Washington State University (WSU) researchers used AI to find a cheaper, more efficient method for 3D printing a high-performance metal alloy, writes Chris Young.

Without the AI model, the researchers would have had to methodically comb through more than 100 million options. According to the scientists behind the study, their findings could allow the alloy to be printed using easily obtainable commercial equipment.

The alloy, called GRCop-42, is composed of three metals: copper, chromium and niobium. Originally developed by NASA, it is used in liquid rocket engine combustion chambers and other similar use-cases because of its high thermal conductivity and strength under extreme heat conditions.

GRCop-42 has many applications, even beyond rocketry. However, its additive manufacturing (AM) can be costly and challenging. The alloy’s copper-rich composition makes it highly reflective and thermally conductive, reducing the efficiency with which laser energy is absorbed during printing. This means GRCop-42 production typically requires laser power in the kilowatt range.

Because the materials required for production are also expensive, a single printing run can cost hundreds of dollars. These factors have, to date, limited experimentation.

For their study, the researchers began with results from 37 unsuccessful configurations from previous research at WSU. Using those results, they developed a method that would estimate the likelihood that a specific untested configuration would produce a successful print. For every promising option it analysed, the AI model would also analyse and improve itself.

Over the course of three months and with a total budget of 40 experiments, the team identified six successful configurations. This allowed them to successfully print the alloy at 500 W for the very first time.

They believe their methodology could be applied to other metal alloys and AM systems, saving costs by allowing researchers to focus their efforts on a tiny fraction of potential test configurations.

“Ninety percent of commercial printers cannot print this metal alloy. So, given that we were able to find these feasible process parameters, it allows us to use those commercial printers and we are essentially democratising the printing of this alloy,” said Jana Doppa, study lead at WSU.