In work performed under a contract with ITER IO, the effects of plasma response on the structure of the edge magnetic field created by ELM-stabilization coils in ITER H-mode operational scenarios are evaluated using the 3D MHD code M3D-C1. Previous studies of the DIII-D ELM-control coils have used the vacuum Island Overlap Width criterion as a figure of merit when evaluating the expected effectiveness of these coils to suppress Edge Localized Modes (ELMs). In order to obtain a more accurate estimate of the capability of magnetic coils to suppress ELMs, new correlation criteria for ELM suppression by Resonant Magnetic Perturbations (RMPs) are obtained by considering the effects of plasma response on the perturbed magnetic field in a database of DIII-D discharges in which RMPs were applied. These criteria are found to correlate with the observed occurrence of ELM suppression more strongly than previous criteria based on vacuum modeling. The efficacy of ITER’s Internal Control Coils is estimated by examining these criteria in eight ITER scenarios. It is found that at full ELM-coil current, the threshold values of all metrics considered are achievable in six of the eight scenarios; in the other two, three of the four threshold values are achievable. The effects of the applied field and the plasma response on magnetic topology, displacement of the plasma edge, and divertor heat flux are also quantified. In the cases considered, the self-consistent inclusion of plasma response from M3D-C1 is found to broaden the divertor footprint relative to the vacuum fields. Small broadening and no elimination of strike line striation as obtained with an ideal MHD like Model earlier were observed. This was modeled with the EMC3-Eirene code taking into account the M3D-C1 plasma response fields in collaboration with Forschungszentrum Julich.
More progress has been made on a new spectral 3D equilibrium code. In the past, we have reported on the nonlinear solver required to compute the field line mapping when it exists. This capability is sufficient for neoclassical transport studies. However, for implementation into gyrokinetic or related codes, an additional pair of linear partial differential equations must be solved. In the 2D limit, we have verified that the new 3D solver exactly recovers the Miller local equilibrium solution, and it does so with spectral accuracy. We find the 3D implementation of the 3D solver to be more systematic and intuitive than the original Miller formulation. Work continues on the 3D numerical solver.
The general purpose mapping code derived from the mapping in GATO was extended to read equilibria generated directly by the CHEASE equilibrium code. The mapping can now process equilibria generated from the major tokamak fusion equilibrium codes and produce the input needed for various codes, including GATO, NOVA-K, and GYRO. Since the MARS-F and MARS-Q codes require equilibria from CHEASE, the facility to treat CHEASE equilibria directly will streamline benchmarks of these codes with GATO. Plans are now underway to develop the facility to handle 2D equilibria generated by the stellarator code VMEC.
Previous NIMROD simulations of massive gas injection (MGI) for disruption mitigation in DIII-D have shown that applied n=1 fields can influence the phase of the unstable m/n = 1/1 mode during the thermal quench (TQ) (see Highlights from April 5 and August 9, 2013 at Theory Weekly Highlights for April 2013 and Theory Weekly Highlights for August 2013). NIMROD simulations have been carried out to compare and validate against recent DIII-D experiments in which n=1 fields were applied prior to MGI. In one experiment, the new CERBERUS MGI valve, with a different poloidal and toroidal location from the old valve, was used, and a clear n=1 variation in radiation asymmetry vs. applied n=1 phase was observed. A source model corresponding to the CERBERUS configuration was added and the particular applied field spectrum obtained during these experiments was supplied to NIMROD. A key question to be addressed by the simulations is whether the applied fields actually lock the 1/1 mode phase or simply influence its growth rate and amplitude, as was seen in earlier NIMROD simulations. In the first simulation for one particular phase, the 1/1 mode appears to saturate in the phase that is dictated by the gas jet location, and not the phase of the applied fields. The radiation toroidal peaking factor (TPF) seen in that simulation is comparable to the values obtained experimentally. Simulations of the remaining three phases will determine what, if any, effect the applied fields have on the mode evolution and how this explains the variation in TPF seen in DIII-D.
These highlights are reports of research work in progress and are accordingly subject to change or modification