Improved understanding of beamlet deflection in ITER-relevant negative ion beams through forward modelling of Beam Emission Spectroscopy

Abstract NBI systems for the ITER tokamak will require large (1 m × 2 m cross section) and powerful negative ion beams to achieve the required 16.5 MW per beamline. Such negative ion beams can only be created from the combination of many individual beamlets – 1280 per beam in the case of the ITER heating neutral beam (HNB). These beamlets are required to have a core divergence of less than 7 mrad. However in these large and powerful beams, individual beamlets cannot be measured, and divergence values can only be obtained for a mixture of multiple beamlets. To investigate this, experiments were performed at the BATMAN Upgrade (BUG) test facility, a negative ion source roughly 1/8 the size of the ITER HNB. Through Beam Emission Spectroscopy (BES) measurements, and synthetic diagnostic results from the code BBCNI, insights into the effects of beamlet mixing on experimental observations have been obtained. For a single beamlet, it is expected that BES measurements obtain a Doppler shifted Hα peak with a roughly Gaussian profile, the width of which is related to the beamlet divergence, and the central wavelength is a function of the beamlet energy and the BES observation angle. In experiments at BUG, it was observed that the amount of Doppler shift in some BES lines of sight depended on the ratio of voltages applied to the grid system, even though the total voltage remained the same. By combining diagnostic data from the experiment and forward modelling from BBCNI, it was found that the change in Doppler shift for the affected lines of sight was caused primarily by changes to the horizontal deflection of beamlets by uncompensated magnetic fields – termed ‘zig-zag’ deflection due to the alternating polarity by aperture row. The amount of Doppler shift was further modulated by the beamlet divergence, which affects the amount of beamlet mixing within the BES lines of sight.