A successful 3-D printing of the metal alloy. (Credit: Nathaniel W. Zuckschwerdt/WSU)
A Washington State University team printed NASA’s GRCop-42 copper alloy at 500 watts for the first time after an AI model picked 40 process trials from more than 100 million possible configurations and found six that worked.
GRCop-42 is a NASA copper-chromium-niobium alloy used in liquid rocket engine combustion chambers because it conducts heat well and stays strong under extreme heat. Printing it has been expensive (a single run can cost hundreds of dollars) and energy-intensive, and it has typically needed high laser power. Ninety percent of commercial printers cannot print the alloy.
Their model estimated how likely an untested setting was to print, then selected small batches that mixed promising options with uncertain ones that could improve the model. Working with Nathaniel Zuckschwerdt, Susmita Bose, and Amit Bandyopadhyay in the School of Mechanical and Materials Engineering, and Aryan Deshwal at the University of Minnesota, the team printed GRCop-42 on the AI-chosen settings and evaluated the samples.
Over three months and a budget of 40 experiments, they identified six successful configurations at different laser power levels. Jana Doppa, a WSU computer science professor who led the research, said each trial returns only success or failure. “Every time you try, you basically get a binary success or failure signal, and you are trying to minimize the number of tries that you have so that you get to those successful needles very quickly.”
The paper appears in the Proceedings of the AAAI Conference on Artificial Intelligence. The team received the conference’s Innovative Deployed Application Award.
Source: news.wsu.edu
