Researchers from the Massachusetts Institute of Technology (MIT) are using machine learning to explore the concept of short-range order (SRO) in metallic alloys at atomic levels. The team believes that understanding SRO is key to creating high-performance alloys with unique properties but this has been a challenging area to explore. High-entropy alloys are of particular interest due to being a composition of equal elements, offering vast exploration options. Undertaking their research with Professors Rodrigo Freitas and Tess Smidt, graduate students Killian Sheriff and Yifan Cao are using machine learning to quantify the complex chemical arrangements of SRO. Their research, which is gaining attention for its potential application in various industries like aerospace, electronics, and biomedicine, has been published in the Proceedings of the National Academy of Sciences. Despite the limitations of traditional methods for understanding SRO, machine learning is playing a crucial role in bridging the gap. The team’s methods involve using 3D Euclidean neural networks and using machine learning to numerically evaluate chemical motifs. Currently, the research team is set to explore how SRO can be altered under standard metal processing circumstances through the US Department of Energy’s INCITE program.
