A simplified lithium vapor box divertor

Abstract

A detached divertor plasma is predicted to be necessary to mitigate the heat fluxes and ion energy, and so sputtering, at the divertor plate of a fusion power plant. It has proven difficult in current experiments, however, to prevent the detachment front, once formed, from running up to the main plasma resulting in deterioration of pedestal performance. The lithium vapor box divertor is designed to localize a dense cloud of lithium vapor away from the main plasma, in order to induce detachment at a stable, distant location. The vapor localization is created by local evaporation and nearby condensation, a configuration inaccessible to non-condensing seed impurities. The UEDGE code has been used to show that stable detachment can be achieved in the Fusion Nuclear Science Facility with lithium vapor alone, giving a widely dispersed heat flux of ~ 2 MW/m2 to the divertor walls. However, the model for lithium vapor transport in UEDGE is purely diffusive, determined by collisions of Li atoms with fixed plasma ions. The paper provides simulations of lithium vapor flow using the SPARTA Direct Simulation Monte Carlo (DSMC) code, in which the lithium vapor dynamics are governed by Li-Li collisions that should dominate in regions with little plasma. The plasma model implemented in SPARTA is based on the UEDGE results for the location of ionization and recombination. We find that a simplified lithium vapor box configuration, without baffles, provides robust stabilization of the detachment front and acceptable lithium vapor flow to the main chamber. Lithium is evaporated from a capillary porous surface on the private-flux side of the divertor, distant from the main plasma, and is condensed on the other side walls of the divertor chamber. The condensed lithium flows back to the evaporator through 2 cm internal diameter pipes. The capillary pressure differential at the surface of the evaporator is capable of overcoming MHD back-force in the piping, with a great deal of margin if a sandwich flow channel insert is implemented. Very little lithium should be accumulated on the first wall of the main chamber if it is maintained at 600 C, as expected for a power-producing fusion system.