Conference Publication Details
Mandatory Fields
P.J. Mc Carthy, L. Giannone, M.G. Dunne, R. Fischer, J.C. Fuchs, K.-H. Schuhbeck, ASDEX Upgrade Team, and EUROfusion MST1 Team
44th European Physical Society Conference on Plasma Physics , Belfast
Time-dependent ELM-resolved equilibrium reconstruction on ASDEX Upgrade
2017
June
In Press
0
()
Optional Fields
P2.131
Belfast
26-JUN-17
30-JUN-17
Type I edge-localized modes (ELMs) on ASDEX Upgrade are accompanied by negative voltage spikes of magnitude ≃ 10 V measured near the lower divertor, current spikes ≃ 10 kA in measured poloidal SOL currents which correspond to several tens of kA toroidally, and a ≃ 10 kA transient increase in the 1 MA plasma current (including the SOL toroidal current) [1]. Inboard and outboard SOL currents, which are measured at a number of toroidal locations, are reasonably axisymmetric (see fig. 1) and hence ELM-resolved equilibrium reconstruction using an axisymmetric equilibrium code incorporating the SOL currents becomes feasible despite the known 3D character of ELMs[2]. An equilibrium reconstruction method that yields loop voltages on a sub-one millisecond time scale in good agreement with experiment has recently been developed and applied to Type I ELMs on ASDEX Upgrade. The method, which simultaneously determines ψ(R,Z) and V (R,Z), parameterizes the source terms in the Grad Shafranov equation by free parameters that fit both static amplitudes and time derivatives to a set of spline basis functions, and includes the dynamics of active coil currents and eddy currents. So far, the time-dependent terms are constrained by multi-timepoint magnetic data (10 kHz acquisition rate) in an interval < 1 msec centred on the selected static equilibrium timepoint. The method is implemented in the CLISTE equilibrium code [3] which extends force balance into the open fieldline region and can achieve high fidelity reconstructions of edge pressure data in both pedestal and SOL regions. Sample results for ASDEX Upgrade discharge #33589 compare finite differencing of a time sequence of ψ values (fig. 2) with the new method (fig. 3) where it should be noted that the measured loop voltage at the location of ‘loop 22’ played no part in the reconstruction. This new technique is currently being applied to reconstructions additionally constrained by internal diagnostic data with the goal of identifying the loop voltage profile and its uncertainties inside the plasma.
Grant Details