About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Computational Thermodynamics and Kinetics
|
| Presentation Title |
Data-Driven Design of Next-Generation Superalloys: Combining Thermodynamic Modeling and Machine Learning for Optimized Mechanical Performance |
| Author(s) |
Mohammad Younes Araghi, Shuozhi Xu |
| On-Site Speaker (Planned) |
Mohammad Younes Araghi |
| Abstract Scope |
Recent progress in computational materials science has accelerated alloy development by combining thermodynamic modeling with data-driven methods. In this study, we use Thermo-Calc to analyze phase formation and quantify yield strength contributions from mechanisms such as precipitation, solid solution, and grain boundaries in a variety of Ni-, Co-, and Fe-based superalloys. We incorporate alloy descriptors like mixing enthalpy, entropy, and the omega parameter to capture complex compositional effects, using these as inputs for machine learning models that predict yield strength and optimize alloy selection. Feature importance and correlation analyses highlight key relationships between alloy chemistry and strengthening. This integrated workflow enables the identification of optimal composition-property ranges and supports the inverse design of superalloys with enhanced mechanical properties. The presented methodology demonstrates an effective framework for accelerating the discovery and development of next-generation, high-performance materials. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Computational Materials Science & Engineering, High-Temperature Materials, Machine Learning |