Scott E. Parker, Yang Chen, and Charlson C. Kim,
University of Colorado, Boulder

Research Objectives

We are currently studying the role of electromagnetic fluctuations in tokamak plasma turbulence; equilibrium and self-generated zonal flows and related techniques for controlling transport; and techniques for long-time meso and macro simulation. Current methods fail at long times, and we are studying an evolving equilibrium approach with an appropriate collision operator which attracts the distribution function back towards the evolving equilibrium distribution.

Computational Approach

Explanation of the apparent anomaly between the two common approaches used to simulate tokamak plasma turbulence. Top: Views of both flux-tube and global computational domains. Bottom: Plot of heat transport versus the derivative of the temperature gradient showing a transition from low-level global-like transport to the larger flux-tube-like transport.

We have developed a drift-fluid-electron gyrokinetic-ion simulation to study electromagnetic turbulence. This is a fully parallel 3D toroidal simulation. We use a 1D domain decomposition in the direction along the magnetic field line. We are also utilizing a domain-cloning technique, in which the grid is replicated on a second set of processors. This is useful when there are more processors then grid cells in the decomposed direction, or for optimal performance on SMP clusters. The drift-fluid electron model uses finite-difference for solving the hyperbolic partial differential equations. The particle-ion part uses particle-in-cell methods. Poisson solvers are used in the direction perpendicular to the magnetic field, and these are done spectrally.

Accomplishments

We have carefully benchmarked themodel in the shearless slab limit to numerical solution of the dispersion relation including full kinetic and finite gyro-radius effects. Recently we have developed a fully toroidal model. We observe finite- stabilization at low ratios of the plasma pressure to the magnetic pressure, then strong destabilization at higher values along with a large increase in transport levels.

We have put a significant effort into identifying the differences between global and flux-tube simulations. We have shown theoretically that the behavior of the purely radial mode can be predicted from the perpendicular flux-surface-averaged ion temperature. With significant profile variation, the heat flux flattens the equilibrium temperature, leading to the generation of the global purely radial mode. On the other hand, when the temperature gradient is constant, there is no preferred location of profile flattening, and the radial mode is then more turbulent, which is observed in constant temperature gradient flux-tube simulations.

Using the knowledge of how the self-generated flows are produced, we did a simple numerical demonstration of a new scheme to reduce the heat transport by slightly rippling the temperature profile. A slight ripple in the equilibrium temperature profile ripples the transport, causing the generation of short-scale zonal flows. This, in turn, reduces the heat transport.

Significance

Publications

S. E. Parker, Y. Chen, and C. C. Kim, "Electromagnetic gyrokinetic simulations using a drift-fluid electron model," Computer Physics Communications (in press).

S. E. Parker, C. C. Kim, and Y. Chen, "Large-scale gyrokinetic turbulence simulations: Effects of profile variation," Physics of Plasmas 6, 1709 (1999).

A. M. Dimits, B. I. Cohen, N. Mattor, W. M. Nevins, D. E. Shumaker, S. E. Parker, and C. Kim, "Simulation of ion-temperature-gradient turbulence in tokamaks," in Proceedings of the 17th IAEA Fusion Energy Conference (paper IAEA-F1-CN-69/TH1/1); Nuclear Fusion (in press).

http://fluid.colorado.edu/