Flow Dynamics in RFXmod

1
Flow Dynamics in RFX-mod

• B.Zaniol, L.Carraro,E.Gazza,M.E.Puiatti,P.Scarin,M
  .Valisa
• Consorzio RFX, Associazione Euratom-ENEA sulla
  fusione

2
OUTLINE

• basic notions on spectroscopic flow measurement
• spectroscopic system in charge during first
  RFX-mod operations - spring/summer 2005
• first indications on ion flows
• future developments

3
carbon flow from doppler spectroscopy

• impurity flows from doppler spectroscopy on not
  fully-stripped impurity ions.
• on RFX-mod (like in old RFX), typical plasma
  parameters are compatible to easily measuring C2
  and C4 lines at 2296 Å and 2270 Å, respectively.
  These emissions come mainly from electron
  excitation processes.

4

• while waiting for operations with the diagnostic
  neutral beam

5
passive spectroscopy the nightmare of a line
integrated signal
two opposite lines of sight for self reference
Ip , B
optical fibres to spectrometer
6
passive spectroscopy the nightmare of a line
integrated signal
intensity
Ip , B
wavelength
7
C V emission _at_ 2271 Å mainly around 0.8 r/a
passive spectroscopy the nightmare of a line
integrated signal
C V 0.60.9 r/a
Emissivity profile 1-D model typical discharge Ip
600 kA ne 26E19 m-3 Te 200300 eV
8
passive spectroscopy the nightmare of a line
integrated signal
C III 0.90.97 r/a
C III emission _at_ 2296 Å mainly around 0.95 r/a
Emissivity profile 1-D model typical discharge Ip
600 kA ne 26E19 m-3 Te 200300 eV
9
passive spectroscopy the nightmare of a line
integrated signal
CT-spectrometer L 75 cm toroidal mirrors 1200
gr/mm grating 750 nm blaze 2D - CCD camera 6
channels _at_ 250Hz 16 bit - 90 QE _at_ 500nm
10
RFX carbon spectra need a multigaussian fit
C I
C III
C V
O II
11
RFX carbon spectra need a multigaussian fit

• KS4FIT
• developed by Manfred von Hellerman
  (Textor,Jet,AUG)
• based on Fortran Nag routines with an IDL wiget
  interface
• to be linked with CHEAP (CHarge Exchange
  Analysis Package) for impurity density
  calculation with beam operations
• CHEAP new version by Lorne Horton (AUG) , main
  features
• - IDL based
• - linked to ADAS
• - machine independent

C I
C V
12
Some results...
13
Typical C III and C V flows in RFX-mod
A
m/s
m-3
eV
14
carbon flow decreases with electron density
C4 flow
v m/s
ne m-3
- time averages on plasma current flat top of
600 kA discharges - error bars reflect the
signals variations on considered time intervals
15
carbon flow decreases with electron density
C4 flow
v m/s
ne m-3
flow dependence on density in the core
consistent with Harvey stochastic model
predicting an outwards directed Er-T/e(?n/n
?T/T)
16
electron density dependence during discharge
time evolution
C4 toroidal flow
start
v m/s
end
ne m-3
17
C4 flow not influenced by other plasma
parameters (f,ip,Te,mode energy)
18
C2 toroidal flow strongly affected by external
perturbation
16248
symbols
line
Toroidal angle of external m0 perturbation
more details on Gas Puff Imaging mesurements in
N. Vianello, tomorrow
19
C2 toroidal flow shows clear dependence on the
position of the external perturbation
C2 toroidal flow
20
In RFX-mod C2 and C4 toroidal and poloidal
flows are consistent with the past
toroidal flow is co- directed with Ip in the core
and opposite at the edge following ExB sign. E
predicted by theory and measured by Langmuir
probes outward directed in the core and inward
directed in the edge
Measured Er?rpi/niZe-(vi x B)r compatible with
stochastic model (Harvey) for C4 at ne
2.5E19 m-3 ? Er 2.5 kV/m
21
RFX-mod NBI ready in these very last days
beam atom energy 50 keV equivalent current up
to 2 A beam angular divergence 10-2
rad pulse duration 50 ms repetition
frequency 3 min
22
New optical heads allow a resolution of 5 cm
along minor radius
BEAM
OPTICAL HEADS
23
First beam emission spectrum before beam duct
E
Unshifted Ha line
E/3
intensity a.u.
E/2
E/18
wavelength nm
Beam optimization still in progress!
24
SUMMARY

• Apart from the electron density, no other clear
  evidence of a dependence of C4 flow on other
  plasma parameters has been found. This is
  consistent with what found on old RFX.
• C2 flow doesnt show any dependence on electron
  density nor on other plasma parameters. Like in
  old RFX, it flows opposite to plasma current and
  is strongly affected by the position of the
  external perturbation.
• Passive spectroscopy on line integrated signals
  is not the better way for studying the features
  and the time behaviour of ion flow profiles. With
  the help of the neutral beam injector we plan to
  obtain a better spatial and time resolution.