MHD Studies by X-ray Imaging -Tangentially Viewing SX Camera-

1
MHD Studies by X-ray Imaging -Tangentially
Viewing SX Camera-

• S. Ohdachi, G. Fuchs1), K. Toi, S. von Goeler2),
  S. Yamamoto3) and LHD Experimental
  GroupNational Intitute for Fusion
  Science1)Institut fuer Plasmaphysik,
  Forschungszentrum Juelich
• 2)Princeton Plasma Physics Laboratory
• 3)Dep. Energy. Eng. And Science, Nagoya University

Based on the presentation _at_ HTPD 2002 at Wisconsin
2
Outline

• Why a fast-framing tangentially viewing SX Camera
  ?
• The SX camera on Large Helical Device
• MHD activities observed by SX camera
• -- Sawtooth-like reluxation, pellet induced
  slow oscllation and m2 mode rotation
• Summary

3
Tangential View in ToroidallyConfined Plasmas.
Magnetic field lines tilted by the rotational
transform
Tangentially Viewing Camera

• Lines of sight can be almost
• parallel to the field lines
• 2D -Perturbations
• extending along the
• field lines can be directly
• visualized with high-contrast
• High-m mode like ITG

m/n 8/8 Simulation in a tokamak
4
Easy installation / Easy calibration
LHD

• With a 2D picture, pattern
• recognition is helpful for
• understanding complicated
• MHD phenomena.
• Tomography using many
• chords is elegant. However,
• it is not always possible to
• put small detectors around the plasma
• in larger devices like LHD.

5
Large-area detector is needed in pin-hole
camera
Scintillator f100mm fast video camera
SX-CCD e.g.f10mm

• SX-CCD is simple. However, to maintain spatial
  resolution,
• pinhole should also be small. Too dark for
  fluctuation study.
• Solution Scintillator screen and conversion from
  SX -gt Visible
• -gtComplicated system is needed.

6
Tangentially Viewing Camera
7
Bundle Fiber (100 x 100 pixels)
100mm
Toray PGR-FB750 0.75mm
Illuminated from the other end
(95 of photons are lost from interface and
transfer)
8
Key Components
Iron Magnetic Shield 2.5cm in thickness
Tapered FOP
MCP controller of I.I.
Fiber bundle
Coupling lens
Proxitronic BV2563 BV2503
Kodak 4540mx
Hamamatsu V4440U-mod
9
Fast-Framing Camera
Kodak(Roper) HS4540mx CMOS Image Sensor
16ch-Parallel Readout 1 pixel 40 mm ( 10.24mm x
10.24mm) 8bit 30 fps - 4500 fps ( 256 x 256 )
13500 fps ( 128 x 128 ) Buffer 512MB
(8192 frame for 256x256) 1.8s
for 4500fps and 256x256 SCSI Interface ( 2000
Frames / 5 min )
10
Noise Level and MCPAmplification

• Photon statistics
• determine the noise
• level of this camera.
• Noise increases
• with MCP voltage.
• Fluctuations larger
• than 2 can be
• detected with
• moderate amplification.

Rax3.6m Bt gt 2.6T
11
Large Helical Device (LHD)

• l 2, m 10 Heliotron type device
• R 3.5 - 3.9m, a 0.6m
• Bt 0.5 2.8 T
• NBI 9MW,ICH2.5MW, ECH1.9MW
• Te ? 10keV, Ti ? 5keV
• ne 1x1019 1020 m-3
• Rotational transform
• profile is different
• from tokamaks.
• q2(core),q1(edge)

n/m 1/2 1/1
12
LHD / SX Camera
We can put the system Far away from the device
LHD

• Tangential port(6T)
• Scintillator Screen
• Fiber Bundle
• Controllers
• CameraIron Shield
• 20G (15m from center)

13
Viewing Area
14
Images of different plasmas Comparison with
simulations
3.9m 3.6m peaked 3.6m hollow
Simulations
Left-top corner of images is cutoff by vacuum
vessel. 64x64 resolution (fibers)
Experiments
15
Example 1 Pellet Injection Movie
The camera is fast enough to study fast
phenomena 10 kHz.
16
Example 2 Sawtooth-like phenomena
Raw movie 8bit Camera
PIN diode 14bit ADC
Subtracting off the first frame
17
How do we analyze the data ?

• Video data contains much information. However, it
  is not easy to extract interesting
• components, because fluctuating component and
  noise level have comparable amplitudes.
• We need a method to extract the fluctuating
  information.

18
Data Analysis using Singular Value Decomposition
(SVD)

• SVD decomposition can decompose multi-channel
  data into independent components.
• Spatial Structure and its time evolution are
  obtained simultaneously.
• Spatial structure becomes recognizable it can be
  analyzed further, if needed, through tomographic
  reconstruction.
• SVD is commonly used to analyze magnetic probe
  array data in investigating magnetically confined
  plasma fluctuations.
• It was also applied to 2D image data.
• SVD C. Nardone, Plasma Phys.Ctrl.Fusion
  34(1992) 1447

19
SVD 2 multi channel data ?comprehensive
components
20
Sawtooth-like phenomena again

• Each component
• is recognizable
• Decrease in the core
• and increase at the edge

21
Slow oscillations are induced
pellet
m1 oscillation 500Hz around q 1
Fluctuating componets are extracted
outboard
Inboard
22
Estimate of q-profile?
Estimate of q-profile will be possible, if the
perturbation is well aligned to the magnetic
field lines. Rotation of perturbation on q1
surface.
23
Extraction of the Fluctuating Component
Simulation
Tangential view
m2 structure
24
Summary

• Tangentially viewing system is useful to study
  the low-frequency ( f lt 2.0kHz) MHD fluctuations
  ( dIsx/Isx 1) on LHD.
• With new video camera with higher framing rate (
  20kHz) and wider dynamic range (10bit), we can
  cover most of the MHD phenomena observed in LHD.
• Experiments on TEXTOR tokamak are planned.
  (2003) Interpretation will be easier
• in tokamaks.

25
Control System
26
Scintillator CsI coated FOP
(fiber optic plate)
FOP
produces 940 photons (at 570nm) 940 x0.5 ½
go to FOP x0.75 packing fraction of
FOP x0.85 Loss at the interface 300
photons are transferred to next stage(fiber
bundle)
CsI(Tl)
CsI Emission Light Spectrum
An 1.5keV photon
f100
Viton O-ring
50mm 12mm
Response time 1 ms
27
Loss of Photons via fiber
Plastic fiber transmission rate
340 x 0.5 NA of plastic fiber x 0.9
Transmittance in the fiber x 0.75
Loss at the Interface x 0.4 Quantum
Efficiency 40 Photons can be
used at the first image intensifier. (
40/940 4 )
1 dB for 9m fiber
28
Image Intensifier
Response time of phosphor
Image intensifiers are fast enough to study
fluctuations. (P47 is used. Freq. 100kHz
) Amplification factor of the MCP is
controlled externally to use the full dynamic
range of the fast camera (Response time of CsI
is also in the order of 1ms.
P47
1ms
29
Estimate of the Noise Level

• Photon statistics determine the noise level of
  this camera.

30
Sample data with
Rax 3.9m 3.6m 3.6m
Ne Discharge
Radiation Profile
Simulation
Measurement
31
SX Camera on LHD

LHD 6T Tangential port
9m Fiber Bundle
Scintillator
Controller
Camera with Iron shield
20G ( 15m from the center of LHD )
32
3D view of m2 component