Radio frequency (RF) sheaths are suspected of limiting the performance of present-day ion cyclotron range of frequencies (ICRFs) antennas over long pulses and should be minimized in future fusion devices. Within the simplest models, RF-sheath effects are quantified by the integral VRF = ∫ E∥ · dl where the parallel RF field E∥ is linked with the slow wave. On 'long open field lines' with large toroidal extension on both sides of the antenna it was shown that VRF is excited by parallel RF currents j∥ flowing on the antenna structure.In this paper, the validity of this simple sheath theory is tested experimentally on the Tore Supra (TS) ITER-like antenna prototype (ILP), together with antenna simulation and post-processing codes developed to compute VRF. The predicted poloidal localization of high-|VRF| zones is confronted to that inferred from experimental data analysis. Surface temperature distribution on ILP front face, as well as ILP-induced modifications of RF coupling and hot spots on a magnetically connected lower hybrid current drive antenna, indicates local maxima of dc plasma potential in both the upper and lower parts of the ILP. This result, qualitatively conforming to VRF simulations, is interpreted in terms of j∥ flowing on ILP frame.Once the validation is done, such reliable theoretical models and numerical codes are then employed to provide predictive results. Indeed, we propose two ways to reduce |VRF| by acting on j∥ on the antenna front face. The first method, more adapted for protruding antennas, consists of avoiding the j∥ circulation on the antenna structure, by slotting the antenna frame on its horizontal edges and by partially cutting the Faraday screen rods.The second method, well suited for recessed antennas, consists of compensating j∥ of opposite signs along long flux tubes, with parallelepiped antennas aligned with (tilted) flux tubes.The different concepts are assessed numerically on a two-strap TS antenna phased [0, π] using near RF fields from the antenna code TOPICA. Simulations stress the need to suppress all current paths for j∥ to substantially reduce |VRF| over the whole antenna height.
Recent developments in ICRF antenna modelling