News & Announcements
|
October 8, 2012
"MSD and MCS form SciDAC-3 Application Partnership for large-scale computational optimization of superconductors"
Researchers in Argonne’s Materials Science Division and Mathematics and Computer Science Division have received funding from the U.S. Department of Energy to conduct a five-year study titled “Optimizing Superconductor Transport Properties through Large-Scale Simulation.” The study is supported by DOE’s SciDAC-3 Application Partnerships in Materials and Chemical Sciences and is funded jointly by the DOE Office of Basic Energy Sciences (BES) and the Office of Advanced Scientific Computing Research (ASCR).
Most energy applications of superconductivity, such as power distribution over superconducting cables and superconducting turbines to tap offshore wind, are based on achieving low energy dissipation in high-temperature superconductors. Dissipation is minimized by restricting the mobility of the vortices carrying magnetic field in the superconducting material.
Understanding the interaction of these vortices is a major challenge and requires the development of accurate numerical algorithms to capture the dynamics of thousands of vortices accounting for all their interactions with each other and the outside world, explained Andreas Glatz, an assistant physicist in the Materials Science Division and overall project lead responsible for coordinating both the theoretical and the computational efforts.
The task is complicated by the high density of the vortices, their mutual long-range interaction, and the dependence of their behavior on external parameters such as temperature and the applied magnetic field. These features preclude analytical description of vortex dynamics and, until recently, have made numerical simulation prohibitively expensive.
“Earlier formulations of vortex dynamics were restricted to modest-sized systems where the full large-scale behavior did not emerge. Now, however, capitalizing on the new capabilities of DOE’s leadership-class computing hardware, we can develop fully implicit time discretization schemes that can simulate such systems in reasonable wall-clock time,” said Dmitry Karpeev, an assistant computer scientist in the Mathematics and Computer Science Division and coordinator of the overall MCS/ASCR efforts.
The Argonne team will use the time-dependent Ginzburg-Landau formulation to capture the dynamics of large-scale arrays of vortices in superconductors. The MSD researchers will work with MCS Division researchers to help select parameters for the numerical model that are most representative of experimental conditions. The team also will extract from the numerical results observable properties that can be validated experimentally.
The researchers plan to carry out the required large-scale computational work on the Argonne Leadership Computing Facility, developing code and implementing the new algorithms on Argonne’s powerful IBM Blue Gene systems.
In addition, Glatz and Karpeev will work closely with MCS and MSD investigators in three DOE centers. Two SciDAC Institutes, FASTMath and SUPER, will collaborate in developing new scalable computational tools, including block preconditioners, adaptive mesh refinement, and derivative-free optimization. The Center for Emergent Superconductivity, a Basic Energy Sciences Energy Frontier Research Center, will collaborate on theoretical formulation of the large-scale dynamics of vortices and on experimental verification of the project’s predictions.
The aim of the partnership is to advance the fundamental understanding of vortex dynamics in superconductors and to determine the optimal size, shape, and concentration of the admixed particles required for achieving optimal power transmission properties.
Media contact: Gail Pieper, pieper@mcs.anl.gov
