A calculation of density profiles in tandem mirrors fueled by pellets

It has been proposed that the central cell fuel ions in a tandem mirror reactor such as MARS or FPD be replenished by the use of high velocity pellets of frozen D-T. The fueling current must replace those ions lost by burnup, diffusion (both radial and axial), and by collisional trapping of ions in the end cell regions. Pellet penetration to only a fraction of the plasma radius is desirable to reduce the required pellet velocity. Central to the issue of whether partial penetration will work for tandem mirrors is that of fueling the plasma by inward radial diffusion. Since axial transport competes with radial transport, one must be careful to deposit particles at a radial location which has adequate axial confinement. Another issue is whether inverted density profiles in the central cell will inhibit the removal of trapped ions in the end cells which is accomplished by induced outward radial transport (drift pumping). To address these issues, the authors modified the LLNL radial transport code TMT to model reactor regime plasmas, fueled by pellets. The source profiles arising from pellet fueling are obtained from existing pellet ablation models. Because inward radial diffusion due to inverted profiles must compete with trapping of central cell ions in the transition region for tandem mirrors, pellets must penetrate fairly far into the plasma. In fact, based on radial calculations, a pellet with a velocity of 10 km/sec cannot sustain the central flux tubes; a velocity more like 100 km/sec will be necessary. They also find that the central cell radial diffusion must exceed classical by about a factor of 100