| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing and Alloy Design: Bridging Fundamental Physical Metallurgy, Advanced Characterization Techniques, and Integrated Computational Materials Engineering for Advanced Materials
|
| Presentation Title |
G-32: Design for Additively Manufactured Refractory High Entropy Alloy: Semi-Autonomous Multi-Objective Bayesian Optimization for Improving Performance, Manufacturability, and Sustainability |
| Author(s) |
Md. Shafiqul Islam, Doguhan Sariturk, Raymundo Arroyave |
| On-Site Speaker (Planned) |
Md. Shafiqul Islam |
| Abstract Scope |
Designing refractory high-entropy alloys (RHEAs) poses significant challenges due to their complex compositions and mechanical properties, affecting manufacturability and performance. We introduce a novel computational framework leveraging Multi-Objective Bayesian Optimization (MOBO) and CALPHAD simulations to explore multi-principal component alloys with promising mechanical properties. Our approach addresses manufacturability trade-offs, optimizing hot tearing tendency (crack susceptibility) and printability (based on porosity-induced defects) by constructing printability maps and calculating the probabilistic printability index. Performance indicators such as high-temperature yield strength (via BCC-B2 dual-phase stability, provided by B2 precipitate to emulate gamma-gamma' of Ni-based superalloys), room temperature ductility (using M3GNet and DFT-KKR), and stiffness (using M3GNet) are optimized. A novel decision-maker aligned optimization framework addresses decision makers' preference. Additionally, supply chain risk assessments ensure the sustainability of the alloy discovery process. This methodology streamlines compositional exploration and enhances the efficiency of identifying optimal alloys guided by decision-maker preferences. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Additive Manufacturing, High-Entropy Alloys, Computational Materials Science & Engineering |