Princeton Plasma Physics Laboratory
NSTX Experimental Proposal Title: Study of the correlation between GAE activity and electron transport
(Approval date unless otherwise stipulated) Revision: OP-XP-822
(2 yrs. unless otherwise stipulated)
Responsible Author: D. Stutman Date 06/18/08 ATI – ET Group Leader: S. Kaye Date RLM - Run Coordinator: M. Bell Date Responsible Division: Experimental Research Operations
Chit Review Board (designated by Run Coordinator)
MINOR MODIFICATIONS (Approved by Experimental Research Operations)
NSTX EXPERIMENTAL PROPOSAL
TITLE: Study of the correlation between GAE activity No. OP-XP-
and electron transport
D. Stutman, L. Delgado-Aparicio, K. Tritz, M. Finkenthal (JHU) N. Gorelenkov, E. Fredrickson, S. Kaye (PPPL)
1. Overview of planned experiment
A large increase in central electron transport with beam heating power was observed in NSTX H-
modes, by changing the beam power at fixed q-profile (XP 612). This increase appears to be correlated
with a quantitative and qualitative change in the Global Alfven Eingemode (GAE) activity. To strengthen
and document this very important observation we propose to run an experiment in which we compare
electron transport in plasmas with and without GAEs. We will use three different scenarios to make this
(i) Create discharges heated by equal beam power, P but at different beam voltage, V bp(ii) Add to a baseline discharge with low GAE activity additional beam power at increasing V b
(iii) Evaluate how RF heats the plasmas with most/least GAE activity, obtained in (i) and (ii) above
These scenarios will be run in H-mode, the main NSTX operating regime.
In addition, time permitting, we will attempt to use slow ramp L-modes in order to cross calibrate
the density fluctuation amplitude measured by high-scattering with that measured by reflectometry.
The estimated run time is one day.
2. Theoretical/empirical justification
Electron transport is the dominant loss channel in beam heated NSTX plasmas. An unusual effect
is that the T profile flattens and the central ? strongly increases with increasing beam power in NSTX H-ee
modes (Fig. 1a). The TRANSP sensitivity analysis, a host of perturbative electron transport experiments,
as well as the recent FIDA data, indicate that the flattening is a genuine electron transport effect and not
the result of a broadening of the heating (i.e., beam ion density) profile by MHD activity. Furthermore,
the large central ? at high P in Fig.1a, together with a number of other observations, suggest that we are eb
dealing with electron transport along stochastic magnetic field lines. (See link below for details).
OP-XP- 2 / 7
The main puzzle is thus what is causing fast electron
transport in the central NSTX plasma in the absence of a significant 22??(m(m/s)/s)ee
Tgradient (or for that matter of any other significant thermal e 100100
gradient). A logical (if unexpected) answer may be that the free-6 MW6 MWa)a)4 MW4 MWenergy needed to drive this transport comes from the gradient of non-5050
thermal particles. Indeed, the TRANSP analysis indicates that the 22
fast ion density has by far the strongest gradient in the region r/a ?
6 MW6 MW Furthermore, a good ‘mediating agent’ between the fast ion
gradient and the electron transport may be expected to be the
persistent Alfven Eigenmode MHD activity driven by the fast ions in GAEGAE
NSTX. In particular, shear Alfven Eigenmodes (*AEs) have been TAETAE
early on predicted to be able to induce electron transport through µ-4 MW4 MW
tearing of the flux surfaces (see e.g., Lee, Okuda and Chance PRL b)b)
We therefore searched for a correlation between changes in
central ce and changes in AE activity. A quite compelling correlation
2 MW2 MWwas found with the GAE (Global Alfven Eigenmode) activity. As
recently shown at NSTX, these are high-n modes localized in the
central plasma and have a substantial shear component [see e.g., N.
Gorelenkov et al., E. Fredrickson et al.]. The correlation is illustrated
in Fig. 1b, which shows that plasmas having high central ? have e
also intense, broadband GAE activity, while plasma with lowest Fig. 1Fig. 1??behavior (a) and behavior (a) and eeGAE behavior (b) as a GAE behavior (b) as a transport is essentially GAEs free. Furthermore, the large ? increase efunction of beam heating function of beam heating
power and at fixed-q. The power and at fixed-q. The for P>2 MW (Fig. 1a) suggests a threshold in the transition to bpower is stepped at 0.42s, power is stepped at 0.42s,
from a steady 4 MW, to the from a steady 4 MW, to the stochastic electron transport, possibly correlated with a threshold in level indicated.level indicated.
the GAE mode superposition, or ‘coalescence’, as seen in Fig. 1b
also at P>2 MW. b
Last but not least, the initial theoretical assessment of a possible GAE/electron transport
connection is encouraging, in that multiple GAE modes seem to be able to induce stochastic transport of
trapped electrons [N. Gorelenkov, preliminary]. If confirmed, this phenomenon could have deep
implications for any burning plasma heated by a large population of fast beam ions and/or alphas.
OP-XP- 3 / 7
3. Experimental run plan
The goal of the proposed experiment is to strengthen and better document the inferred connection
between electron transport and GAE activity. To this end we propose to compare Tprofiles and electron e
transport in three scenarios. They are primarily based on the fact that the GAEs, being excited by the high
energy end of the fast ion distribution, have a threshold in beam voltage.
(i) discharges heated by equal beam power, P but at different beam voltages, V bp
(ii) baseline discharge with low GAE activity to which we add beam power at increasing V b
(iii) RF heating of NBI plasmas with/without GAE activity
Neon injection will be used in the first scenario to compare also particle transport.
If possible, we will also try to obtain an estimate of the GAE density perturbation amplitude using
high-k scattering, in conjunction with reflectometry. To this end we will cross-calibrate the high-k signal
in terms of??n, using a comparison with reflectometry in slow-ramp, low density L-modes (see e.g., shot e
112996). This calibration will then be used to estimate ?n in H-modes not accessible to reflectometry. e
The XP will be run mostly in H-mode, for reasons of MHD stability at high power and for
reproducibility. In addition, the elevated central-q typical of H-mode operation in NSTX, seems to favor
GAE coalescence and rapid electron transport. As baseline condition we will use a recent, 4 MW, 0.9 MA,
4.5 kG shot, nr. 129902. The run plan is as follows:
Part I: Compare plasmas with same P but different V bb
1. High V: A/90 + B/100 (2 shots) b
2. Low V: A/90 + B/65 + C/65 (2 shots) b
3. Decrease I ramp if q tends to reverse at low V (2 shots) pb
4. Neon injection at high/low V (2 shots) b
Part II: P step at increasing V bb
5. Establish baseline A/90 + C/65 (2 shots)
6. Step B at 450 ms: 60, 75, 90, 105 kV (8 shots)
Part III: Compare RF heating in NBI deuterium plasmas with/w.o. GAEs
7. Apply 2 MW RF to plasmas having largest difference in GAE content (4 shots)
Part IV: High-k cross calibration with reflectometry (time permitting)
8. High V slow-ramp L-mode (2 shots) b
Total number of shots = 24
OP-XP- 4 / 7
4. Required machine, NBI, RF, CHI and diagnostic capabilities
(1) All neutral beams operational between 60 and 105 kV; required
(2) RF with phasing optimal for heating in deuterium, P ?2 MW; required
(3) High-frequency and low frequency Mirnov coils; required
(4) MPTS at 16 ms spacing, with timing synchronized for a measurement at 450 ms; required
(5) CHERS operational and synchronized for frame starting at 450 ms; required
(6) MSE operational and synchronized for measurement starting at 450 ms; required
(7) High-k scattering taking data at R?115 cm, if possible in ‘interferometric mode’; required
13(8) Reflectometer 44.5 GHz frequency (3 10 cm-3 cutoff) ; required
(9) Three-color tangential optical SXR array; required
(10) USXR arrays in two-color configuration: Hor. Up – Be10, Hor. Down - Be100; required
(11) FIDA; desired
5. Planned analysis
TRANSP, multi-color SXR, impurity transport, ORBIT, GS2.
6. Planned publication of results
Contributions to international conferences and in refereed journals.
OP-XP- 5 / 7
PHYSICS OPERATIONS REQUEST
TITLE: No. OP-XP- AUTHORS: DATE: Machine conditions (specify ranges as appropriate)
I(kA): 4.5 kG Flattop start/stop (s): -0.02/1.0 s TF
I (MA): 0.9 Flattop start/stop (s): 0.12-0.22/0.8 sP
Outer gap (m): 0.05-0.10 Inner gap (m): 0.01-0.06
Elongation ?: 2.25 Upper/lower triangularity ?: 0.6/0.6
Z position (m):
Gas Species: D, Ne Injector(s):
NBI Species: D Sources: Voltage (kV): 60-105 Duration (s): 1s
ICRF Power (MW): 2 Phasing: TBD Duration (s): 0.3
CHI: Off Bank capacitance (mF):
Either: 121135, 121172
Or: Sketch the desired time profiles, including inner and outer gaps, ?, ?, heating,
fuelling, etc. as appropriate. Accurately label the sketch with times and values.
OP-XP- 6 / 7
TITLE: No. OP-XP-
Note special diagnostic requirements in Sec. 4 Note special diagnostic requirements in Sec. 4
Diagnostic Need Want Diagnostic Need Want
? MSE Bolometer – tangential array ?
NPA – ExB scanning Bolometer – divertor
NPA – solid state ? CHERS – toroidal ?
Neutron measurements ? CHERS – poloidal ?
Plasma TV ? Divertor fast camera
Reciprocating probe Dust detector
Reflectometer – 65GHz ? EBW radiometers
Reflectometer – correlation Edge deposition monitors
Reflectometer – FM/CW Edge neutral density diag.
Reflectometer – fixed f ? Edge pressure gauges
Reflectometer – SOL Edge rotation diagnostic
RF edge probes Fast ion D_alpha - FIDA ?
Spectrometer – SPRED ? Fast lost ion probes - IFLIP ?
Spectrometer – VIPS ? Fast lost ion probes - SFLIP ?
SWIFT – 2D flow Filterscopes ?
Thomson scattering ? FIReTIP
Ultrasoft X-ray arrays ? Gas puff imaging
Ultrasoft X-rays – bicolor ? H? camera - 1D
Ultrasoft X-rays – TG spectr. High-k scattering ?
Visible bremsstrahlung det. ? Infrared cameras
X-ray crystal spectrom. - H Interferometer - 1 mm
X-ray crystal spectrom. - V Langmuir probes – divertor
X-ray fast pinhole camera Langmuir probes – BEaP
X-ray spectrometer - XEUS Langmuir probes – RF ant.
Magnetics – Diamagnetism ? Magnetics – Flux loops ? Magnetics – Locked modes Magnetics – Pickup coils ? Magnetics – Rogowski coils ? Magnetics – Halo currents Magnetics – RWM sensors Mirnov coils – high f. ? Mirnov coils – poloidal array Mirnov coils – toroidal array Mirnov coils – 3-axis proto.
OP-XP- 7 / 7