Measurements of dust movement by fast TV camera in JT-60U Nobuyuki Asakura , H.Kawashima, N.Ohno 1) , T.Nakano, S.Takamura 2) , Y.Uesugi 3) 1) Japan Atomic Energy Agency, Naka, 2) EcoTopia Science Institute, Nagoya Univ., Nagoya, 3) Graduate School of Engineering, Nagoya Univ., Nagoya, 4) Faculty of Engineering, Kanazawa Univ., Kanazawa 10th meeting of ITPA "SOL and divertor physics" Topical G roup Jan.7-10, 2008, Avila
20
Embed
Measurements of dust movement by fast TV camera in JT-60U Nobuyuki Asakura, H.Kawashima, N.Ohno 1), T.Nakano, S.Takamura 2), Y.Uesugi 3) 1) Japan Atomic.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Measurements of dust movement by fast TV camera in JT-60U
2)EcoTopia Science Institute, Nagoya Univ., Nagoya, 3)Graduate School of Engineering, Nagoya Univ., Nagoya,
4)Faculty of Engineering, Kanazawa Univ., Kanazawa
10th meeting of ITPA "SOL and divertor physics" Topical Group
Jan.7-10, 2008, Avila
CONTENTS
1. Introduction:
Dust collection results Dust transport in plasma Fast visible TV camera measurement
2. Dust movement in main chamber (Case-1, -2)
3. Dust movement at divertor in ELMy H-mode
(Case-3, -4, -5)
4. Influence of dusts from outer divertor on main plasma (2007 Dec.)
5. Summary
Under baffle tiles 10.5ms (26mg/m2)
1. Introduction
Dusts were collected in one poloidal section (1/32 toroidal area), and they have been deposited in W-shaped divertor operation (1997/1999 to 2003).
Ref. Masaki, et al.J. Nucl. Mater. 337-339 (2005) 609
Dusts were cumulated on lower PFCs:• not only tile surface but also under
tiles of the outer baffle.( fig.)
• larger number of dusts were cumulated under the divertor plates.
( next page)
Dust collection from PFC(#10-section):
Dusts in fusion devices potentially cause:• Damage to Plasma Facing Components(PFCs) and viewing windows,• Impurity source, • Tritium reservoir in the shadowing area.
・ Collected dust was 170 mg (for 1/36 toroidal area): 95% was on the lower VV, and 90% was under the divertor plates:
Number of dusts were found under divertor (shadow area)
most dense area was under outer (Low-Field-Side) divertor and dome, i.e. on the exhaust route of gas flow by divertor pumping in experiments.
Dust production was smaller than C erosion at LFS divertor
Erosion rate at LFS divertor surface
Dust production rate is ~7% of averaged- erosion rate at outer divertor surface (3 mg/s).
However, T-retention is potentially a problem in reactor: high (D+H)/C~0.7-0.8 of co-deposition layers was found under dome.
Extrapolation to total dust weight in the whole VV is 7 g⇒ averaged Dust production rate is estimated to be ~0.2mg/s (7g is divided by NBI period of 3x104s in 13000 shots)
Carbon from the outer divertor may be transported and deposited to inner target.
an order larger than co-deposition ratio at target and dome tiles. (Proc. 21th IAEA, Masaki et al.2006, EX-P4-14)
⇒ retention measurement in dusts should be required.
Dust transport in fusion plasma
SOL flow
Fg (gravity)
B//
SOL plasma
-q
ErFE (Electro-static)
Froc (ablation)
Ffric (friction)Ffric
Fg
B//
Sheath plasma
-q
Ep
FE
Drift flow
SOL flow
Dust in fusion plasma is generally charged negatively : Z ~ 103-104
Dust movement ( MddVd /dt = F ) is determined mainly by ・ Ffric (friction force by SOL flow and diamagnetic & ExB drift flows), and FC (Coulomb scattering),・ FE (electro-static force by potential) and FL (Lorentz force: small),・ Fg (gravitational force: small for normal size dust rd < 0.1mm),・ Froc (rocket force by ablation: small for lower Td < 2000-3000K),Dust movement is influenced by PFC geometry (first wall, divertor), background plasma (core, SOL, sheath), plasma turbulence (blob, ELM).Dust mass (Md) is reducing by sputtering, RES, sublimation Life time
neutron shield
Fast TV camera& Optical system
Mechanical shutter Interference-filter
Camera Memory & Controller
Timing control system
Ip start
Ethernet+Optical link
JT-60U machine room
15m
scope
JT-60U
Viewing divertor area with Δ~3mm
-Image guide
Control room:PC
,Camera control ,Data acquisition Timing pulse control
Dust measurement with fast TV cameraVisible light image was measured with fast TV camera from tangential port: Typical frame rate of 2 kHz (1024x1024 pix, 3s) - 6 kHz (512x512 pix, 6s). Narrow (9º) and wide (35º) viewing angles for divertor and main plasma, respectively, can be selected.
Timing control (CAMAC)
Camera &Image-Intensifier
Lens shutter
Memory& Controller
Neutron shieldViewing main plasma with Δ~1cm
(Photoron FastCAM II with full 6.5Gbyte memory)
2. Dust movement in main chamber
(L-mode Hydrogen plasma during NBI in all graphite first wall)
Fast TV scope was installed at the same port of CCD monitor camera
Tangential viewing from P17 viewing port: CCD monitor camera
Dust movements in ELMy H-mode plasma (case 1)A number of dusts were observed, in particular, at start of the first NB shot after high Ip plasma disruptions and overnight-GDC.
Shot#46034: Start of an ELM even (at 5.26s) is defined as t = 0ms
1
2
3
4
Dust-1 moves in nearly toroidal (ion drift) direction: Vt ~4.5m(toroidal distance) / 14ms(28 frames) = 0.32km/s similar to SOL flow direction, but Vt was small (plasmaV//
SOL~10km/s)
28 frames (14ms)
18 frames (9ms)ELM
Dust movements in ELMy H-mode plasma (case 1)Dust 2: moving in nearly radial direction (exhausted from NBI port): Vr ~0.5m(radial distance)/9ms = 0.06km/s, then changing toroidally.Dust 3: moving in toroidal (ion drift) direction: Vt ~2m/6.5ms =0.3km/s
then accelerating toroidally or sublimated in edge ?.
Shot#46034
?
3. Dust movement at divertor in ELMy H-modeBetter spatial and time resolutions are required for dust measurement in divertor: narrow viewing scope was used (r ~3 mm, 4-6 k-frame/s)
512x512 (6k frame/s)
A few dusts were observed in most ELMy H-mode plasmas for normal strike-point location.
LFS baffle
HFS baffle
domeLFS divertor
512x5121024x512
Viewing port edge
Dusts were ejected into inner (HFS) divertor (Case 3)Dusts were ejected into HFS divertor after ELMs:when the inner strike-point was at upper target (on carbon-deposition layers),
⇒ Part of deposition layers was removed by ELM heat and particle fluxes.
Plume of C- and D-ionization from dust along the field line.
512x512 (6k frame/s)
Dusts move outward and toroidally (ion drift direction).
Dust-2 from inner divertor moves in toroidal direction:Vt ~1m/4ms=0.25km/s
1024x512
ELM
This was not seen at main plasma edge: more examples will be necessary.
dust2
dust1
then, away from separatrix in radial direction, Vr~0.2m/4ms =0.05km/s
Shot#46953
4. Dust from outer divertor and influence on main plasma
However, large numbers of dusts were recently seen in some std. ELMy H-mode plasmas (PNBI=12-16MW): strike-point is located on toroidal W-band C5+ line at edge was enhanced by 3-5 times (but W was not enhanced).
Mostly C dusts were ejected from the outer target with small toroidal velocity(here, W ion was large in core region)
16 frames (8ms)
Dust movements from outer divertor (one of worst example)
Dust 1: moving to core, change the direction: Vr/p ~1m/17ms = 0.06km/s.Dust 2: moving to main plasmasublimation?: Vp ~1m/8ms =0.13km/s
outer (LFS) midplane
Divertor
NBI port (P12)
Bt, Ip
Port edge (P17)
CFC tile with ribbon-W (P14) was not seen by port edge
34 frames (17ms)
dust1
dust2
dust1
dust2
Dusts ejected from outer target move poloidally/radially (small friction?).
5. Summary
Trajectory and velocity of dust (emission) in ELMy H-mode plasmas were measured with a fast TV camera from tangential port.
In main chamber, number of dusts with various directions were observed, particularly after hard disruptions (large ΔWdia) and overnight GDC.
Many dusts were ejected from deposition layer near INNER strike-point after ELMs: large heat and particle loading enhance thermal expand/RES.
・ Velocity of nearly toroidal /poloidal movement (0.2-0.5km/s) was faster than that of nearly radial movement (0.03-0.06km/s).
・ The toroidal movement was mostly ion drift (Ip) direction: it is consistent with SOL flow measurement in HFS and LFS SOLs.・ The radial movement (at LFS divertor) was affected by ELM events: inward movement from LFS divertor changed to outward (after ELM).・ Dust ejection did not correlate to the main nC & Zeff. But influence of many
dust ejections on main plasma were recently observed.Future work and proposals: UEDGE/DUSTT simulation for dynamics study started by Kanazawa Univ. Measurement from different angles is required for accurate trajectory.