About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Magnesium Technology 2026
|
| Presentation Title |
When Processing Leads the Way: ML-Powered Optimization of Biodegradable Mg Alloys |
| Author(s) |
Sreenivas Raguraman, Maitreyee Sharma Priyadharshini, Tunde Ayodeji, Andrew Kim, Veronica Ivanovskaya, Camryn Byrum, Adam Griebel, Paulette Clancy, Timothy P Weihs |
| On-Site Speaker (Planned) |
Sreenivas Raguraman |
| Abstract Scope |
Magnesium alloys are promising for orthopedic implants due to their biodegradability, biocompatibility, and mechanical strength, but rapid corrosion limits their clinical adoption. While machine learning is increasingly applied to alloy design, its use in optimizing processing, critical for clinical translation, remains rare. In this work, we processed ZX10 Mg alloy using over 40 distinct thermal, mechanical, and thermomechanical treatments conducted in‑house. All property data were generated through rapid hardness and immersion corrosion testing, providing a robust dataset for a physics‑informed Bayesian Optimization approach (PAL 2.0) that links processing routes to performance. Through two closed‑loop experimental rounds, we increased hardness by ~50% of the maximum observed in the initial dataset, while bringing corrosion rates closer to acceptable limits. These results highlight the promise of combining machine learning and processing pathways, without altering composition, to balance strength and corrosion resistance, thereby accelerating the design of clinically relevant and scalable magnesium implants. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Magnesium, Machine Learning, Process Technology |