Researchers at the Pennsylvania State University (Penn State) trained a generative adversarial network, the same artificial intelligence technology responsible for deepfakes, to create refractory high-entropy alloys that can maintain their strength in ultra-high temperatures.
The researchers created a training dataset of hundreds of published examples of alloys, and after training, the model was tasked with creating alloy compositions suitable for use in turbine blades.
Said Penn State's Zi-Kui Liu said, "Our preliminary results show that generative models can learn complex relationships in order to generate novelty on demand.
"This is phenomenal. It's really what we are missing in our computational community in materials science in general."
The research shows progress toward the inverse design of alloys. With inverse design, Penn State's Wesley Reinhart explained, "You can ask for a material with defined properties and get 100 or 1,000 compositions that might be suitable in milliseconds."
From Pennsylvania State University News
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found