Technologies for alternative energy are humming right along, and the power grid gets greener every year. But fossil fuels are still the primary players in the global energy game and will be for decades to come. So important new technologies for those power resources are humming along too.
Traditional coal-burning power plants were built with the best technology of the time—and with a certain lifespan in mind. The predicted lifespans of these facilities were based on continual, nearly steady-state operation to meet all of the nation’s energy needs. Now, energy from alternative sources like the sun or wind is meeting some of those needs, so traditional facilities have to adapt—ramping production down temporarily—to make room in the power grid for the alternatively sourced energy. This ramping up and down strains the system, and can accelerate aging, thus invalidating lifespan predictions. Furthermore, modern methods that improve energy-extraction efficiency create more extreme operating conditions—higher temperatures and pressures—than these facilities were built for.
The industry needs new lifespan predictions for extant facilities, and it needs new facilities built with new materials that have been expressly designed for high-temperature and high-stress environments.
A national effort funded by the U.S. Department of Energy (DOE) Fossil Energy program called eXtremeMAT (XMAT) is addressing both of these challenges by harnessing the expertise and capabilities within the national laboratory complex. A large part of XMAT is computation—modeling and simulating a material’s microstructure to predict how it will change with age—and Los Alamos is a key contributor on that front.
The discovery of new materials has historically been slow because it is Edisonian, that is, pure trial and error. Los Alamos scientists are using state-of-the-art computational methods, like physics-based machine learning, to considerably reduce the iteration and accelerate the timeline of new material development.
Materials scientist Laurent Capolungo leads the effort at Los Alamos. He explains, “We are developing computational tools and methods to assess the evolution of microstructures in ferritic and austenitic steels—materials that will be subjected to extreme conditions.” (Ferritic and austenitic are two kinds of steel alloy that have different physical properties.)
“There are two advances we’ve made so far,” Capolungo continues. “First, we’ve developed models that are sensitive to the material’s microstructure and its thermomechanical processing history to predict the material’s rupture life. Second, we’ve developed very sophisticated surrogate-material models for engineering applications that are highly accurate but also simple enough to be user friendly.”
Both kinds of models—the lifetime predictors and the material evaluators—will provide a science-based understanding of the relationship between a material’s microstructure and its performance. This information will be used by collaborating XMAT scientists to physically produce and test candidate new materials.
One reason why Los Alamos is so well positioned to take on much of the computation for XMAT is that some of the leading physics-based tools and models for this type of work were developed at the Laboratory—specifically, the Visco-Plastic Self-Consistent (VPSC) computer code, developed by Capolungo’s predecessor and mentor Laboratory fellow Carlos Tomé and Los Alamos scientist Ricardo Lebensohn, and the Elasto-Visco-Plastic Fast Fourier Transform (EVP-FFT) code, developed by Lebensohn. These codes deliver physics-based predictions of a material’s lifespan, based on microstructural changes in the material. Although they are open-source and used by materials scientists around the world, because they were invented at Los Alamos, the Laboratory maintains the highest concentration of expertise in their implementation.
The VPSC and EVP-FFT computer codes were first developed more than 10 years ago with support from DOE’s Basic Energy Science and Nuclear Energy programs, as well as the Advanced Simulation and Computing program and the Joint Munitions program. Those investments have continually paid off, even more so now with the highly visible and collaborative XMAT project.
“I think this is a good example of how things at Los Alamos work,” says Capolungo. “We are leveraging the institutional knowledge that exists here—the expertise and experience acquired from previous work on DOE programs—to make improvements in the fossil energy arena, and quickly.”
The goal of XMAT—to use data science to reduce the time and cost for alloy development and lifetime prediction—is ambitious. But the tools needed to meet this goal—data mining, machine learning, multiscale modeling—are Los Alamos specialties. By bringing together experts from across the national laboratory complex, XMAT is improving fossil-energy power systems so that they, and the country, can keep humming right along.