Self-driving lab leverages AI to develop tough new 3D-printable metal alloys for aerospace and advanced manufacturing

Researchers at U of T Engineering have utilized an AI-driven "self-driving lab" to discover a new class of metal alloys specifically designed for additive manufacturing in extreme environments. By combining machine learning with robot-assisted manufacturing, the team identified materials that maintain strength and oxidation resistance at temperatures exceeding 600 degrees Celsius. This breakthrough addresses the critical need for high-performance, 3D-printable materials in the aerospace and power generation sectors, where traditional alloys often fail under intense pressure and heat.

Led by Professor Yu Zou, the research team at the University of Toronto’s Department of Materials Science & Engineering (MSE) developed an "active learning" technique to navigate the vast design space of potential metal alloys. This closed-loop discovery platform integrates computer modeling and automated manufacturing to identify compositionally complex alloys that can be printed layer by layer. The project, supported by U of T’s Acceleration Consortium, aims to create components with variable compositions—such as a hard exterior and a lightweight interior—that are impossible to produce through traditional manufacturing methods.

The self-driving lab successfully identified six new alloys within a few weeks by focusing on a three-element system of nickel, cobalt, and chromium. PhD student Ajay Talbot, the lead author of the study published in npj Advanced Manufacturing, explained that the model uses "data-lean" strategies to strategically select and test samples, feeding results back into the AI to refine future searches. One specific alloy, composed of 12% nickel, 62% cobalt, and 26% chrome, demonstrated superior hardness at 600°C, outperforming the industry-standard Inconel 625 by 4.5% despite having a much simpler chemical makeup.

Another significant discovery was an alloy consisting of 36% nickel, 14% cobalt, and 50% chrome, designed for the even more extreme temperatures found in the rear sections of jet engines. This material exhibited an 85% improvement in oxidation resistance over Inconel 625 at 1000°C, preventing the "oxide scale" formation that typically degrades materials in high-heat environments. Moving forward, the researchers intend to increase the complexity of their discovery process to include up to 12 different elements, targeting even higher operating temperatures of 1,200°C for next-generation aerospace and nuclear power applications.