About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
Machine Learning Domain Knowledge-Based Design of Alloys with High Strength |
| Author(s) |
Yasir Sohail, Jinyu Zhang, Evan Ma |
| On-Site Speaker (Planned) |
Yasir Sohail |
| Abstract Scope |
The pursuit of strong yet ductile alloys has continued for centuries. However, most developed alloys, including high-entropy ones with good ductility rarely achieve 2-GPa yield strength at room temperature. The few that do—mainly ultra-strong steels—often exhibit stress–strain plateaus and serrations due to plastic instabilities like Lüders strains, resulting in only pseudo-uniform elongation. Here, we report a group of carefully engineered multi-principal-element alloys, with a composition of Fe35Ni29Co21Al12Ta3, designed using domain knowledge-informed machine learning, that achieve an unprecedented balance of strength and ductility. One such alloy shows 1.8-GPa yield strength with 25% truly uniform elongation. This synergy arises from extreme microstructural heterogeneity, incorporating high fractions of both coherent L12 nanoprecipitates and incoherent B2 microparticles. The latter, being multicomponent and having low chemical ordering energy, form a deformable phase that accumulates dislocations and supports high strain hardening, thereby extending the uniform elongation. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, High-Entropy Alloys, High-Temperature Materials |