Controlled tungsten melting and droplet ejection studies in ASDEX Upgrade

Abstract

Tungsten rods of 1×1×3mm were exposed in single H-mode discharges at the outer divertor target plate of ASDEX Upgrade using the divertor manipulator system. Melting of the W rod at a pre-defined time was induced by moving the initially far away outer strike point close to the W-rod position. Visible light emission of both the W-pin and consecutively ejected W droplets were recorded by two fast cameras with crossed viewing cones. The time evolution of the local W-source at the pin location was measured by spectroscopic observation of the WI line emission at 400.9 nm and compared to the subsequent increase of the tungsten concentration in the confined plasma derived from tungsten vacuum UV line emission.Combining these measurements with the total amount of released tungsten due to the pin melt events and ejected droplets allowed to derive an estimate of the screening factor for this type of tungsten source. The resulting values of the tungsten divertor retention in the range from 10-20 agree with those found in previous studies using a W-source of sublimated W(CO) 6 vapour at the same exposure location. Ejected droplets were found to be always accelerated in general direction of the plasma flow, attributed to friction forces and to rocket forces. Furthermore, the vertically inclined target plates cause the droplets, which are repelled by the target plate surface potential due to their electric charge, to move upwards against gravity due to the centrifugal force component parallel to the target plate.

Evolution of surface melt damage, its influence on plasma performance and prospects of recovery

“…From the above presented experiments several conclusions with respect to the performance of PFCs under melt conditions can be drawn in addition to the already presented results [1,2,4,7,15]. With leading edges exposed to the divertor plasma melting is almost unavoidable, thus shaping of divertor components intersecting with high heat and particle flux to remove leading edges is mandatory.…”

“…In contrast to previous studies of melt layer loss behavior in AUG with a molten pin exposure [4,7] a larger melt pool as seen from figure 9 is now subject to plasma impact. Studying the actual droplet dynamic after detaching from the melt is out of the scope of this paper and discussed in detail in [7].…”

“…Studying the actual droplet dynamic after detaching from the melt is out of the scope of this paper and discussed in detail in [7]. Figure 11 shows the correlation of the local WI(400.9nm) line emission together with the Dα divertor recycling-signal during the ejection of droplets of tungsten.…”

“…In both AUG and C-Mod the disruptivety increases, in the case of C-Mod up to 100% while for AUG it rises only towards 30% -still intolerable for ITER. A tungsten source caused by melting thus needs to avoided be it in the form of droplets [1,2,4,7,15] during the melt process or loss of re-solidified components.…”

“…The main differences are the impinging angle of the heat flux and magnetic field, castellation geometry and the impact on plasma operation due to screening of impurities being different in divertor geometry as seen in previous pin-melt exposures at AUG [4,7]. In this contribution mainly two 'experiments' are considered: firstly, accidental melting during tungsten dedicated run campaigns ('07-'09) in C-Mod as well as the dedicated exposures of castellated structures with an inherent leading edge at AUG.…”