2 nd threshold of error field penetration in n=1 and n=2 mixed NA - - PowerPoint PPT Presentation

2nd threshold of error field penetration in n=1 and n=2 mixed NA field experiment

2014 KSTAR Conference February 26, 2014

Jayhyun Kim (jayhyunkim@nfri.re.kr)1, Y.In1, G. Kim2, G. Yun2, Junyoung Kim3, J.W. Yoo3, J. Seol1, J.K. Park4, Y. Park5, S.A. Sabbagh5 and the KSTAR team

Acknowledgement to C. Paz-Soldan (General Atomics, United states)

1National Fusion Research Institute, Korea 2Pohang University of Science and Technology, Korea 3University of Science and Technology, Korea 4Princeton Plasma Physics Laboratory, United states 5Columbia University, United states

1

• ITER and also KSTAR plan to use various non-axisymmetric (NA) field controls such

as EF correction (n=1, 2), RWM (n=1, 2…), and ELM (n=4, 3).

• The fields used in the controls have their own optimal spectra.
• They affect each other as resonant and also non-resonant way.
• We should address the effect of overlap (especially among different toroidal modes)
• In theory (A.J. Cole et al., 2007 Phys. Rev. Letters):

– The plasma is less susceptible to error-field penetration and locking, by a factor that depends on the non-resonant error-field amplitude.

• In experiment (M.E. Fenstermacher et al., 2008 Phys. Plasmas):

– ELM suppression is obtained over an increasing range of q95 by adding n=1 perturbations to “fill in” gaps between islands across the edge plasma.

• Multi-purpose in-vessel control coil (IVCC) in KSTAR is a good tool for validating the

effect of overlap.

Motivation of work

2

• Y. M. Jeon
• n, IAEA

EA FEC 2012 12, EX/3-3

• All the coils are internal thus NA field could be effectively coupled to plasmas.
• n=1 or 2 field is applicable per each row with various toroidal phase.
• Three rows of FEC coils can provide various poloidal magnetic spectra.

Configuration of Field Error Correction (FEC) in KSTAR

3

+ +

• +
• +

+ +

• L

D H P

+ +

• +
• +

+ +

• L

D H P n=2 (even) + n=1 n=2 (odd) + n=1 Port Top Mid Bot Port Top Mid Bot

0° 180° 270° 90° 0° 180° 270° 90°

Spectrum of field

n=2 odd n=2 even n=1 Reference plasma (KSTAR no. 8889): Ip=600 kA, BT=2 T, pure Ohmic, limited Vacuum analysis based on the equilibrium

• nly with magnetic diagnostics (no MSE)

q=2 q=2 q=2

4

Typical procedure of error field penetration

5

Slide- away

Locking Minor disruption Density pump-in(?) Minor disruption Locking

• n=1 field is gradually increased to cause the disruption.
• n=2 field is constantly applied during n=1 field increase.

Error field penetration has two thresholds.

6

Slide- away

Density pump-in(?) 1st threshold 2nd threshold 1st threshold 2nd threshold

ECEI location and Poincare map

7

Poincare map with ECEI window from H port q=2 surface

ECEI (L,H)

n=1 field (exaggerated) Top view

ECEI catches island opening and minor disruption.

Minor disruption Island opening

8

Plasma is locked earlier under higher n=2 field regardless of poloidal spectra of n=2 field.

9

Late locking: #9338 (1 kA/t, odd) #9367 (1 kA/t, even) Early locking: #9339 (2 kA/t, odd) #9368 (2 kA/t, even)

Plasma is less susceptible to minor disruption under higher n=2 even field.

10

#8889 (no n=2) > #9367 (n=2, 1 kA/t) > #9368 (n=2, 2 kA/t) Early disruption No disruption Relative change Slide

• away

Higher n=2 even field seems to hinder the growth of n=1 mode amplitude.

• Pure n=1 field discharge (blue trace) without applying n=2 field

does not show the change of growth rate after locking phase.

11

No n=2 field n=2 even 1 kA/t n=2 even 2 kA/t

n=2 odd field shows almost same minor disruption among different n=2 field strengths.

12

• #9338 (n=2, 1 kA/t) and #9339 (n=2, 2 kA/t) exhibit similar minor disruption.
• They are still robust to minor disruption when compared with #8889 (no n=2).

Relative change Slide

• away

n=2 odd field also hinders the growth of n=1 mode although the effect is less effective than n=2 even field.

13

• #9338 (n=2, 1 kA/t) catches up #9339 (n=2, 2 kA/t) despite of late locking.

No n=2 field n=2 odd 1 kA/t n=2 odd 2 kA/t

Just before final disruption, the mode amplitude shows rapid growth again.

• After the locking phase, the growth rate of the mode amplitude decreases.
• This kind of growth pattern is common in both n=2 even and odd discharges

although the saturation (?) level is different.

14

Locking phase Final growing phase Saturation (?) phase 2nd threshold of error field penetration

• Application of n=2 even field makes plasma less susceptible to minor disruption by

n=1 field.

• The delay of minor disruption by n=1 field is proportional to n=2 field strength in

n=2 even configuration.

• n=2 odd field does not show the proportionality to the field strength in preventing

minor disruption. However, when compared with no n=2 field discharge, they still exhibit the robustness against minor disruption by n=1 field.

• The investigation of mode amplitude reveals that n=2 field hinders the growth of

n=1 mode after locking phase.

• In the mixed NA field discharges, the mode amplitude shows rapid growth again

just before final disruption.

• The timing of locking (density pump-in, Te drop, and rotation change) looks related

to n=2 field strength regardless of n=2 field configuration (even or odd).

• The experiments will be conducted at various q95 values.

– If mode-mode coupling/competition really matters, it could affect on the coupling.

Summary and future work

15

Supplements

16

Spectrogram of Mirnov coil: no n=2 and n=2 even

n=1 + n=2 even 1 kA/t n=1 + n=2 even 2 kA/t n=1 only m/n=?/1 m/n=?/1 m/n=?/2 m/n=?/1 m/n=?/1 m/n=?/1 Pre-existing mode before applying NA field n=2 field on n=2 field on n=2 field on Locking Almost same frequency but long-lasting in n=2 even 2 kA/t (?) Locking

17

n=1 + n=2 odd 1 kA/t n=1 + n=2 odd 2 kA/t m/n=?/1 m/n=?/1 n=2 field on n=2 field on Locking Locking

Spectrogram of Mirnov coil: n=2 odd

Pre-existing mode before applying NA field m/n=?/1 m/n=?/1 Almost same frequency but eventually disappear before disruption

18

Sawtooth activities disappear when plasma locks.

• Electron temperatures in core region are dropped with locking.

19