2D ECEI OBSERVATION OF TEARING MODES IN KSTAR RMP EXPERIMENTS - PowerPoint PPT Presentation

2D ECEI OBSERVATION OF TEARING MODES IN KSTAR RMP EXPERIMENTS MINJUN J. CHOI * , G.S. YUN, J. LEE, W. LEE, H.K. PARK Center for Fusion POSTECH, POHANG, KOREA Plasma Diagnostics & Steady-State Operation C.W. DOMIER, N.C. LUHMANN, JR., UC DAVIS, CALIFORNIA, U.S.A. Y.M. JEON NATIONAL FUSION RESEARCH INSTITUTE, DAEJEON, KOREA * cmj0417@postech.edu

MOTIVATION OF RESEARCH In the 2011 KSTAR campaign, static 𝑜 = 1 resonant magnetic perturbations (RMPs) have been applied to control the edge localized modes (ELMs) as well as to study error fields effects *. While the ELMs control experiments demonstrated that the 𝑜 = 1 RMPs were effective to suppress or mitigate ELMs, unwanted effects of RMP have also been observed such as plasma locking and/or tearing mode, both of which could potentially lead to a confinement degradation or even disruption. We report three observations with the KSTAR ECEI system; (1) T e fluctuation of typical m/n=2/1 tearing mode, (2) 1/1 and 2/1 mode coupling, and (3) tearing mode related transport event. Also, a method to estimate Δ′ based on relatively calibrated ECEI data is introduced. * Y.M. Jeon et. al., the 53rd Annual Meeting of the APS Division of Plasma Physics, Salt 2 Lake City, USA (2011)

KSTAR RESONANT MAGNETIC PERTURBATION (RMP) Coils for magnetic perturbation fields n=1 +B R Top-RMP -B R Mid-RMP I p , B T Bot-RMP Figures are reprinted from * * Y.M. Jeon et. al., Mini-workshop for NTM experiments on KSTAR, Pohang, Korea (2011) 3

KSTAR ELECTRON CYCLOTRON EMISSION IMAGING SYSTEM 2 × 24 × 8 local electron cyclotron emission measurement* 𝑔 𝑓𝑑𝑓 Large imaging Heterodyne optics detector array (24) Heterodyne 𝑆 detector array (24) Vertical resolution : ~1.4 cm/ch Radial resolution : ~1.5 cm/ch Time resolution : 0.5, 1, 2 us 1 𝜐 𝑑 2∆𝑔 𝑤 e resolution : 1 × 𝑂 𝑗𝑡 = 𝜐 𝑗𝑜𝑢 = Δ𝑔 𝑗𝑔 ∼ 3% T * G.S. Yun et. al., RSI 81 (2010) 4

TYPICAL m/n =2/1 TEARING MODE ON KSTAR (#6123) 1-D diagnostics and KSTAR ECEI system Edge Localized Mode (ELM) is suppressed / mitigated by RMP L-mode After the turned-off static RMP, large amplitude oscillation is observed at Mirnov coils, interferometer, and ECE signals Oscillations turned out to be the 2/1 tearing mode numbers from Mirnov coil and ECE/ECEI signals 5

ECE images of 2/1 tearing mode of KSTAR Plasma is rotating in laboratory frame, and so the fixed ECEI frames sweep the plasma in the plasma frame Clear phase inversion across 𝜀𝑈 𝑓 = 0 , normalized fluctuation amplitude is larger inside ECE images are made from normalized ECEI data against average values Given 𝑔 𝐹𝐷𝐹 ∼ n ∗ 𝑔 𝑢𝑝𝑠 , 𝑜 = 1 , and 𝑔 𝐹𝐷𝐹 = 1.3 kHz, plasma rotates with 𝑤 𝑢𝑝𝑠 ∼ 18 km/s at 𝑆 = 2.17 m Fast visible bremsstrahlung images obtained from 2/1 tearing mode at DIII-D* are shown for comparison White boxes are the virtual ECEI view * Van Zeeland et. al., Nuclear Fusion 81 (2008) 6

Find shaping parameters of the un-perturbed flux from ECE image #5956 Fit 𝜺𝑼 𝒇 = 𝟏 contour with 𝑆 = 𝑆 0 + 𝑠 cos(𝜄 + 𝜀 sin 𝜄) 𝑨 = 𝑨 0 + 𝜆𝑠 sin 𝜄 𝑛 = 2 tearing mode is observed at the both high and low field sides 𝜀 = 0 ± 0.1 𝜆 = 1.0 ± 0.1 Inner separatrix position estimation with ECEI data only Separatrix position can be estimated by finding Δ𝑈 𝑓 maximum position* Assuming the flat temperature profile 〈𝑈 𝑓 〉 𝑢1 = 𝑑𝑝𝑜𝑡𝑢 inside the island, the inner Δ𝑈 𝑓 maximum position is same with the inner Δ𝑈 𝑓 /〈𝑈 𝑓 〉 𝑢1 maximum position 𝜀𝑈 𝑓 / 𝑈 𝑓 𝑢1 = max 𝜀𝑈 𝑓 / 𝑈 𝑓 𝑢1 = 0 * P.C. de Vries et. al., PPCF 39 (1997) 7

MODE LOCKING AND COUPLED MODES (#5951) When plasma is completely locked, no Mode field precursor is observed. Oscillation starts appear Locked again in the RMP ramp-down phase core ECE spectrum : precursor oscillation of ECE core 𝑛/𝑜 = 1/1 sawtooth. The sawtooth crash observation shows 𝑟 = 1 surface at 𝑠 = 13 cm Spectrogram core ECE middle ECE spectrum : T e perturbation of noise 𝑛/𝑜 = 2/1 tearing mode. 𝑟 = 2 surface is found to be 𝑠 = 31 cm (phase inversion Spectrogram middle ECE position) Both modes have n=1 and frequencies of oscillation are almost same After the turned-off of RMP, vanishing of the RMP torque may increases core plasma rotation (white arrow). The stronger RMP in this shot compared to #6123 could explain further effect on plasma 8

ECE images during the mode coupling ECE measurements of mode coupling on TEXTOR* q=2 T e perturbation (a) w/o 𝑛 = 1 mode (b) w/ 𝑛 = 1 mode Phase of m=1 and m=2 modes are synchronized at the high field side where the flux surfaces are denser * P.C. de Vries et. al., PPCF 39 (1997) 9

BIG TRANSPORT EVENT WITH RMP FIELD (#6056) ECE signals Big transport events related to perturbation fields 𝑠 = − 2.4 cm -5.6 -8.6 -11.5 -14.4 -21.1 -23.6 -26.0 -28.4 -30.7 -32.9 -38.2 10

2-D T e dynamics observed in ECE images Note that strong static RMP is applied (b — d) crescent shape decrease (inside) and increase (outside) around typical 𝑟 = 2 region. It may be due to T e flattening of tearing mode (e) core part moves upward, and small heat transport occurs (f — h) heat channel is formed at the midplane, and large signal decrease of core and increase of edge are observed (e) (f — h) 11

Δ′ ESTIMATION WITH RELATIVELY CALIBRATED ECEI DATA Ren’s tearing mode flux model* equilibrium flux TM perturbed flux constant flux condition 1 2 𝑦 2 1 + 𝜗 1 𝑦 + 𝑥 2 1 + 𝜗 2 𝑦 cos 𝑛𝜄 = 𝑧 4 ′ (𝑠 ′′ (𝑠 ϵ 1 ≡ 𝐶 𝜄 𝑡 )/3𝐶 𝜄 𝑡 ) 2 1 1 w 𝑠 𝑡 𝐶 𝑠 ϵ 2 ≡ 𝛽 ± + ≡ 𝑠 𝑡 and ′ 4 𝑛𝐶 𝜄 * C. Ren et. al., Physics of Plasma 5 (1998) 12

Application to 2-D system (#6123) relative calibration : assuming (1) flat T e profile inside the island and (2) axis symmetry of average T e profile 𝑈 𝐹𝐷𝐹 (𝑠 ′ ) 𝑢 = 𝑈 0 (𝑠) 𝜀𝑈 𝑓 / 𝑈 𝑓 = max 𝜀𝑈 𝑓 / 𝑈 𝑓 = 0 𝑠 ’ projection using 𝑠 the shaping parameters 13

Measure the T e contour and find fitting parameters 1 2 𝑦 2 1 + 𝜗 1 𝑦 + 𝑥 2 1 + 𝜗 2 𝑦 cos 𝑛𝜄 = 𝑧 4 𝑥 ∼ 3.0 , 𝜗 1 ∼ 0.082 , 𝛽 + ∼ −1.22, and 𝛽 − ∼ −0.58. T herefore, Δ ′ ∼ 0.90 Note that tearing mode in this plasma survives for a longer time after the turned-off of RMP, which is contrast to the case of #5951. Uncertainty is considered to be large and not able to quantize the error at this moment. Measurement error can be improved by increasing channel resolution 14

SUMMARY 2-D T e perturbation due to 2/1 mode tearing mode clear phase inversion in 2-D space larger fluctuation amplitude inside rational surface than outside estimation of un-perturbed fluxes surface shape Coupling between 1/1 and 2/1 mode mode locking (rotation speed changes) and synchronization of mode phases at high field side n=1 RMP related big transport event sudden temperature (and density) drop over large radial range Application of 2-D ECE diagnostics to estimate 𝚬′ 15

BACKUP normalized ECE data with values of O-point 16

BACKUP 3/2 tearing mode (#6304) 1/4 (1/6) toroidal (poloidal) turn (20 m s later) 17

TM Poster : study crash more, Mirnov coil data mode spectra analysis, 3/2 mode and also 2/1 mode onset study what happens tracking island position with ECE data EFIT data manage : q, B data 18

inside inside 19

upper channels show increase and lower decrease. the core seems to be deformed ? tearing whole this area mode ? show a little decrease whole this area show significant increase can’t find any localized heat channel in this case 20

inner HFS small large increase increase LFS almost unchanged less decrease the inner, the steeper increase of signal. there is the region where the signal does not change much (channel 1,2,3 in HFS) 21

increase no HFS change decrease LFS channels of HFS 2,1 and LFS 8,7 show decrease and channels of LFS 6 — 2 show increase 22

23