What if a metal-printing search could test fewer configurations while still finding successful conditions?
Washington State University researchers used artificial intelligence to find a more efficient, less expensive way to 3D-print GRCop-42, a high-performance copper, chromium, and niobium alloy developed by NASA.
Why the alloy is difficult to print
GRCop-42 combines high thermal conductivity with strength under extreme heat, properties that support its use in aerospace applications including liquid rocket engine combustion chambers.
Printing GRCop-42 is expensive and energy-intensive because the process requires significant laser power.
Researchers had previously tried to print the alloy at the lower wattages and laser powers used by more common commercial printers without success.
Turning failed trials into a search strategy
A single printing run can cost hundreds of dollars, while detailed post-print quality analysis can take days.
The team began its AI approach with results from 37 previously tested unsuccessful configurations.
Using those results, the AI estimated the likelihood that an untested configuration would produce a successful print.
The problem used binary success-or-failure signals while seeking to minimize the number of trials.
Balancing likely wins and unknown territory
Rather than selecting only the configurations that appeared most promising, the model chose small batches that balanced those options with uncertain areas that could improve the model.
The researchers then used the AI-selected process configurations to print GRCop-42 and evaluated the resulting samples.
According to the study’s first author, every experimental result, including failures, improved the AI model.
Some trial configurations produced melted parts that were not printable.
What the experiment found
Within the three-month, 40-experiment budget, the team identified six successful configurations across different laser power levels.
The team also successfully printed GRCop-42 at 500 watts for the first time.
The researchers knew that only a very small fraction of more than 100 million possible configurations would produce successful GRCop-42 prints.
Potential value and remaining limits
The researchers say that printing GRCop-42 at lower laser power could reduce energy use, equipment wear, and post-processing costs.
They also say the approach could broaden access for groups without specialized high-power equipment.
The researchers believe the AI-guided framework could be adapted for other metal alloys and additive-manufacturing systems.
Sources:
Reader links:
Top comments (0)