Search

Suppression or mitigation of edge-localized mode (ELM) crashes is necessary for ITER. The strategy to suppress all the ELM crashes by the resonant magnetic perturbation (RMP) should be applied as soon as the first low-to-high confinement (L–H) transition occurs. A control algorithm based on real-time machine learning (ML) enables such an approach: it classifies the H-mode transition and the ELMy phase in real-time and automatically applies the preemptive RMP. This paper reports the algorithm design, which is now implemented in the KSTAR plasma-control system, and the corresponding experimental demonstration of typical high-δ KSTAR H-mode plasmas. As a result, all initial ELM crashes are suppressed with an acceptable safety factor at the edge (q 95) and with RMP field adjustment. Moreover, the ML-driven ELM crash suppression discharges remain stable without further degradation due to the regularization of the plasma pedestal.

We explore the inboard-limited internal transport barrier (ITB) as an alternative advanced operation scenario for KSTAR. This paper presents in detail the progress of the ITB experiment at KSTAR. In an earlier study, the ITB formed in both ion and electron thermal channels, and an early neutral-beam injection (NBI) power of over 4 to 5 MW under a limited L-mode was crucial to ITB access. In the present study, we access the ITB experimentally with about 3 MW of NBI power by using the upper single null (USN), which is an unfavorable H-mode condition with a higher L–H power threshold. Finding an ITB access condition with a lower heating power should allow for a more efficient control of the heat flux and for maintaining stable plasma performance. The key control parameters of the experiment are the vertical position and the location of outboard striking point of the plasma. The shape-control attempts to divert the plasma to a vertically shifted USN with a marginal touch of the inboard limiter so that the plasma can remain in the L-mode at the boundary, while the striking-point control maintains the ITB performance for a longer period of time.

This study aims to develop a facility that can irradiate subjects with a desired low dose, which can be used to assess the biological effects of low-dose radiation. We develop a single-occupancy mouse-cage and shelf system with adjustable geometric parameters, such as the distances and angles of the cages relative to the collimator. We assess the irradiation-level accuracy using two measurement methods. First, we calculate the angle and distance of each mouse cage relative to the irradiator. We employ a Monte Carlo n-particle simulation for all of the cages at a given distance from the radiation source to calculate the air kerma and the relative absorbed dose in the in-house designed shelving system; these are found to be approximately 0.108 and 0.109 Gy, respectively. Second, we measure the relative absorbed dose using glass dosimeters inserted directly into the heads and bodies of the mice. For a conventional irradiation system, the irradiation measurements show a maximum discrepancy of 42% between the absorbed and desired doses, whereas a discrepancy of only 6% from the desired dose is found for the designed mouse apartment system. In addition, multi-mouse cages are shown to yield to significantly greater differences in the mouse head and body relative absorbed doses, compared to the discrepancies found for single-occupancy cages in the conventional irradiation system. Our findings suggest that the in-house shelving system has greater reliability for the biological analysis of the effects of low-dose radiation.