Quantification of X3 absorption for ITER L-mode parameters in ASDEX Upgrade

For an early H-mode access in hydrogen, ITER considers operating at 1/3 of the full field using 170 GHz X-Mode for heating at the 3rdharmonic. The optical thickness for such a heating scheme depends onTe2. It is rather low in the ohmic phase (withTeabout 1-2 keV), but reaches high single pass absorption for the strongly EC heated plasma withTeexceeding 10 keV. Launching ECRH into an ohmic plasma may trigger a boot-strap process onTeif the additional power absorption due to increasingTeexceeds the additional power losses due to increased transport (which often tends to increase with input power). In this contribution we present measurements of the X3 absorption for the parameter range relevant for ITER, i.e.ne2 1019m−3,Te2keVin order to back up theoretical estimates used for the modeling so far. In ASDEX Upgrade (AUG) such low densities cannot be reached in H-mode such that dominant heating with NBI is not an option. For moderateTe, it is also not an option to use X3 heating as main heating, due to the excessive stray radiation threatening in-vessel components. This dilemma is solved with the 2-frequency EC system of AUG. The main central heating is done with the lower frequency of 105 GHz at the 2ndharmonic and full single pass absorption. Up to 3.5 MW of ECRH are used at that frequency to varyTe. Two other gyrotrons are used at 140 GHz to probe the X3 interaction close to the plasma center with a sequence of short blips. The expected values of single pass absorption are calculated with TORBEAM and vary from 7% to 70%. Below 40% single pass absorption the non-absorbed power triggers an arc in the tile gaps of the inner heat shield which screens the thermo-couples from the incoming beam such that they cannot be used. Between 40% and 80% single pass absorption, the predictions and measurements agree within the uncertainty of the measurement, unless we have clear evidence for non-linear interactions, which are not described by TORBEAM and which are not expected in ITER, but are due to some specific experimental choices for an isolated subset of our results.