Non-linear MHD modelling of edge localized modes dynamics in KSTAR

The explanation of the existence of the rotating MHD modes in the pedestal region before Type I edge localized mode (ELM) crash and in the inter-ELM periods (ELM precursors) observed in KSTAR is provided for the first time in the present paper. The dynamics of ELMs, observed using electron cyclotron emission imaging (ECEI) in KSTAR tokamak, is compared to the modelling results of the non-linear reduced resistive MHD code JOREK. The realistic KSTAR pulse parameters and geometry including X-point and scrape off layer (SOL) were used. The full ELM crash modelling was performed using JOREK code for single and multi-harmonic representation and in multi-cycles ELMy regimes including relevant flows. The most unstable toroidal modes numbers (n  =  5–8), velocity (~5 km s−1 for n  =  8 mode) and the direction of the mode rotation were reproduced in modelling. The two fluid diamagnetic effects and toroidal rotations included in the model were found to be the most important factors in explaining the experimentally observed rotation of the ballooning modes before the ELM crash and in the inter-ELM phase. In multi-harmonic multi-cycle simulations the spectrum of temperature fluctuations is similar to the experimental one in the inter-ELM phase, where several rotating modes with medium n numbers were detected in 5–30 kHz frequency range. The rotating modes can contain single or several harmonics which last from 0.2 ms to few ms in time, and can appear and disappear in the inter ELM period or persist until a new ELM crash.