Transient induced tungsten melting at the Joint European Torus (JET)

Melting is one of the major risks associated with tungsten (W) plasma-facing components (PFCs) in tokamaks like JET or ITER. These components are designed such that leading edges and hence excessive plasma heat loads deposited at near normal incidence are avoided. Due to the high stored energies in ITER discharges, shallow surface melting can occur under insufficiently mitigated plasma disruption and so-called edge localised modes-power load transients. A dedicated program was carried out at the JET to study the physics and consequences of W transient melting. Following initial exposures in 2013 (ILW-1) of a W-lamella with leading edge, new experiments have been performed on a sloped surface (15 degrees slope) during the 2015/2016 (ILW-3) campaign. This new experiment allows significantly improved infrared thermography measurements and thus resolved important issue of power loading in the context of the previous leading edge exposures. The new lamella was monitored by local diagnostics: spectroscopy, thermography and high-resolution photography in between discharges. No impact on the main plasma was observed despite a strong increase of the local W source consistent with evaporation. In contrast to the earlier exposure, no droplet emission was observed from the sloped surface. Topological modifications resulting from the melting are clearly visible between discharges on the photographic images. Melt damage can be clearly linked to the infrared measurements: the emissivity drops in zones where melting occurs. In comparison with the previous leading edge experiment, no runaway melt motion is observed, consistent with the hypothesis that the escape of thermionic electrons emitted from the melt zone is largely suppressed in this geometry, where the magnetic field intersects the surface at lower angles than in the case of perpendicular impact on a leading edge. Utilising both exposures allows us to further test the model of the forces driving melt motion that successfully reproduced the findings from the original leading edge exposure. Since the ILW-1 experiments, the exposed misaligned lamella has now been retrieved from the JET machine and post mortem analysis has been performed. No obvious mass loss is observed. Profilometry of the ILW-1 lamella shows the structure of the melt damage which is in line with the modell predictions thus allowing further model validation. Nuclear reaction analysis shows a tenfold reduction in surface deuterium concentration in the molten surface in comparison to the non-molten part of the lamella.