Study of B+1, B+4 and B+5 impurity poloidal rotation in Alcator C-Mod plasmas for 0.75 ≤ρρρρ≤ 1.0.
Igor Bespamyatnov, William Rowan, Ronald Bravenec, and Kenneth Gentle
The University of Texas at Austin, Fusion Research Center
Robert Granetz and Dexter BealsMIT Plasma Science and Fusion Center
Session QP1: Poster Session VI, Abstract: QP1.000562:00pm-5:00pm Wednesday, November 1, 2006
Philadelphia Marriott Downtown - Franklin Hall AB
48th Annual Meeting of the Division of Plasma Physicsof the American Physical Society
October 30–November 3 2006; Philadelphia, Pennsylvania
Abstract
Poloidal velocities of B+1, B+4 and B+5 impurity ions are measured in Alcator C-Mod
tokamak plasmas using charge exchange recombination spectroscopy (CXRS) for B+5
and ambient emission for B+1 and B+4. The set of 25 poloidal optical channels, 10
toroidal optical channels, modulated diagnostic neutral beam and fast Roper CCD
camera allow 2mm poloidal spatial resolution in the region 0.75 < r/a < 1.0 and 10 mm
toroidal spatial resolution with 13 msec temporal resolution at all times during the 1.5
sec plasma pulse. The variation in the poloidal rotation as the plasma transitions from
ohmic to L- to H- mode will be described. Implications for Er will be discussed. Data
for EDA H-modes and ITB discharges will also be presented. The emphasis of this
work is on comparing the poloidal rotation of B+1, B+4 and B+5 impurities, cataloging
t h e e f f ec t s o f d i f f e re n t p l a sma mod e s a n d f i na l ly o n a t t empt ing
to understand the poloidal rotation based on neoclassical theories.
Supported by USDOE grant DE-FG03-96ER54373
and Coop. Agreement DE-FC02-99-ER54512.
Motivations
• Poloidal rotation has been shown to deviate significantly from neoclassical expressions in DIII-D, TFTR and in JET
- R. E. Bell, Physical Review Letters 81, 7, 1429 (1998)
- K. Crombe, Physical Review Letters 95,155003 (2005)
- K. Crombe, EPS conference on Contr. Fusion and Plasma Phys. St. Petersburg, 7-11 July 2003 ECA Vol. 27A, P-1.55
- W.M. Solomon, Physics of Plasmas 13, 056116 (2006)
- G. M. Staebler, Nuclear Fusion, Vol. 41, 7, 891 (2001)
• JET proposes that large Vp deviation may be due to an effect of heating beams.
• JET observed the largest excursions of Vp from neoclassical values temporally coincide with the reduction in turbulence fluctuations which occurred 250-450 ms before the ITB formation.
• TFTR observed a vast Vp overshoot (precursor triggering the transition) during the enhanced reversed shear (ERS) discharges.
• TFTR also found a sudden change in the Vp during Li pellet injection which relaxed to the initial value within τ ~ 0.5 sec.
• DIII-D made a comparison of Vp with the neoclassical theory in the steady state portion of thedischarge assuming that it is in equilibrium with no consideration of possible Vp relaxation from previous transition.
• JET observed ta Vp relaxation (τ ~ 0.5-1 sec) after the transition to ITB
• It has been conjectured that drift waves may be responsible for sudden changes of the Vp during the plasma transitions.
Summary and conclusions
• In C-Mod,
– The impurity density must be inferred from CXRS measurements using cross sections which take into account high density and multiple beam components
– Cross section effects on poloidal rotation measurements are small
– Poloidal rotation compares favorably to neoclassical predictions of Kim.
– After H-mode transitions, the rotation is apparently driven away from the neoclassical value and then relaxes to the value on a long time scale.
– There is no heating beam to complicate analysis
• Transiently, the Vp can be significantly larger (one order of magnitude) than the neoclassical theory predicts but Vp settles to the neoclassical value in a quiescent discharge.
• The approach to neoclassical equilibrium is examined with an empirical model in which the time derivative of the density is employed as a surrogate for the “other physics”.
• The time scale of the Vp relaxation (τ ~ 0.4 sec) is consistent with TFTR and JET results.
• The Vp plays an important role in fusion plasmas confinement, including the Er×B shear suppression of turbulence, stabilization of the plasma instabilities and the formation of the internal transport barrier (ITB). For more details on Er please see the poster GP1.00029 (R. Bravenec, Gyrokinetic Microstability Analysis of the Inner Boundary of the H-mode Pedestal) and invited talk GI1.00001 (A. Hubbard, H-mode pedestal and threshold studies over an expanded operating space on Alcator C-Mod)
Analysis Physics
• The analysis employs a complete set of fine structure components for the CXRS transition.
• Cross section effects due to view components parallel to the beam are negligible as the views are essentially perpendicular to the beam.
• Cross section effects due to finite lifetime of the excited states are negligible due to the low ion temperature.
• Ion temperature and measured impurity temperatures are the same due to collisionality.
• The density of B+5 is based on detailed analysis of excitation cross sections and on a combination of in-situ and post campaign calibration procedures.
Charge Exchange Recombination Spectroscopy (CXRS)
•CXRS emission (active charge
exchange), 4944.6 hydrogen- like B+4
(n=7→6) (visible spectrum) for
measurements of nB , TB, VB
•Blended with a line from a lower
ionization stage 4940.38,
(2s3d 1D2 → 2s4f 1F3) B+1
•Blended with 4944.6 (n=7→6) B+4
excited by thermal charge exchange
and electron collisions
•Accompanied by line 4950.77
(n=11→8) B+4
•Red spectrum: acquired during
DNB pulse
•Blue spectrum: acquired just after
the DNB was turned off
•Red - Blue = CXRS enhancement
( ) +−++ +→+ HnAAH iqq )1(0
R=0.847 m
(4940.38) 2s3d 1D2 →2s4f 1F3, B+1
(4944.67) (n=7→6) B+4
EB=50 keV
(4950.77) (n=11→8) B+4
( ) ( ) νhnAnA fq
iq +→ −+−+ )1()1(
Neoclassical prediction of Vp of B+5 impurity
• The neoclassical theory employed here is described in “Neoclassical poloidal and toroidal rotation in tokamaks,” Y. B. Kim, P. H. Diamond, and R. J. Groebner, Physics of Fluids B 3, 2050 (1991).
• The poloidal component of the velocity is given by
• TI and rotation velocity is derived from width and shift of CXRS spectral lines in the usual way
• The absolute density for B+5 is described here in some detail as it is a recent development
2
21
1111
2
3
B
B
LZ
Z
LLZ
ZKK
Z
Tv t
nII
i
niTII
i
i
iIp
+−
+−+=
profilesBnTfromcalculatedKandK
arrayVBfromZandnoftsmeasuremenTSn
tsmeasuremenCXRSvnT
litycollisionahighplasmadensityhighTT
profilesnnToflengthsscaleinverseLLL
ii
effei
IpII
Ii
IiInIniTI
,,
,,
,
,,,,
21 −
−
−
−≈
−
Finding the absolute boron density
• Absolute density optical measurements are always complicated and usually require in-vessel calibration of the full optical path. The more frequent the calibration is made the more suitable the calibration result for the following experiments. During the plasma discharges the optical parts can be displaced and darkened which makes the preceding calibration obsolete.
• This is a complex procedure for large tokamaks which are usually opened for in-vessel work once or twice a year.
• Another way is proposed here for shot to shot calibration measurements. It employs the bremsstrahlung background of each optical channel to calculate the given calibration coefficient.
• The method uses the Thompson scattering ne, te and VB array data to predict the bremsstrahlung background for each spatial channel.
• The validity of this method was confirmed by absolute in-vessel calibration made on 08/10/06.
• In order to acquire the density of B+5 one may need to know the CXRS effective emission rate coefficients for proper transition. This non-trivial task includes the calculation of the neutral beam propagation and excitation, atomic model of photon emission and correction of CXRS cross sections.
� I.O. Bespamyatnov, W.L. Rowan, Effects of neutral-beam excited states on charge-exchange emission cross sections, Review of Scientific Instruments 77, 10F123 2006
The bremsstrahlung calibration # 1060726031 (1.1 sec)
B+5 density
The bremsstrahlung calibration # 1060726031 (1.0 sec)
B+4 density
The scope of the plasma shot # 1060726031
Neoclassical comparison, # 1060726031
□ - CXRS measurements
▬ - Neoclassical prediction□ - CXRS measurements
▬ - Neoclassical prediction
□ - CXRS measurements
▬ - Neoclassical prediction□ - CXRS measurements
▬ - Neoclassical prediction
time
Local ne comparison and Vp fits, # 1060726031
- ne
◊ Vp
- Vp fit
sec4.0
4.0
0sec0
=
=
=
τ
α
kmV
sec1 km
eqV =
sec1 km
eqV =
sec1 km
eqV =
sec1 km
eqV =
B+1 and B+4 poloidal velocities (chord averaged), # 1060726031
edge chord
internal chord
edge chord
internal chord
B+1 and B+4 poloidal velocities # 1060726031 and 1060726024
Conclusions:
• B+4 poloidal velocity undergoes the similar increase as B+5 poloidal velocity
• B+1 poloidal velocity also increases during the transition and relaxes to the initial level rapidly during the final part of the transition. The formation of the edge pedestal may control the B+1 velocity evolution since B+1 emits mainly from the region outside the separatrix and edge pedestal position will determine the behavior of the edge, cold, low-charge-state impurities.
edge chord
internal chord
The scope of the plasma shot # 1060726022
Neoclassical comparison, # 1060726022
□ - CXRS measurements
▬ - Neoclassical prediction□ - CXRS measurements
▬ - Neoclassical prediction
□ - CXRS measurements
▬ - Neoclassical prediction□ - CXRS measurements
▬ - Neoclassical prediction
time
Local ne comparison and Vp fits, # 1060726022
- ne
◊ Vp
- Vp fit
sec4.0
4.0
0sec0
=
=
=
τ
α
kmV
sec1 km
eqV =
sec1 km
eqV =
sec2 km
eqV =
sec0 km
eqV =
The scope of the plasma shot # 1060525013
Neoclassical comparison, # 1060725013
□ - CXRS measurements
▬ - Neoclassical prediction□ - CXRS measurements
▬ - Neoclassical prediction
□ - CXRS measurements
▬ - Neoclassical prediction
□ - CXRS measurements
▬ - Neoclassical prediction
time
Model for temporal evolution of Vp
• It was found that poloidal velocity doesn’t follow the neoclassical equilibrium solution but strongly depends on temporal transients of the local electron density.
• It may well be that it is not electron density transients that drive the velocity but some different process which drives both density and poloidal velocity similar in ways.
• In addition it was observed that poloidal velocity undergoes an equilibration process which pushes the velocity to the neoclassical equilibrium solution.
• The first attempt was made to produce the simple poloidal velocity model which implies all preceding arguments.
• Only 4 parameters (V0, Veq, α, τ) describe the model. Some of them may be similar for typical plasma discharges.
{ }0)0( VV
VV
dt
dn
dt
dV eqpep
=
−−=
τα
timeionequilibrat
velocitymequilibriuV
tcoefficiendriving
velocityinitialV
eq
−
−
−
−
τ
α
0
The scope of the plasma shot # 1060726024
The scope of the plasma shot # 1060524018 (reversed B)
Neoclassical comparison, shots # 1060726024, 1060524018
□ - CXRS measurements
▬ - Neoclassical prediction
□ - CXRS measurements
▬ - Neoclassical prediction
Conclusions:
• There is noticeable correlation between the Vp and the local ne.
• Equilibration of the Vp in the absence of the ne transitions is also observed.
• Multiple transitions on shot 1060726024 do not allow for Vp to equilibrate.
• There is no such vivid correlation for reversed field plasma discharges.
Local ne and Vp comparison, # 1060726024
- ne
◊ Vp
- Vp fit
sec4.0
4.0
0sec0
=
=
=
τ
α
kmV
sec2 km
eqV =
sec1 km
eqV =
sec2 km
eqV =
sec0 km
eqV =
Summary and conclusions
• In C-Mod,
– The impurity density must be inferred from CXRS measurements using cross sections which take into account high density and multiple beam components
– Cross section effects on poloidal rotation measurements are small
– Poloidal rotation compares favorably to neoclassical predictions of Kim.
– After H-mode transitions, the rotation is apparently driven away from the neoclassical value and then relaxes to the value on a long time scale.
– There is no heating beam to complicate analysis
• Transiently, the Vp can be significantly larger (one order of magnitude) than the neoclassical theory predicts but Vp settles to the neoclassical value in a quiescent discharge.
• The approach to neoclassical equilibrium is examined with an empirical model in which the time derivative of the density is employed as a surrogate for the “other physics”.
• The time scale of the Vp relaxation (τ ~ 0.4 sec) is consistent with TFTR and JET results.
• The Vp plays an important role in fusion plasmas confinement, including the Er×B shear suppression of turbulence, stabilization of the plasma instabilities and the formation of the internal transport barrier (ITB). For more details on Er please see the poster GP1.00029 (R. Bravenec, Gyrokinetic Microstability Analysis of the Inner Boundary of the H-mode Pedestal) and invited talk GI1.00001 (A. Hubbard, H-mode pedestal and threshold studies over an expanded operating space on Alcator C-Mod)