MHD-stabilized reactor-relevant fusion plasmas (with full integration of edge, divertor and core)

Abstract

Since 2013, the UNIST Fusion center has been studying the fusion plasma stability and confinement, while spearheading advanced imaging diagnostics [1]. This led us to directly measure the nonlinear interaction between resonant magnetic perturbation (RMP) and turbulent eddies during RMP-driven, edge-localized-mode (ELM) crash suppression, clarifying the role of the ExB rotation shear [2, 3]. Currently, we are conducting a comprehensive investigation to realize MHD-stabilized conditions in a fully integrated manner at edge, divertor and core areas, essential for fusion reactor. Specifically, three research focuses are grouped in terms of 1) “Control”, 2) “Advanced Diagnostics” and 3) “Numerical Simulation” respectively. In the area of “Control”, the main emphasis is given to the physics mechanism of RMP-driven, ELM-crash-suppression, as well as that of divertor heat flux broadening under RMP, while securing its compatibility with core plasmas even in a long pulse. To further advance the imaging diagnostic capability, a next-generation receiver of electron cyclotron emission imaging (ECEI) diagnostic is being developed. In the area of “Advanced Diagnostics”, we continue to explore the possibilities of advanced diagnostics for reactor-relevant fusion plasma through full utilization of ECEI and high-speed RF systems on KSTAR, leading to localized measurement of high frequency electromagnetic fluctuations [4], and semi-automated images of MHD and turbulence fluctuations. In the area of “Numerical Simulation”, we have launched a gyro-kinetic simulation study about the interaction between plasma turbulence and Neoclassical Tearing Mode (NTM) that could degrade the performance of magnetic confinement, as well as a numerical prediction of NTM creation and its behavior. This talk will not only introduce the progress in each area, but also outline the latest research highlights. For example, recent KSTAR experiments helped us discover a strong up-down asymmetric dependence of RMP on ELM-crash-suppression [5]. Also, in view of divertor heat fluxes, the RMP-drivenELM mitigation appears slightly more favorable than RMP-driven, ELM suppression [5]. Meanwhile, the multi-year long ECEI improvements led us to expedite the post-processing of turbulence images via machine learning [6]. Preliminary GENE simulation results show non-monotonic temperature profiles inside magnetic island, whose aspect ratio dependence has been systematically addressed [7]. The details are to be presented in various conferences, as well as to be published.

[1] H. K. Park, Advanced in Physics: X 4, 1633956 (2019) [2] J. Lee et al, Phys. Rev. Lett. 117, 075001 (2016) [3] J. Lee et al, Nucl. Fusion 59, 066033 (2019) [4] M. Kim et al., accepted at Nuclear. Fusion (2020) [5] Y. In et al, IAEA-FEC (2020) [6] G.S. Yun et al, AAPPS-DPP (2020) [7] E.S. Yoon et al, to be published (2020)