Computational Support for Alternative Confinement Concepts Basic Plasma Science

Abstract

This is the final report for contract DE-FG03-99ER54528, ''Computational Support for Alternative Confinement Concepts''. Progress was made in the following areas of investigation: (1) Extensive studies of the confinement properties of conventional Reversed-field Pinch (RFP) configurations (i.e., without current profile control) were performed in collaboration with the Royal Institute of Technology (KTH) in Stockholm, Sweden. These studies were carried out using the full 3-dimensional, finite-{beta}, resistive MHD model in the DEBS code, including ohmic heating and anisotropic heat conduction, and thus for the first time included the self-consistent effects of the dynamo magnetic fluctuations on the confinement properties of the RFP. By using multi-variant regression analysis of these results, scaling laws for various properties characterizing the conventional RFP were obtained. In particular, it was found that the, for constant ratio of I/N (where I is the current and N = na{sup 2} is the line density), and over a range of Lundquist numbers S that approaches 10{sup 6}, the fluctuations scale as {delta}B/B {approx} S{sup -0.14}, the temperature scales as T {approx} I{sup 0.56}, the poloidal beta scales as {beta}{sub {theta}} {approx} I{sup -0.4}, and the energy confinement time scales as {tau}{sub E} {approx} I{sup 0.34}. The degradation of poloidal beta with current is a result of the weak scaling of the fluctuation level with the Lundquist number, and leads to the unfavorable scaling laws for temperature and energy confinement time. These results compare reasonably well with experimental data, and emphasize the need for external control of the dynamo fluctuations in the RFP. (2) Studies of feedback stabilization of resistive wall modes in the RFP were performed with the DEBS code in collaboration with the CNR/RFX group in Padua, Italy. The ideal growth rates are ''passively'' reduced by the presence of a resistive wall within the radius for perfectly conducting wall stabilization of these modes. In this work we consider cases with up to two resistive walls. Moreover the feedback system is assumed to react to any given Fourier harmonic with an ideal response, in the sense that no spurious harmonics are generated. Successful feedback schemes are shown to be possible. However, a careful choice of the gains, along with the simultaneous feedback on at least 4 or 5 modes, is found to be necessary. (3) Studies of a stable rampdown operating regime for the RFP were performed in collaboration with Los Alamos National Laboratory and the University of Wisconsin. It was found that completely stable mean profiles can be obtained by properly tailoring the decaying time dependence of the toroidal current and magnetic flux. Deviations from optimal decay rates were shown to lead to single helicity (SH) and quasi-single helicity (QSH) states. In all cases the prospects for improved confinement properties were obtained. These results may account for the experimental observation of QSH states when the toroidal current is allowed to decay.