Washington State University scientists used AI to find printable settings for GRCop-42 among more than 100 million possibilities, cutting the needed laser power to 500 watts.
Turning 100 Million Options Into 40 Tests
3D printing NASA's GRCop-42 alloy used to need specialized, high-power machines that most universities and small companies cannot afford. A team at Washington State University decided to fix that with AI instead of brute-force testing.
The alloy is a mix of copper, chromium, and niobium. NASA designed it to survive the extreme heat inside rocket combustion chambers while still moving heat away quickly. Those properties make it valuable, but they also make it hard to print. Earlier attempts on lower-power commercial machines often ended in melted puddles.
How the AI Picked the Right Settings
The WSU team started with 37 failed prints and built a model that could guess which untested combinations of laser power, speed, and other parameters might work. Instead of trying every possible setup, the AI suggested small batches of promising candidates. Each batch updated the model, so the next round was smarter.
Over three months, the researchers ran just 40 experiments. From those, six settings produced good prints. One of them worked at only 500 watts of laser power, the first time GRCop-42 has been printed successfully at that level.
Why 500 Watts Matters
Most commercial metal printers top out below the power levels normally used for GRCop-42. The team says about 90 percent of those machines could not print the alloy before. Dropping the power requirement means the alloy could become available to a much wider group of researchers and manufacturers.
The work was published in the Proceedings of the AAAI Conference on Artificial Intelligence and won the conference's Innovative Deployed Application Award. The researchers think the same approach could help with other alloys, and even with problems like drug discovery, where testing every option would cost far too much time and money.
For rocket builders, the near-term payoff is simpler: a proven path to printing a key aerospace alloy on hardware they may already own.
Comments (0)
No comments yet. Be the first!
Leave a Comment