About this Abstract |
Meeting |
2025 TMS Annual Meeting & Exhibition
|
Symposium
|
Advanced Biomaterials for Biomedical Implants
|
Presentation Title |
Machine Learning-Driven Magnesium Alloy Design for Biomedical Implants Through Process Optimization |
Author(s) |
Sreenivas Raguraman, Maitreyee Sharma Priyadharshini, Adam Griebel, Paulette Clancy, Timothy Weihs |
On-Site Speaker (Planned) |
Sreenivas Raguraman |
Abstract Scope |
Magnesium alloys present a compelling alternative to conventional orthopedic implant materials due to their biodegradability, biocompatibility, and impressive mechanical characteristics. However, their rapid deterioration challenges practical applications, by compromising mechanical integrity. This study develops a predictive numerical model to identify optimal processing routes for strong and corrosion-resistant Mg alloys. We processed ZX10, a dilute Mg-Zn-Ca-Mn alloy, through over 40 combinations of thermal, mechanical, and thermomechanical techniques, and rapidly quantified the properties via hardness and immersion experiments. This data is integrated into a machine learning framework using Bayesian Optimization to link processing routes to desired properties. Employing a closed-loop paradigm, we experimentally validate our models and refine them based on experimental data, recommending optimal processing routes. This iterative approach achieves the best combination of high hardness and low corrosion rates, fine-tuning the model to accurately predict corrosion and hardness as a function of processing. |
Proceedings Inclusion? |
Planned: |
Keywords |
Biomaterials, Magnesium, Machine Learning |