) |
Kun Wang |
| Abstract Scope |
Machine learning has shown strong promise for predicting phase formation and properties in complex materials, including high-entropy systems. However, its application to high-entropy materials for extreme environments, such as nuclear reactors and hypersonic vehicles, remains limited by data scarcity and inconsistent data quality. This presentation will highlight recent work from the presenter’s group on applying machine learning to predict single-phase formation and mechanical properties in high-entropy alloys. It will further discuss an integrated framework that combines high-throughput experiments with machine learning for the design of high-entropy ultra-high-temperature ceramics. Because the experiments are conducted under controlled and consistent conditions, high-throughput methods generate high-quality datasets for model training. Experimental validation is then used both to assess model accuracy and to generate new data, establishing an iterative closed-loop strategy that accelerates materials discovery and optimization. |
