Major Computational Applications in Magnetic Fusion
B. I. Cohen, M. J. Caturla, T. Diaz de la Rubia, A. M. Dimits, A. Koniges, W. M. Nevins, G. D. Porter, M. E. Rensink, T. D. Rognlien, D. E. Shumaker, and X. Q. Xu, Lawrence Livermore National Laboratory
Research Objectives
The major simulation codes in Lawrence Livermore National Laboratory's Magnetic Fusion Program address the physics of fusion plasmas and materials in three areas: (1) anomalous energy and momentum transport in the tokamak core, using gyrokinetic particle-based simulation; (2) the dynamics of energy displacement cascades in vanadium, vanadium alloys, and SiC ceramics, using the molecular dynamics code MOLDYCASK; (3) tokamak edge-region plasmas, modeled by a series of codes using domain-decomposition algorithms on parallel computers. These codes are used to design tokamak divertors that have acceptable heat fluxes on material surfaces and allow pumping of helium ash.
Computational Approach
The principal computer codes reviewed here are run on massively parallel computers, mostly on the NERSC T3E, and make use of message passing and domain decomposition in most cases. The computational physics algorithms employ state-of-the-art methods in the kinetic simulation of plasmas with particle codes, molecular dynamics simulations, and hydrodynamic simulation.
Accomplishments
Gyrokinetic (GK) turbulent transport simulation: Detailed parameter studies have been conducted with our GK simulations, addressing discharge #81499 in the General Atomics DIII-D tokamak as a base case because of its relevance to the International Thermonuclear Experimental Reactor (ITER). Careful comparisons have been made between the results of the fluid and kinetic simulations of shot #81499 and variants to determine parametric dependences and points of agreement between the simulation algorithms (as part of the Cyclone Project, the purpose of which was to study the physics basis and reliability of the various transport models used for ITER projections).
In the Cyclone Project, a GK flux-tube particle-in-cell (PIC) simulation has been the object of a vigorous study, focusing on its convergence properties with respect to particle number, spatial resolution, and system size. Results indicate strong convergent properties. The GK simulation results concerning turbulent ion thermal diffusivity for the ITER-relevant DIII-D discharge are in the range of a factor of 2 to 3 lower than GLF results, with larger differences closer to marginal stability.
The gyrokinetic simulations have also demonstrated the importance of flow shear in reducing drift-wave turbulence in tokamaks, as observed in experiments. In particular, a new phenomenon in which zero-transport states occur for values of the temperature gradient significantly greater than the linear marginal value has been found.
Neutron interactive materials: The MOLDYCASK code was converted from PVM to MPI in FY98 and has been used to investigate the primary damage state from recoil cascades in vanadium and SiC. The simulations provide the database for defect production in these materials for input into kinetic Monte Carlo simulations of damage accumulation and microstructure evolution over macroscopic length and time scales.
Edge plasma simulation: For the UEDGE transport code, we have generalized the basic domain-decomposition algorithm used in our parallel version to include multiply-connected regions of the edge plasma. This allows simulation of realistic toroidal geometry for devices.
Two-dimensional hydrodynamic simulations with the UEDGE transport code accurately predict the electron temperature in detached plasmas near the divertor plate in the General Atomics DIII-D tokamak experiment. False-color plots of the electron temperature as functions of vertical height Z and major radius position R are shown for data from the UEDGE simulation and the experiment for shot number 87506.
An algorithm is also included for calculating the electric fields that are believed important for suppressing turbulence and that contribute significantly to in/out divertor asymmetries.
Within the last year, we have developed and parallelized the first 3D global-turbulence simulation code BOUT in a realistic toroidal x-point geometry spanning the region including the separatrix.
Significance
The simulation and modeling research described above has been successful in developing a deeper understanding of the physics, analyzing experiments, aiding efforts to improve experimental performance, and designing new experiments.
Publications
R. S. Averback and T. Diaz de la Rubia, "Displacement damage in irradiated metals and semiconductors," Solid State Physics 51, 281-401 (1998).
T. D. Rognlien and D. D. Ryutov, "Analysis of classical transport equations for the tokamak edge plasma," Contr. Plasma Phys. 38, 152 (1998).
X. Q. Xu and R. H. Cohen, "Scrape-off layer turbulence theory and simulations," Contributions to Plasma Physics 38 (1998).