Modelling of tokamak glow discharge cleaning II: comparison with experiment and application to ITER

Abstract

The primary function of the ITER glow discharge cleaning (GDC) system will be the preparation of in-vessel component surfaces prior to the machine start-up. It may also contribute to tritium removal in the nuclear phase. In GDC, conditioning efficiency is strongly dependent on the homogeneity of the flux of ions impinging onto wall surfaces. In order to assess the wall particle flux distribution in ITER, a novel 2D multi-fluid model, described in a companion paper, has recently been developed and is benchmarked here against both experimental glow discharge data obtained in a small laboratory chamber with cylindrical geometry and from two large toroidal devices: the JET tokamak and the RFX reverse field pinch. In the laboratory plasma, simulated and measured plasma electron density and temperature are in a good agreement in the negative glow region, while discrepancies exist in the anode glow, where the fluid description of the model is inaccurate due to long mean free paths of electrons. Calculated and measured ion flux distribution profiles in RFX are found in good agreement, whereas in JET comparison it is more difficult, due to the complex geometry of the first wall which leads to local inhomogeneities in the measured flux. Simulations of H2-GDC for ITER with one or two anodes indicate fairly homogeneous plasma parameters and wall ion flux in the negative glow at 0.5 Pa, a commonly used gas pressure for GDC in existing fusion devices. Although the axisymmetric geometry in the model does not allow all seven ITER anodes to be powered simultaneously in the simulations, the results can be extrapolated to the full system and predict ion current densities on wall surfaces close to the simple expectation of total anode current divided by wall surface area (0.21 A m−2), which is relevant to GDC in JET and other machines.