Qualification of Tungsten Coating on Carbon Components for the Wall of the RFX-mod Device

RFX-mod is a toroidal device for the magnetic confinement of plasmas in the Reversed Field Pinch (RFP) configuration, with major/minor radius R/a = 2/0.46 m. It operates with deuterium (or hydrogen) and helium plasmas at currents up to 2 MA and typical electron densities in the range n/nG = 0.2-0.4. Neither limiter nor divertor is present, hence nominally the thermal outflux is spread over the whole surface, giving, at the maximum plasma current, an average thermal load to the wall of 2 MW/m2. In reality, the RFP configuration is characterized by the presence of a spectrum of magnetic modes that phase lock, giving rise to a macroscopic perturbation of the plasma column that causes localised Plasma Wall Interaction (PWI) events of duration up to 10 ms where the thermal load can reach values as high as tens of MW/m2. Up to now the thermal load has been sustained by a first wall of tiles made of polycrystalline graphite that completely cover the vacuum vessel. This solution is very safe due to the excellent thermal properties of graphite, but has the drawback of making difficult the control of density during deuterium operation. In fact carbon wall retains huge amount of D, and releases it in uncontrolled way under the PWI events. A possible alternative solution for the plasma facing components of RFX-mod is under test: the coating of the carbon tiles by a tens of microns layer of tungsten. The explored techniques for the coating are all based on Plasma Vapour Deposition (PVD), and are the High Power Impulse Magnetron Sputtering (HiPIMS) PVD at IENICNR laboratory, Padova, and the Combined Magnetron Sputtering and Ion Implantation (CMSII) technology, developed at MEdC-Romanian Euratom Association. They have been preliminarily tested on small scale specimens exposed to RFX-plasmas in controlled experimental sessions of about 15 discharges, during which the interaction with the plasma has been enhanced by the active control of the magnetic boundary in order to simulate the conditions of the PWI events. They all suffered of coating damages. The CMSII technique, that gave the minor faults at the first stage, was then tested on full scale dimension by replacing 4 actual RFX-mod carbon tiles with tungsten coated ones. The tiles have been exposed to four months of normal RFX-mod operation over a large range of operational conditions, after which large damaged areas have been observed. In this contribution, we describe the small and full scale exposure experiments and the analysis of the coating damages that occurred in both cases.