About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing of Refractory Metallic Materials
|
| Presentation Title |
Design Towards Next-Generation Refractory Alloys via A Multi-Modal, Multi-Field Generative AI Framework |
| Author(s) |
Bo Ni, Daniel Daniel Sinclair, Sauda Namiiro, Kareem Abdelmaqsoud, John Kitchin, Bryan Webler, Ana Inés Torres, S. Mohadeseh Taheri-Mousavi |
| On-Site Speaker (Planned) |
Bo Ni |
| Abstract Scope |
Additively manufactured (AM) refractory alloys hold unique potential for structural applications at extreme conditions. However, limited mechanical integrity due to their room-temperature brittleness renders their processing and use challenging. In this presentation, we summarize our recent progress in designing next-generation W-based refractory alloys with high strength at high temperatures and sufficient ductility at room temperature by exploring the effects of alloy composition optimization, oxide dispersion, and potential rare-earth element alloying. For a full-cycle design, we combine interdisciplinary efforts including ore-to-element extraction costs and sustainability metrics, composition design via multiscale modeling, and manufacturing optimization with AM. We highlight the development of generative AI models, including alloy-specific language models (AlloyGPT) and multi-agent frameworks, as the “super glue” to integrate the whole design workflow. Our explorations and results may lay the foundation for in-depth understanding and efficient design of novel additively manufacturable refractory alloys for extreme applications. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, High-Temperature Materials, Additive Manufacturing |