Novel comparative measurement of H- beam divergences at the BATMAN Upgrade test facility: single beamlet and a group of beamlets

For the ITER fusion experiment, two neutral beam injectors (NBIs) are required for plasma heating and current drive. Each NBI supplies a power of about 17 MW, obtained from neutralization of 40 A (46 A), 1 MeV (0.87 MeV) negative deuterium (hydrogen) ions. Additionally ITER NBIs involve strict requirements regarding the beamlet divergence (< 7 mrad); the full beam is composed out of 1280 beamlets, formed in 16 beamlet groups. BATMAN Upgrade is a NBI test facility at IPP, which contributes to the R&D roadmap towards the ITER NBI. BATMAN Upgrade uses an ITER-like grid with one beamlet group, which consists of 70 apertures; beamlet calculations predict a minimum divergence in the range of 11-14 mrad at this setup. Beam Emission Spectroscopy (BES) is one of the main beam diagnostic tools used at the IPP NBI test facilities. BES measures an integrated signal along its line-of-sight covering the beam composed of the overlap of several beamlets. Spatially resolved CFC tile calorimeters are tools for the measurement of the footprint of individual beamlets. During a recent experimental campaign performed by IPP and Consorzio RFX to better assess the beam optics, the divergence of a single beamlet was compared to a group of beamlets. To this purpose, the top part of the BATMAN Upgrade grid system was selectively masked to leave just a single aperture open; the bottom part was left fully open for measurements of a group of beamlets. The single beamlet is measured by the CFC tile calorimeter and BES, whereas the divergence of the group of beamlets is measured with BES. Both diagnostics measure at a similar distance of 1.3-1.4 m from the last grid. These measurements showed for the first time two major results: first, the evaluation of the divergence from the two beam diagnostics showed an excellent agreement for the single beamlet. Second, the divergence of the single beamlet is in the range of the predicted minimum divergence and about a factor of two smaller than the estimation of the signal of the group of beamlets. These results initiate now further thorough investigations with performance scans in the experiment accompanied by modeling in order to get a deeper insight into the different behavior of the divergence of a single beamlet and the divergence obtained from the overlap of several beamlets.