| Abstract Scope |
Potential alloy design spaces are broader than ever. Simultaneously, application requirements are becoming more stringent. As design complexity grows, a seamless integration between experiments, computation, and AI becomes critical. However, experimental testing poses a bottleneck. Conventional, manual alloy testing is time-consuming, expensive, difficult to scale, and associated with long feedback delays. High-throughput, automated experiments can overcome this barrier to efficiently navigate complex thermodynamics and kinetics in multicomponent alloys: A first example on metastable phase and glass formation in thin films illustrates the power of examining complex behavior in the big picture across thousands of data points. The second example on BCC-B2 superalloys integrates rapid computational screening with broad experimental assessment, but also highlights the need for automatic, closed-loop coupling. Building towards this goal, a preview of our autonomous, self-driving laboratory platform “AlloyBot” developed at UW-Madison highlights our new on-demand arc-melting synthesis capability for producing 100+ distinct alloy compositions per week. |