Transcript Slide 1

Overview of Globus-M Spherical Tokamak Results
V.K.Gusev, B.B.Ayushin, F.V.Chernyshev, I.N.Chugunov, V.V.Dyachenko,
L.A.Esipov, D.B.Gin, V.E.Golant, N.A.Khromov, S.V.Krikunov, G.S.Kurskiev,
M.M.Larionov, R.G.Levin, V.B.Minaev, E.E.Mukhin, A.N.Novokhatskii, M.I.Patrov,
Yu.V.Petrov, K.A.Podushnikova, V.V.Rozhdestvensky, N.V.Sakharov,
O.N.Shcherbinin, A.E.Shevelev, S.Yu.Tolstyakov, V.I.Varfolomeev, M.I.Vildjunas,
A.V.Voronin
Ioffe Physico-Technical Institute, RAS, 194021, St. Petersburg, Russia
E-mail: [email protected]
V.G.Kapralov , I.V.Miroshnikov, V.A.Rozhansky, I.Yu.Senichenkov, A.S.Smirnov,
I.Yu.Veselova
St.-Petersburg State Polytechnical University, 195251, St. Petersburg, Russia
S.E.Bender, V.A.Belyakov, Yu.A.Kostsov, A.B.Mineev, V.I.Vasiliev
D.V. Efremov Institute of Electrophysical Apparatus, 196641, St. Petersburg, Russia
E.A.Kuznetsov, V.N.Scherbitskii, V.A.Yagnov
SSC RF TRINITI, 142192, Moscow region, Russia
A.G.Barsukov, V.V.Kuznetsov, V.M.Leonov, A.A.Panasenkov, G.N.Tilinin
NFI RRC “Kurchatov Institute”, 123182, Moscow, Russia
E.G.Zhilin
Ioffe Fusion Technology Ltd., 194021, St. Petersburg, Russia
Presented at 11th IST Workshop, 11-13 October, 2006, Chengdu, PR China
Outline
1.
Density limits in Globus-M: Technology and scenario,
features of high density regimes (MHD etc.)
2.
NB heating: Ions heating – beam specie AM influence,
temperature measurement, beam thermalization
Electrons heating – dependence on density and Ebeam
3.
ICR heating at fundamental harmonic: General features,
energy exchange, antenna spectrum, simulation and experimental
result of CH dependence, fast particles (tail) temperature and
fraction
4.
Plasma jet injection with high velocity: Density control by
injection at stationary stage, plasma jet penetration through
magnetic field, penetration into hot plasma – experiment and
simulations, discharge initialization by plasma jet
5.
Summary
Globus-M parameters
Parameter
Designed Achieved
Toroidal magnetic
field
0.62 T
Plasma current
0.3 MA
Major radius
0.36 m
Minor radius
0.24 m
Aspect ratio
1.5
Vertical elongation2.2
Triangularity
0.3
Average density 11020 m-3
Pulse duration
200 ms
Safety factor, edge4.5
Toroidal beta
25%
0.55 T
0.36 MA
0.36 m
0.24 m
1.5
2.0
0.45
1.21020 m-3
130 ms
2
~10%
ICRF power
frequency
duration
1 MW
0.5 MW
8 -30 MHz 7.5-30 MHz
100 ms 30-80 ms
NBI power
energy
duration
1.3 MW
30 keV
30 ms
1 MW
30 keV
30 ms
Improvement in vacuum technology, plasma
control, and experimental scenario gave
synergetic effect in density limit increase
Until last year high B/R ratio potential of Globus-M was not realized
Oil free pumping
Vertical plasma control improvement
Belt limiters
made possible to operate routinely with high plasma currents (up to 0.25MA)
with the small gap 3-4cm between plasma and vessel wall
High density OH operating without current
stabilization
NBI 0.55 MW
Listed above steps and:
• Careful
wall
conditioning
and
boronization improved density control by
inner wall gas puff (contribution of the
walls could be neglected)
• Experiment scenario, when high density
shot was followed by several low density
shots to prevent wall saturation by
deuterium.
Results:
•
Stable operating at high average
densities in the target OH regime.
• Near the density limit no radiation
collapse – current degradation
• Line average densities
<ne> ~ 1-1.21020 m-3 were achieved,
(n/nG)~1
Gusev NF 46 7 2006
High density OH operating with current
stabilization
Listed above steps and:
• Density control by inner and outer
wall gas puff (contribution of the
walls could be neglected).
• Plasma current feedback
stabilization
Results:
• Stable operating at high average
densities in the target OH regime.
• Radiation collapse at density limit
(n/nG)~1 was achieved
Petrov 33 EPS conf Roma 2006
Top
Features of high density OH operating with
current stabilization
Bottom
1
0
144
145
146
Time (ms)
147
148
Top
143
Bottom
1
0
163
164
165
166
Time (ms)
167
168
Density limit depend neither on the gas puff position (inner – outer wall) nor on
gas puffing rate (high, moderate, small)
Main MHD instability is saw-tooth oscillations, amplitude and period increase
with gas puffing rate. Saw-tooth seems not restrict the density limit
Sometimes benign m=1/n=1 tearing mode (snake) develops ( no saw-teeth in
this case), but does not restrict the density limit
Operational space of Globus-M increases
due to higher densities
Achieved is n/nG~1 and
approached is n/nMur~1
Average densities up to
1.21020m-3 obtained at
0.4T
Radiation collapse at
density limit restrict
further density rise
Density limit is easily
accessed at lower plasma
currents
Operational space of Globus-M increases
due to higher densities and temperatures
2 1
(1-(r/a) )
2 1.7
(1-(r/a) )
19
1000
2.0x10
#13802
155ms
19
1.5x10
ne, m-3
Te, eV
800
600
400
Max electron
temperature
Te(0)  0.95 keV
is achieved at
<ne>~1.51019m-3
19
1.0x10
18
5.0x10
200
0.0
0
0.2
0.2
0.4
R, m
2 1.5
(1-(r/a) )
300
0.3
0.3
0.4
R, m
2 1.1
Max electron density
Ne(0)  1.5 1020m-3
is achieved at
<Te>~175 eV
(1-(r/a) )
20
2.0x10
#13727
160ms
20
200
ne, m-3
Te, eV
1.5x10
20
1.0x10
100
19
5.0x10
0
0.2
0.3
R, m
0.4
0.0
0.2
0.3
R, m
0.4
NBI heating in Globus-M
400
nH/(nH+nD) = 20%
TD (eV)
300
200
NBI
100
130
140
150
160
170
180
t (ms)
Principal difficulties due to small size of the target plasma compared to the
beam dimensions and small size of the vacuum vessel compared to fast
particle orbit extent - in Globus-M RL  a/2 for 30keV deuterons at outboard
edge . Moreover in Globus-M plasma is tightly fitted into the vacuum vessel.
Ion NB heating efficiency weakly depends on of the beam specie AM. D-beam
is slightly more effective due to lower atoms velocity at the same beam
energy. Minaev 33 EPS conf 2006 Roma
Globus-M
2004.06.21 (#9191)
t = 136.5 ms
NBI ion heating in Globus-M
600
t = 156 ms
TD
OH
187 eV
H-NBI
376 eV
D-NBI
393 eV
10
3/2
10
9
10
8
16947 NBI
16991 NBI
500
400
Ti (eV)
10
11
2
(eV cm ster s)
-1
10
300
cx / E
0.5
200
Thermalized
particles
10
NBI
slowing down
particles
7
0
1
E (keV)
2
100
0
120
130
140
150
160
170
t (ms)
Ion temperatures measured by NPA in principle coincide with first CHERS
measurements at densities < (3-3.5)х1019m-3, for higher densities correction
for plasma opacity should be done.
t = 136.5 ms
NB thermalization in Globus-M
Thermalized
particles
10
8
10
6
1/2
cx / E
NBI T i=357 eV
(E
E b /3
10
3/2
3/2 -1
+ Ecr )
Beam slowing down
particles (trapped)
3/2
2
(eV cm ster s)
-1
10
OH T i=197 eV
10
E b/2
E cr
Eb
4
0
5
10
15
E (keV)
20
25
“Perpendicular’ spectrum of fast particles, measured by 12 channel NPA.
Above Ecrit (12-22.5 keV) electron drag predominates, pitch angle scattering is poor.
Below Ecrit (0–12 keV) collisions with plasma ions provides pitch angle scattering.
Specrtum coincides with Fokker-Plank predictions ( no losses).
Beam ion slowing down is well described by classical Coulomb scattering theory and
the particle losses at least in the energy range below Ecr are insignificant.
NBI heating at high densities in Globus-M
Ion heating vanishes at high densities at beam energy 25keV, (low
P=0.5MW, opaque plasma), contrary electrons heating is improved
with density rise and increase of NB energy and power. Agree with
ASTRA simulations
Ion cyclotron resonance heating experiments
and simulations in Globus-M
Showed that: Ion heating is effective in the low frequency range (fundamental
IC resonance for protons as “minority” in deuterium plasma). At low RF power
input ~0.3POH – Ti increases two times.
Energy exchange between plasma components is classical, i.e. deuterons is
mostly heated through energy exchange with protons. Ion energy confinement
is neoclassical, or even better (ASTRA gives 0.7 χNEO) [Shcherbinin,…Leonov NF 46 7
2006]. Electron heating is small at such power level.
Ion cyclotron resonance heating in STs
has specific features
•Simultaneous existence of several IC harmonics in the plasma crosssection.
•Low width of resonance absorption layers (much smaller than excited
wave length) and lower efficiency of single-pass absorption of FMS waves.
•Wave propagation similar to a resonator kind, when the whole tokamak
vessel plays the role of a multimode resonator of low quality.
•Significant non-resonance absorption ( high plasma dielectric constant).
Prad, arb.un
Spectrum of single loop antennae in
Globus-M
-200
-100
0
Nz
100
The peaks correspond to resonator
modes excited in the chamber. Short
wavelength components (lNzl~150) are
strong enough.
Dashed - the idealized spectrum if all
excited waves are completely absorbed in
the plasma without any reflection from
inner plasma layers.
Shcherbinin NF 46 7 2006
The elimination
of ICRH
secondon
harmonic
of
Specific
Features of
ST
hydrogen resonance improves ICR heating
ωH
2ωD 3ωD
ωH
2ωD
2ωH
2D, 1H
40
3D
0
1D
y, cm
20
-20
-40
Several Resonances
9MHz, 4T
-20
0
One
20
x, cm
7,5 MHz, 4T
Resonance
Simulated RF Energy Absorption Profiles
dP/dr, arb.un.
CH=10%
|Nz|<150
-20
-10
0
10
CH=10%
|Nz|<20
20
-20
-10
r,cm
0
dP/dr, arb.un.
-10
0
r,cm
20
r,cm
CH=50%
|Nz|<150
-20
10
10
CH=50%
|Nz|<20
20
-20
-10
0
10
20
Left - calculated absorption for
the whole excited wave
spectrum (|Nz|≤150).
Right - calculated absorption
for the narrow part of wave
spectrum (||≤20).
Green – electrons (TTMP,
Landau damp)
Red – protons (cyclotron)
Blue – deuterons (cyclotron,
Bernstein wave abs)
Shcherbinin NF 46 7 2006
r,cm
High absorption at r  - 5cm – for protons at high CH is the
consequence of short wave part of the exited spectrum
Ion heating improves with
H-concentration increase
B0 = 0.4 T, f = 7.5 MHz, Pinp =
120 kW, ne(0) ≈ 3.1019m-3,
IP = 195 – 230 kA.
The 2nd H-harmonic is absent
in the plasma volume.
CH increases from 10% to 70%
Sensible ICRH efficiency
improvement with increase of
hydrogen fraction may be
explained by strong short
wavelength component of
antennae spectrum
Triangles – deuterons
Circles – protons
Fast proton population approximately
constant in the wide range of Hconcentration during ICRH
The effective hydrogen “tail”
temperature (measured in the
energy range 1.7 – 4 keV)
decreases with CH rise.
The ratio of the tail proton
concentration to the thermal
proton concentration drops
sharply with increase of CH
(see the Table).
CH
15 % 25 % 35 % 50 % 60 % 70 %
Ntail/Ntherm 15,7% 11,4% 11,5% 7 % 4,8 % 3,5 %
The total quantity of fast
proton population in the
plasma remained
approximately the same in the
course of experiment.
Plasma jet injection with double
stage plasma gun
Plasma gun
Vacuum
shutter
Jet parameters:
• density up to 1022 m-3
• total number of accelerated
particles - (1-5)1019
• flow velocity of 50-110 km/s
Shot parameters:
• Bt=0.4 T,
• Ip= 0.2 MA
• initial central electron density
~ 31019 m-3.
Criterion of penetration through
magnetic field :
ρV2/2 > BT2/2μ0
Plasma jet injection into steady state
discharge period
220
110
0
2
1
0
Plasma current, kA
Gun current
D-alpha, V
Plasma Jet
Gas Jet
14
12
8
4
15
12
9
Injection from the equatorial plane,
along the major radius from the low
field side. The distance between the
plasma gun output and plasma was ~
0.5 m
Shots 12968, 12975
-2
Line integrated density, 10 cm , R = 24 cm
14
Magnetic field at the center of the
vessel was 0.4 T
-2
Line integrated density, 10 cm , R = 42 cm
The jet speed 110 km/s.
The jet density 2×1022m-3 at the gun
edge.
O III (559 nm), V
0,50
0,25
0,00
Comparison with low speed (~2 km/s)
gas jet injection
C III (465 nm), V
0,50
0,25
0,00
Mirnov signal, a.u.
2,5
0,0
-2,5
144
146
148
150
Time, ms
152
154
156
Time constant of density increase with
plasma jet injection is much smaller,
than with gas jet. Discharge is not
disturbed
Plasma jet penetration through the
magnetic field
If ρV2/2 < BT2/2μ0, how the plasma jet
penetrates through the magnetic
field?
Jet pressure, Atm
0,5
Study of jet penetration between poles
of DC magnet of 0.3 T induction. The jet
specific kinetic energy at the velocity of
75 km/s is less than magnetic pressure
of 0.3 T field.
0,4
0,3
0,2
0,1
Position of
gun edge
Position of
The pressure signal varies as the
pressure
detector magnet is moved from the gun towards
detector from zero level ( the jet is
0,0
blocked) to full pressure (~0.5 Atm – the
0 10 20 30 40 50 60 70 80 90 100 jet passes freely). Time-of-flight
Magnet position, cm
recombination of dense cold (1 eV)
plasma jet into the jet of neutrals with
Outside plasma - Time-of-flight
the same density and velocity makes
recombination into dense neutral
the jet insensitive to the magnetic field.
jet occur
Plasma jet penetration into the plasma
0
5
15
20
1.5
2.0
R (cm)
100
Te |z=0 (eV)
10
10
1
0.1
0.0
0.5
1.0
t (s)
The TS measurements (left) show the jet penetration deep in plasma [Gusev NF 46 7
2006]
Simulations inside plasma show – ionization of jet by hot electron influx and
braking due to emission of Alfven waves (ionization is very fast -0.5mks). grad B
drift accelerates jet towards LFS. Resulting effect allows the jet deposition beyond
separatrix, unlike the case of molecular supersonic beam, which is definitely
deposited outside separatrix [Rozhansky 33 EPS 2006 Roma]
Discharge initialization by plasma jet
injection
90
Plasma current, kA
Shots 15064, 15065
Preionisation by
plasma jet
60
Injection at maximum loop voltage
(UHF preionozation and prefill of the
vacuum vessel are off).
a
Preionisation by
magnetron
30
b
0
1,8
D-alpha, V
b
Number of injected particles is
comparable with total number of the
particles in tokamak (5×1018 –
1×1019).
0,9
a
Plasma current ramps up faster than
with traditional method.
0,0
0,50
C III (465 nm), V
a
0,25
b
0,00
112
113
114
115
116
Time, ms
117
118
119
120
D-alpha and CIII start earlier.
Plasma current is higher which
confirms more intensive plasma
heating at the initial stage of the
discharge.
Gas generator of the plasma gun
modification
Coaxial
accelerator
C1
C2
b
Grid filter
Jet flow
Dependences of number hydrogen
molecules on shot number
Titanium
hydride
grains
C2
Grid filter
Coaxial
accalerator
a
C1
a
Jet flow
Titanium
hydride
grains
Double stage plasma gun
A.V. Voronin 21 IAEA FEC 2006
Two versions of gas generating
stage:
a- fresh grains loaded before
each shot
b-fresh grains loaded before
series
b
Summary
• New results were obtained during the reported period practically at all main
direction of Globus-M tokamak research program.
• Greenwald limit densities are obtained both in OH and NBI heating regimes.
Average densities, obtained in low field of 0.4 T reaches (1.1-1.2)×1020m-3
with gas puffing.
• NBI thermalization was studied. Slowing down of beam ions is well
described by classical Coulomb scattering theory and the particle losses at
least below Ecr are insignificant. Regimes with overheated ions are achieved
at densities below 2.5×1019m-3, heating of electrons is observed at densities
higher (5-6)×1019m-3.
• The ICR heating study at the fundamental harmonic range were continued
on Globus-M tokamak to make clear and self consistent the picture of RF
heating in ST. ICR heating efficiency improvement with hydrogen minority
concentration increase in the wide range of 10 – 70% was recorded
experimentally and confirmed by simulations. The negative role of second
harmonic hydrogen resonance positioning was outlined in the experiments.
• The reliability of the plasma gun as the source for plasma feeding and the
instrument for the discharge initialization was confirmed. Numerical
simulations of plasma jet interaction with core tokamak plasma were
started, giving first results in tolerable agreement with experiments.