AI Finds Six Ways to Print Rocket-Grade Alloy on Commercial 3D Printers

A problem of scale

GRCop-42 is NASA's favorite high-heat copper alloy. It handles the inside of rocket engines where temperatures approach 6,000 degrees Fahrenheit. The problem: printing it is brutal. The material reflects laser energy rather than absorbing it, so traditional setups need very high power. Most commercial 3D printers top out around 500 to 700 watts. GRCop-42 usually wants more than that, and testing every possible combination of laser speed, power, and feed rate is impractical. There are over 100 million configurations. A single failed print can cost hundreds of dollars and take days to evaluate.

AI does the legwork

Washington State University researchers built an AI model that learns from failed attempts and predicts which untested settings are worth trying. They started with 37 already-failed configurations and let the algorithm propose small batches of new ones. Over three months and 40 total experiments, the system found six working parameter sets. The standout result: the first successful GRCop-42 print at just 500 watts. That is a record low and well within the range of standard commercial machines.

Democratizing a difficult material

The lower power requirement matters for more than just cost. It means universities, small labs, and emerging aerospace companies do not need specialized high-power lasers to work with this alloy. Energy use drops, equipment lasts longer, and post-processing costs shrink. The same AI approach could apply to other tricky materials and even fields like drug discovery, where experiments are expensive and outcomes are sparse.

Not ready for flight yet

This is a process development win, not a qualified aerospace part. The study does not certify GRCop-42 components for service in rocket engines. Future work will test the method on other alloys and compositions. But the proof that AI can compress months of trial and error into weeks is the real story here.