A Monte-Carlo study of ICRF-sustained mode operation in tandem mirrors

Abstract

A study, using a Monte-Carlo simulation code, of ICRF-sustained mode operation in tandem mirrors by way of ICRF end-cell fuelling and heating is described. Although the basic parameter space considered corresponds to the Phaedrus experiment, the central-cell density and temperatures are extended towards the reactor regime. It is found that significant end cell ion potential barriers can be generated with ICRF, but that, owing to choking of the central-cell ion source stream by the plugging potential, saturation occurs and power requirements rapidly increase, so that the potential rise is limited to about twice the central-cell ion temperature. Although performance is improved as the ion cyclotron resonance approaches the end-cell mid-plane, no significant difference is found between inboard, outboard or double resonance location. As the central-cell density and temperatures are increased, the RF power requirement is found to increase dramatically. Optimum performance for end cell fuelling results when the central-cell electron temperature is higher than the ion temperature, but the magnitude of this ratio is limited by an increase in threshold power level with electron temperature.

Ion cyclotron wave effects on the ion velocity distribution in a tandem mirror

“…These two effects, however, are not important near the end cell midplane on axis. To the extent possible, this program has been compared with a similar computer program by Todd [17], and reasonable agreement has been found. )…”

“…The dynamics of an ion in the electric field on the RF wave have been discussed by a number of investigators [15][16][17][18]. The change in velocity as the ion passes through the resonance can be computed from…”

Simulation of ion confinement during ion cyclotron heating in the TMX-U tandem mirror

“…has been discussed in an earlier paper [1]. Here we focus on the ion heating by using a Monte Carlo code where an ensemble of test particles are subjected to collisions and wave heating [7][8][9]. This method has several advantages over a bounce average Fokker-Planck code.…”