Beamlet formation in negative ion sources for fusion applications: A computational grid system comparison

Divergence is a crucial parameter for neutral beam injection heating systems, since it affects the transmission of the beam through a duct. The divergence is determined by properties of the extracted ions, in combination with a grid system that extracts ions from the plasma and subsequently accelerates them to full energy. To disentangle different contributions to the divergence, eight different negative ion-based grid systems are studied with the IBSimu code, treating volume produced deuterium ions only. To ensure that the observed differences can be attributed to grid features, the grid systems are modeled without magnetic field. The ratio between acceleration and extraction potential that gives the lowest divergence shows good agreement with calculated and measured literature values. The divergence in various grid systems is studied at the ITER-heating neutral beams D- current density and extraction potential. At similar normalized emittance, there is a large range of divergences due to the variation in the axial velocity at different acceleration potentials. The normalized emittance is mostly determined by the starting emittance and emittance growth in the extraction stage. The starting emittance is due to the initial perpendicular temperature of the negative ions, which is assumed as 1 eV in simulations. The emittance growth is due to space charge in the extraction stage. Some of the investigated grid systems produce lower divergence beamlets at similar emittance by trading off beamlet size and divergence. The optimum divergence is limited by three parameters at fixed beamlet size: the D- perpendicular velocity distribution upstream of the grid system, the space charge in the extraction stage, and the available high-voltage potential to accelerate the beamlet.