Asymmetry of collapse of density pedestal by type I ELM on JT-60U
N. Oyama, Y. Miura, A. V. Chankin, H. Takenaga, N. Asakura, Y. Kamada, T. Oikawa, K. Shinohara and S. Takeji
Naka Fusion Research Establishment, Japan Atomic Energy Research Institute, Naka-machi, Naka-gun, Ibaraki-ken 311-0193, Japan
Abstract.
The study of the mechanism for the collapse of the pedestal structure due to ELM is important for evaluation the transient sever heat loads, which can erode the divertor target plate in the next step devices such as ITER-FEAT. The understanding of ELM physics makes it possible to control the ELM itself. In JT-60U, the detailed behavior of type I ELMs was studied using a heterodyne reflectometer system with high time and spatial resolution. From the analysis of time series and evaluation of particle balance during an ELM, the collapse of density pedestal was supposed as poloidally asymmetric event, such as localized only at low-field side of the plasma. This kind of poloidal mode structure was also supported by stability analysis by Lao et al. In order to confirm the poloidal asymmetry of the collapse of density pedestal caused by ELMs, the edge pedestal density for high- and low-field side was measured by vertical interferometer and reflectometer, respectively.
Since the JT-60U tokamak has a good capability of changing plasma position, the FIR interferometer can be set to measure the edge plasma from the scrape-off layer (SOL) to the core with 20 cm horizontal plasma sweep. The O-mode reflectometer with the frequency of 35.9 GHz, which corresponds to the cut-off density of 1.6x1019m-3, measures what at the pedestal. The phase change of reflected wave exhibits the radial movement of the cutoff layer. From the time series analysis using the same digitizer system for the reflectometer, Dα intensity and interferometer, the experimental evidence for poloidal asymmetry of the collapse was clearly observed in JT-60U ELMy H-mode discharge for the first time. The maximum displacement of cutoff layer reached about 5 cm at 290 μs later than starting time. This movement was not due to the plasma shift, because the variation of plasma equilibrium during ELM was estimated to be less than 1 cm. About 100 μs later from the starting time, the Dα signal at the outer-divertor region started to increase. On the other hand, there was no response on the line-integrated density at the high-field side of plasma until Dα signal started to increase, although the half of the line of sight of the interferometer integrated the top of density pedestal. The collapse of density pedestal, therefore, does not occur at the high field side of the plasma.