AI and 3D Printing Enable Design of High-Performance Refractory Alloys for Aerospace and Defense

Researchers at Arizona State University and UNSW Sydney are leveraging artificial intelligence (AI) and 3D printing to redesign refractory alloys, metals that maintain strength at extreme heat and stress, for aerospace and defense applications. Traditional refractory alloys, based on elements like tungsten, niobium, and molybdenum, were developed decades ago and optimized for casting or forging rather than additive manufacturing. When subjected to 3D printing, these materials often crack or warp due to rapid melting and solidification. To address this, the team uses reinforcement learning AI to explore thousands of alloy combinations, optimizing for high-temperature strength, oxidation resistance, manufacturability, cost, and weight. Promising alloys are then tested in the lab, with experimental results feeding back into the AI model to improve predictions. The approach enables faster development of alloys suitable for hypersonic engines, nuclear submarines, and other advanced systems, while reducing material waste by producing components additively instead of machining from bulk metal. Challenges remain, including limited experimental data, high cost of refractory powders, and scaling laboratory results to larger parts. The collaboration aims to align AI-designed alloys with real-world defense needs, combining computational design with practical testing to advance both materials science and defense technology.