About this Abstract |
Meeting |
2023 TMS Annual Meeting & Exhibition
|
Symposium
|
Thermodynamics and Kinetics of Alloys
|
Presentation Title |
Understanding Precipitation and Age Hardening Of FeCrAl Alloy Using Explainable Artificial Intelligence |
Author(s) |
Indranil Roy, Subhrajit Roychowdhury, Sandipp Krishnan Ravi, Bojun Feng, Rajnikant Umretiya, Andrew Hoffman, Raul Rebak |
On-Site Speaker (Planned) |
Indranil Roy |
Abstract Scope |
FeCrAl alloy has been historically proven to be a good oxidation-resistant material at both high and low temperatures and hence a top candidate for nuclear reactor cladding material to replace Zr-based alloys. One of the major challenges of using FeCrAl is the formation of Cr-precipitates that make the alloy brittle after exposure to high temperatures. With the age-hardening data from literature along with experiments performed at GE Research, we demonstrate a systematic data-driven modeling approach to understanding precipitation in FeCrAl alloys. Historically, it was accepted that Aluminum (Al) suppresses precipitation. But with the accumulation of a high volume of data over a larger temperature range, we discuss how Al can incite precipitation at lower temperature ranges while suppressing it at higher temperatures. Along with predictive modeling, explainable AI, which is a novel tool in material science, can help understand the competition of thermodynamics and kinetics of precipitation for FeCrAl alloy. |
Proceedings Inclusion? |
Planned: |
Keywords |
Phase Transformations, Machine Learning, Nuclear Materials |