Alloys are at the heart of human civilization, but developing alloys in the Information Age is much different than it was in the Bronze Age. Trial-by-error smelting has given way to the use of high-performance computing to micromanage alloy development down to individual atoms. At Oak Ridge National Laboratory (ORNL), Markus Eisenbach and his colleagues are tackling the computational challenges of developing those ultra-precise, futuristic alloys.
“The specific position of an atom of one element in relation to the atom of another element can make a huge difference in an alloy’s properties and how it impacts tensile strength and resistance to damage,” said Eisenbach, who is a computational scientist specializing in condensed matter and materials science with ORNL’s Leadership Computing Facility (OLCF).
The research team utilized an electronic structure code (Locally Self-consistent Multiple Scattering, or “LSMS”) created at ORNL to conduct single-atom calculations, aiming to test theories about the phase transitions of certain alloys. “With this research,” Eisenbach explained, “we are trying to determine when the atom placement changes from ordered to disordered, meaning at what temperature these phase transitions happen.”
Analyzing an alloy’s magnetic behavior, crystal structure and individual atomic positions produces enormous amounts of data, so Eisenbach and his colleagues turned to the OLCF’s Cray XK7 Titan supercomputer, a system that leveraged 18,688 AMD Opteron CPUs and 18,688 Nvidia K20 GPUs to deliver nearly 18 Linpack petaflops – before it was decommissioned in August.
Specifically, the LSMS code assigned individual atoms of the simulated alloy to individual nodes within the supercomputer. Thanks to the computing power at play, the research was able to demonstrate that the specific ordering of the atoms was important to understanding how the alloy would perform in the real world.
“This requires quite a lot of computing power—hundreds of thousands of calculations of many different configurations of energies—which wouldn’t have been possible before Titan,” Eisenbach said.
Now, with Titan decommissioned, the researchers are turning to a far more powerful system: the world’s fastest publicly-ranked supercomputer.
“We are looking now at Summit to continue this work, running calculations on more complicated materials that make up certain alloys, and incorporating artificial intelligence techniques to improve the performance of the statistical mechanics simulations,” Eisenbach said. Beyond Summit, they have their eyes set on Frontier (ORNL’s exascale system, expected in 2021), which they anticipate will be able to analyze defects in the microstructures of complex materials.
About the research
The research discussed in this article was published as “First-principles study of order-disorder transitions in multicomponent solid-solution alloys” in Journal of Physics: Condensed Matter. It was written by Markus Eisenbach, Zongrui Pei and Xianglin Li and can be found at this link.
The original release from OLCF discussing this research can be found here.