An artificial intelligence system has cracked a problem that stumped human engineers for years: how to 3D-print a NASA rocket alloy using far less power than previously thought possible. Researchers at Washington State University in the United States used AI to search through more than 100 million potential printing configurations for GRCop-42, a copper-chromium-niobium alloy developed by NASA for extreme heat environments. After just 40 physical experiments, the system identified six successful settings, including one that worked at a record-low 500 watts. That breakthrough could slash the cost and complexity of printing this high-performance material, making it accessible to far more manufacturers.
A Metal That Most Printers Can't Handle
GRCop-42 is prized in aerospace for its ability to withstand intense heat while efficiently transferring thermal energy. It's used in liquid rocket engine combustion chambers, where both strength and heat resistance are critical. But the alloy has a major drawback: printing it typically requires very high laser power, which is only available on specialized, expensive machines. About 90 percent of commercial 3D printers can't handle the material at all. Previous attempts to print GRCop-42 at lower wattages failed, and testing all 100 million possible settings manually was impossible. Each trial print costs hundreds of dollars in material alone, and analyzing a single sample can take days. Some configurations simply melted the product, wasting time and money.
AI Cuts Through the Noise
The WSU team, led by computer science professor Jana Doppa, developed an AI algorithm that could predict which settings were most likely to work. Instead of brute-force testing, the AI learned from each experiment and narrowed down the possibilities. After 40 physical runs, it found six viable configurations, including the 500-watt setting that had never been achieved before. The research, published in the Proceedings of the AAAI Conference on Artificial Intelligence, also earned the Innovative Deployed Application Award at the organization's annual conference. The team believes the same AI approach could be applied to other scientific challenges with enormous search spaces, such as drug discovery.
Opening the Door to Wider Use
For local researchers and the broader manufacturing community, this is more than a lab curiosity. The ability to print GRCop-42 on common commercial printers could democratize access to the alloy, allowing smaller companies and research labs to work with a material previously reserved for well-funded aerospace programs. The AI-driven method also reduces the cost and time required to develop new manufacturing processes. While the team's immediate focus is on aerospace, the implications extend to any industry that could benefit from a metal that stays strong under extreme heat. The discovery shows that sometimes the smartest way to solve a complex problem is to let a machine do the searching.
