38 / 2021-04-23 16:58:57

Formation and control of the 3/1 locked island in the plasma boundary on J-TEXT

magnetic island,,RMP,J-TEXT,island divertor

The formation of a locked island in the plasma boundary is important for the RMP-ELM suppression [1] and the idea of island divertor [2]. The differential phase Δφ, defined by the phase difference between the top and bottom RMP coils, plays an important role in the locked island formation [3]. The experiments of the Δφ scan were carried out on J-TEXT to investigate the plasma response and formation of 3/1 locked island. Successful formation of 3/1 locked magnetic island was obtained due to 3/1 RMP penetration at a much higher electron density (n_e > 2.5) using Δφ = 270°. The density scaling of 3/1 RMP penetration shares the same threshold (b_r,vac^3/1) in different Δφ [4]. It suggests that the 3/1 RMP component dominates the 3/1 RMP penetration process while other components have few effects in this region. Unfortunately, the RMP coils contribute a large 2/1 component with Δφ = 270°, and hence 2/1 locked island is easily excited when q_a approaching 3 and triggers disruption.

It is further observed that the application of ECRH (50 ~ 400 kW) modifies the width of the 3/1 locked island. By depositing the ECRH power around the q = 1 surface, sawteeth with long periods and big crashes were excited. Following these big sawteeth, the width of the 3/1 locked island was periodically reduced. With the increase of ECRH power, the reduction of 3/1 island width became more significant. One hypothesis on this observation is that the big sawtooth crashes release a large heat pulse towards the edge and hence increase the local temperature around the 3/1 locked island, which leads to the healing of the locked island.

These results reveal the challenges of establishing an island divertor configuration in a tokamak using the 3/1 island: a) To form 3/1 island at high density regime requires large RMP coil currents, which either exceeds the engineering limit (for Δφ = 180°) or induces 2/1 island and disruption via the large 2/1 RMP component (for Δφ = 270°); b) the core/edge collapses (e.g. sawteeth or ELMs) might modify the width or the phase of the edge island frequently, which is unfavorable for a divertor configuration. Future researches will try higher m number, e.g. 4, for the island divertor study.

Reference

[1]  R. Nazikian, et al., Phys. Rev. Lett. 114 (2015) 105002

[2]  F. Karger, et al., Phys. Lett. A 61 (1977) 385

[3]  H.Wang, et al., Nucl. Fusion 58 (2018) 056024

[4]  F.Y. Mao, this Conference.