Upload
eustace-rice
View
220
Download
3
Tags:
Embed Size (px)
Citation preview
J. Boedo, UCSD
Fast Probe Results and Plans
By J. Boedo
For the UCSD and NSTX Teams
J. Boedo, UCSD
Probe Installed and Commissioned
11 deg angle
6.8”
Location is 6/8” below midplane
J. Boedo, UCSD
New head and tips to accommodate curvature and heat flux
Thermal ConductivityThermalgraph Tips (parallel) 800 W/mKCopper 401 W/mKPOCO graphite 72 W/mK
J. Boedo, UCSD
IsatVf1
Vf2Epol
DoubleProbe
Vf3
Mach1
Mach2
Er
Vf4
Te fluct
Existing Probe Head has 10 Tips
•Tips in blue will be active on day one, the rest implemented as upgrades (if funded)•Fluctuations to 1 MHz•Two Vf tips used for Epol (and fluctuations)•Two Vf tips used as Er (and fluctuations) >> Reynolds Stress•One tip as Isat >> ne•Two tips as double probe (Te and Ne profiles)
J. Boedo, UCSD
Data Taken at High Spatial and Temporal Resolution
-5.0 100
0.0 100
5.0 100
1.0 101
1.5 101
2.0 101
0.05 0.1 0.15 0.2 0.25
sp109033
Radius
Time
~ 100 ms
~ 30 ms
Insertion completed in 100 ms. ~30 ms in plasma
Time resolution for Te and Ne is 1 ms
Time resolution for fluctuations is 1 micro sec
Spatial resolution is 1.5-2 mm (tip size + probe motion)
J. Boedo, UCSD
0.0 100
1.0 1012
2.0 1012
3.0 1012
4.0 1012
5.0 1012
-2 0 2 4 6 8 10 12
109052 H 4.2 MW λ=1.51 cm109035 2.0 H MW λ=2.55 cm109033 2.0 L MW λ=3.25 cm
Radius
Within the limited shot space, density and power were varied. So scaling is inconclusive
So far:
decreases in H mode with power 2.5>1.5 cm
Will obtain wider operating space in the future and produce scalings
€
λn
J. Boedo, UCSD
H-mode Radial Electric Field Flat in SOL
-80
-60
-40
-20
0
20
40
-5 0 5 10 15
Er
(V/cm)
R-Rsep
(cm)
H-Mode
L-Mode
J. Boedo, UCSD
Fluctuation Quantities to Characterize in NSTX
density fluctuations:
€
n
~
n
=
Isat
~
Isat
+
Te
~
2 Te
plasma potential fluctuations
€
φpl
~
= φfl
~
+ α Te
~
cross-field particle flux:
€
Γturb
= ⟨ n
~
Eθ
~
⟩ Bφ
convective heat flux:
€
Qconv
=
5
2
Te
Γturb
conductive heat flux:
€
Qcond
=
3
2
⟨ Te
~
Eθ
~
⟩ Bθ
− Te
Γturb
Need to add Te fluct capability
J. Boedo, UCSD
Harmonic Diganostic
250 WRF amp
Pearsoncoil
Probe
AmpR1
V_fast
V
Ifdrive= 400 kHz
AmpAmp
Amp
400 kHzBandpass
filter
800 kHzBandpass
filter
Ampli-tude
detector
I2
I_fast
Te_slow
TunableLPF
Compa-rator
Analogmultiplier
Analogdivider
Ampli-tude
detector
Ampli-tude
detector
I /I2
Amp
Analogmultiplier
Compa-ratorAdder
RefRef
V_min I_minDummy
transmissionline
R2C3C1
C2
J. Boedo, UCSD
Harmonics Technique
Current to a DC-floating probe driven by sinusoidal voltage
can be expressed as a series of sinusoidal harmonics [1]:
where - amplitude of mth harmonic
- Bessel functions of integer order k
For eU0/kTe << 1:
Thus Te can be determined from the ratio
of the amplitudes of 1st and 2nd harmonics
The error of this approximation for eU0/kTe= 1 is only about 5% [1]
ekT
eU
ekTeU
2I 01
0 ≈⎟⎠⎞⎜
⎝⎛
20
2 81
I 0⎟⎟⎠
⎞⎜⎜⎝
⎛≈⎟
⎠⎞⎜
⎝⎛
ekTeU
ekTeU
2
0
2
104I
I4 I
IeUeUkTe =≈
[1] Boedo et al, Rev. Sci. Instrum. 70 (1999), 2997
)cos()cos(II
2 00
0 110
tmItmI
IekT
eU
ekTeU
ekTeU m
mm
msi
pr ⎟⎠⎞⎜
⎝⎛∑=∑ ⎟
⎠⎞⎜
⎝⎛
⎟⎠⎞⎜
⎝⎛
=∞
=
∞
=
⎟⎠⎞⎜
⎝⎛⎟
⎠⎞⎜
⎝⎛=⎟
⎠⎞⎜
⎝⎛
ekTeU
ekTeU
ekTeU
msim II 0000II2ω
)(I zk
J. Boedo, UCSD
Need to Include Magnetic component
⎟⎟⎠
⎞⎜⎜⎝
⎛ ×∇−⎟⎟
⎠
⎞⎜⎜⎝
⎛ ×∇−=
22
~~
BBnm
BBì
φφt
⎥⎥
⎦
⎤
⎢⎢
⎣
⎡+
×≡
B
q
B
p BBEQ
~~~~
2
3~ ||
2
B
vn
B
n BBÃ
~~~~~ ||
2+
×∇−=⊥
φParticle Flux
Reynolds Stress(neglecting ion pressure
fluctuations)
Heat Flux
B
p
B
j r
J
BBÃ
~~~~~ ||
2
||
|| +×∇
−=φParallel
Current Flux
⊥= BÃ~~~
φKHelicity Flux
J. Boedo, UCSD
Normalized N fluctuations very large
0.0
0.5
1.0
1.5
-1 0 1 2 3 4 5 6 7
Position
J. Boedo, UCSD
Radial Turbulent Flux Lower in H-mode
-7.0 1017
-6.0 1017
-5.0 1017
-4.0 1017
-3.0 1017
-2.0 1017
-1.0 1017
0.0 100
1.0 1017
-5 0 5 10 15 20
-Γr (cm-2 s -1 )
- H mode-L mode
Position
Value is ~3-5 1017 cm-2 s-1
J. Boedo, UCSD
Intermittent Plasma Events Pervasive
J. Boedo, UCSD
Power Spectrum Comparison
10-5
0.0001
0.001
0.01
0.1
1
0 20 40 60 80 100
Power spectrum vs f (in kHz) from probe data during gas injection (solid circles) and without(x) and comparison to that obtained form the GPI system (open circles). Note that the power spans y 4 orders of magnitude indicating excellent S/N ratio.
J. Boedo, UCSD
Intermittency (Radial Convective Transport) Lower in H-Mode
L-Mode 109033 H-Mode 109052
Intermittency is very strong in NSTX!
J. Boedo, UCSD
Intermittency (Radial Convective Transport) Lower in H-Mode
L-Mode 109033 H-Mode 109052
Intermittency is very strong in NSTX!
J. Boedo, UCSD
L-Mode 109033DL >
DL >
=I DL >
Density is highly intermittentRate of bursts is ~1E4Poloidal field is much less intermittentCoherent modes show in EthetaInstantaneous flux is 1-3E15 cm-2s-1
J. Boedo, UCSD
H-Mode 109052=I DL >
=I DL >
=I DL >
Density is intermittent (ELMs?)Rate of bursts is 2E3Poloidal field is much less intermittentCoherent modes show in EthetaInstantaneous flux is ~2E15 cm-2s-1
J. Boedo, UCSD
Initial Reynolds Stress and Bicoherency Measurements
€
∂uθ∂t
+∂uruθ∂r
=−μuθ
Codes Written and Tested
H-Mode
L-Mode
J. Boedo, UCSD
Bispectrum
H-mode L-Mode
f3=f1+f2
f3=f1-f2
J. Boedo, UCSD
Bicoherency
H-mode L-Mode
J. Boedo, UCSD
Profile Conclusions
Edge/SOL profiles available with high spatial (2-2.5 mm) resolutionLimited to 1086xx, 1089xx, 109032-109062Fully calibrated data for 109032-xxx062
NSTX profiles seem different from larger aspect ratio devicesTe profile in the edge/SOL is flat at 25-30 eVNe profile has a very long decay length. Reduced by power!H-mode Er profile flat in SOL
How do flat Te profiles are maintained?Do we have (more) anomalous radial transport? (intermittency)Adding fast Te diagnostic
J. Boedo, UCSD
Future Work and Goals
Data obtained in L and H-mode. SOL length scales with power. Ongoing
Many puzzles need addressingFeatures usually associated with LCFS vicinity are different in NSTX, such as a potential well, steep H-mode profiles, etc Ongoing
Intermittency is quite strong in NSTX (very strong radial transport?)
Need to quantify electrostatic turbulence and Intermittency (enough to explain strong transport?) Codes just ported for electrostatic turbulence, ongoing for intermittency!
Bicoherency and Reynolds Stress codes just finished. Investigate energy cascading and self-organized flows
Need to quantify electromagnetic components of transportUpgraded head partially designed and to be built in FY04
J. Boedo, UCSD
J. Boedo, UCSD
J. Boedo, UCSD