Abstract Details
Abstracts
Author: Gabriele Merlo
Requested Type: Pre-Selected Invited
Submitted: 2018-02-28 10:15:45
Co-authors: J. Dominski, A. Bhattacharjee, C. S. Chang, E. D’Azevedo, S. Ethier, R. Hager, A. Hakim, J. Hittinger, F. Jenko, S. Klasky,3S. Ku, M. Parashar, S. Parker, L. Ricketson, A. Siegel, B. Sturdevant, and E. Suchyta
Contact Info:
The University of Texas at Austin
201 E 24th St
Austin, 78712
Texas
Abstract Text:
Understanding, predicting, and optimizing the performance of ITER and of future fusion power plants is a necessary step towards the success of nuclear fusion. Within the Exascale Computing Project (ECP), the Whole Device Model Application is a project that aims at providing a first-principles-based computational tool that integrates all the crucial elements required to simulate a burning plasma.
The first stage of this project consists in coupling the two existing state-of-the-art gyrokinetic codes GENE and XGC, which are used to simulate turbulence in respectively the core and the edge of a tokamak. (During a second stage, other codes will be coupled to this basic framework.) This task has been accomplished, and the two codes are now coupled, enabling one to carry out simulations of turbulent transport from the magnetic axis to the wall.
The coupling algorithm, based on exchanging moments and electromagnetic fields, allows to couple the core-edge codes in a tight and self-consistent way at the level of the distribution function, regardless of the very different numerical schemes and discretization techniques employed by the two codes (GENE is a continuum code, whereas XGC is a particle-in-cell code).
The coupling scheme initially implemented for XGC-XGC coupled simulations has also been applied to GENE-GENE runs. These simplified code-coupling allowed us to learn on the behavior of the coupled system, including the effect of boundary conditions, avoiding the complications due to mappings between the different meshes. The interpolation scheme used for transferring data back and forth between GENE’s structured grid and the unstructured XGC one will also be introduced and discussed. First results obtained with the coupled GENE-XGC code will be presented.
This research was supported by the Exascale Computing Project (17-SC-20-SC), a collaborative effort of the U.S. Department of Energy Office of Science and the National Nuclear Security Administration.
Comments: