Alloys like steel, bronze and copper have become fundamental materials for many aspects of human life and industry. But the needs of the modern world are ever-evolving, and at the Texas Advanced Computing Center (TACC), researchers are leveraging supercomputing and deep learning to investigate new, higher-tech alloys: “high-entropy alloys.”
Traditional alloys, like steel, contain just one “principal element” (in steel’s case, iron, which constitutes 98 to 99 percent of the alloy) that is supplemented by much smaller amounts of other elements (in steel’s case, carbon). “High-entropy alloys represent a totally different design concept. In this case we try to mix multiple principal elements together,” explained Wei Chen, senior author of the study and associate professor of materials science and engineering at the Illinois Institute of Technology, in an interview with TACC’s Jorge Salazar.
Chen’s team first ran high-throughput quantum mechanical calculations of thousands of different combinations of 14 elements to assess the stability and elasticity of those combinations. (“This is to our knowledge the largest database of the elastic properties of high-entropy alloys,” Chen said.) Then, they applied a deep learning model to that database to predict the properties of “more than 370,000” new high-entropy alloy compositions.
“We derived some design rules for high-entropy alloy development. And we proposed several compositions that experimentalists can try to synthesize and make,” Chen said, referring to the deep learning process known as association rule mining. That initial database made the deep learning process possible. “That’s why we perform the high-throughput calculations, in order to survey a very large number of high-entropy alloy spaces and understand their stability and elastic properties,” Chen said.
Of course, for an intensive process like this, correspondingly intensive compute resources are necessary. For that, the researchers turned to TACC’s Stampede2, a Dell-built supercomputer with Intel CPUs that rates at 10.68 Linpack petaflops and which ranked 47th on the most recent Top500 list. The system was allocated to Chen’s team by the Extreme Science and Engineering Discovery Environment (XSEDE) program. (Stampede2 is pictured in the header image.)
“The sheer number of calculations are basically not possible to perform on individual computer clusters or personal computers,” Chen said. “That’s why we need access to high-performance computing facilities, like those at TACC allocated by XSEDE.”
To efficiently run the team’s code on a supercomputer, Chen used a TACC-provided code called Launcher, which Chen said “helped [them] pack individual small jobs into one or two large jobs, so that we [could] take full advantage of Stampede2’s high performance computing nodes.”
“Hopefully more researchers will utilize computational tools to help them narrow down the materials that they want to synthesize,” Chen added. “High-entropy alloys can be made from easily sourced elements and, hopefully, we can replace the precious metals or elements such as platinum or cobalt that have supply chain issues. These are actually strategic and sustainable materials for the future.”