Energy Department launches new HPC initiative
Connecting state and local government leaders
The new program focuses on materials for use in severe environments, and is part of DOE's larger HPC4 Energy Innovation Initiative.
The National Labs have five of the 10 most powerful supercomputers in the world. Jeff Roberts, the deputy director for Energy and Climate Security at the Lawrence Livermore National Lab, said that U.S. private industry -- which often runs laps around government when it comes to technology -- lags behind the government by about five years when it comes to supercomputing capabilties.
For that reason, Roberts told GCN, “we wanted to create a program that allows industry to try and use high performance computing at national labs and the dedicated teams that know how to use it in an applied way to support industry’s toughest challenges.”
That desire led first to the High Performance Computing for Energy initiative, and then HPC for Manufacturing. The latest iteration is HPC for Materials; the Department of Energy is offering up its computing power and modeling expertise to help industry improve materials used in extreme environments.
This could mean extreme pressure, radiation or temperature; corrosion; chemical environments; vibration; fatigue; or stress states. The DOE will accept research proposals from industry on these subjects; those selected will work with National Lab researchers to leverage the supercomputing power at multiple lab locations.
David Teter, the Los Alamos National Lab's Materials Science and Technology Division leader, said the fossil energy, nuclear energy and transportation sectors are just a few of those that could benefit from materials that can stand up to these tough conditions. Materials in an engine, for example, can limit its maximum running temperature, but engines are more efficient if they can operate at higher temperatures.
There are many things to consider when modeling materials, Roberts explained. If you build something in a new alloy then it won’t always behave in the way you expect.
But the National Labs have a repository of models ready for the task of modeling problems that industry puts forward. “Using high-performance computing, you can optimize the design ahead of time and maybe make a new product that performs better, more quickly,” he said.
There will be an informational workshop next month in Pittsburgh, Pa., where interested industry has can learn more. A formal solicitation for research areas will go out at the beginning of next year.