D. A. Spong, D. B. Batchelor, L. A. Berry, B. A. Carreras, M. D. Carter, E. F. Jaeger, S. P. Hirshman, J.-N. Leboeuf, V. E. Lynch, J. F. Lyon, and J. C. Whitson,
Oak Ridge National Laboratory
D. E. Newman, University of Alaska L. Garcia and R. Sanchez,
Universidad Carlos III de Madrid, Spain
A. Ware, University of Montana

Research Objectives

The ORNL Fusion Theory Group is pursuing computational research in three areas that encompass understanding plasma behavior in existing devices and the design of future experiments. These topics are: stellarator optimization and physics, toroidal plasma turbulence and its effects on transport, and radio frequency (RF) heating of plasmas.

Computational Approach

Stellarator optimizations are carried out using using a steepest-descent method to minimize a variational form for the 3D plasma equilibrium. The plasma optimization is then carried out with a Levenberg-Marquardt algorithm. Plasma turbulence models evolve coupled sets of partial differential equations for the ion density, parallel velocity, and temperature in time in the presence of a noise source (to simulate heating). Finite differences in radius and Fourier expansions in the toroidal and poloidal angles are used. The time stepping scheme is time-implicit for the linear terms and time-explicit for the nonlinear terms. Particle models are used both in the stellarator transport physics studies and the self-organized criticality sandpile calculations.

Accomplishments

	Outer magnetic flux isosurface and filamentary magnet coils for a compact three field period transport-optimized stellarator. Color contours indicate the magnetic field strength: red = high field, magenta = low field.

We have developed compact stellarator configurations that provide improved plasma confinement and stability over previous approaches. These efforts are part of the National Compact Stellarator Experiment (NCSX) project and are expected to lead to the construction of proof-of-principle (POP) and concept exploration (CE) devices during the next few years. The POP device will be based on the quasi-axisymmetric optimization technique, while the CE device will be based on the quasi-omnigenous (QO) approach. Our optimization techniques and stellarator analysis codes have helped translate both of these new optimization strategies into realizable experimental designs. These developments have opened up a new niche for the U.S. within the world stellarator program. Successful completion of these designs could result in a $40-50 million investment by DOE in new experimental facilities that have been designed predominantly through the application of NERSC's high performance computing resources.

Self-organized criticality sandpile models are used to study the nonlinear dynamics of plasma instabilities. These have now been run in parallel using enough particles and for long enough times to collect large statistical samples. This is leading to an improved understanding of L-H transition dynamics in tokamak experiments, control of internal transport barriers in reversed shear discharges, evaluation of superdiffusive transport regimes, and better analysis of the long time correlations in plasma edge turbulence. In addition, Landau fluid calculations of ion temperature gradient-driven turbulence have been incorporated into a simple gyrofluid model that evolves equations in time for the ion density or vorticity, the parallel ion velocity, and the ion temperature.

RF calculations have been performed in support of plasma heating efforts on the NSTX device at Princeton Plasma Physics Lab. Mechanisms have been identified and analyzed by which RF can drive wave-induced plasma flows.

Significance

Publications

L. A. Berry, E. F. Jaeger, and D. B. Batchelor, "Wave-induced momentum transport and flow drive in tokamak plasmas," Phys. Rev. Lett. 82, 1871 (1999).

S. P. Hirshman, D. A. Spong, J. C. Whitson, et al., "Physics of compact stellarators," Phys. Plasmas 6, 1858 (1999).

L. Garcia, B. A. Carreras, and V. E. Lynch, "Spatio-temporal structure of resistive pressure-gradient-driven turbulence," Phys. Plasmas 6, 107 (1999).