Princeton University Plasma Physics Laboratory, Princeton ... · -1 320 340 360 380 Df [kHz] 80.47...
Transcript of Princeton University Plasma Physics Laboratory, Princeton ... · -1 320 340 360 380 Df [kHz] 80.47...
1
Results from
ICR
F Heating and C
urrent Drive
on C-M
od*
G. Schilling
1, P. T. Bonoli 2, A
. Hubbard
2, L. Lin2, Y
. Lin2,
A. Parisot 2, M
. Porkolab2, J. R
. Wilson
1, S. J. Wukitch
2,and the C
-Mod Team
1Princeton University Plasm
a Physics Laboratory, Princeton NJ 08543
2MIT Plasm
a Science and Fusion Center, C
ambridge, M
A 02139
Presented at the 45th Annual M
eeting of the Division of
Plasma Physics
October 27-31, 2003, A
lbuquerque, NM
*Work supported by U
S DoE C
ontract DE-AC
02-76-CH
0-3073 and Cooperative
Agreement D
E-FC02-99ER54512
2
Abstract
On C
-Mod, ion cyclotron range of frequencies (IC
RF) is the only
auxiliary heating. To inject high power, the antennas need to be able to
withstand high RF voltages. To m
aximize voltage handling, the antennas
have been modified to avoid or m
inimize the R
F E-field in regions where
the RF E-field is parallel to the tokam
ak B-field. For 0.5 second pulses,the IC
RF antennas have now
delivered up to 5 MW
with m
inimum
densityand im
purity generation in H-m
ode. Furthermore, initial tests of the 4-
strap antenna in non-symm
etric phasing were begun. In m
inority Hheating scenario, the 4-strap antenna has coupled 2.7 M
W w
ith bothcounter and co-current phasing w
ithout significant negative impact on the
plasma. The saw
tooth period was observed to lengthen w
ith co-currentphase and shorten w
ith counter-current phase. In addition, initial mode
conversion current drive experiments w
ere begun. Direct electron heating
was observed, independent of phase, for a D
( 3He) m
ode conversionscenario. A
n analysis of the antenna performance and plasm
a responsew
ill be presented.
3
Outline
•Physics achievem
ents- M
ode conversion studies- Current drive initiation
•ICRF system
technical achievements
- Power
- Phased operation
•ICRF antenna upgrades- D
and E-port- J-port
4
C-Mod ICRF resonance location at 78 M
Hz
0 1 2 3 4 5 6 7 80.40.5
0.60.7
0.80.9
B(0.66m) (T)
R(m
)
ΩB
-11,5+Ω
D , ΩH
e-4
ΩH
e-3
ΩH
5
ICR
F Mode C
onversion
Narrow
deposition may
provide tool for modifying
or controlling pressure andcurrent profiles.
Mode conversion cancom
pete with other
damping m
echanisms.
•D
etailed measurem
entsprovide stringent test forIC
RF code predictions.
•M
ore complicated than
expected.•
Absorption details could
affect current and flow drive
predictions.
n3H
e /ne =
0.14B
T=8 T, Ip=0.8
D(3H
e) discharge
TO
RIC
GPC
expG
PCth
Unique com
bination of diagnostics, codes, andR
F system to study various scenarios D
( 3He),
H( 3H
e), and D(H
).•PC
I measures chord-integrated fluctuation
level.•H
igh resolution (~0.7 cm) EC
E radiometer
measures pow
er deposition profile.•First principle 2-D
full-wave calculation
using ICR
F code TOR
IC.
6
Phase Contrast Imaging diagnostic m
easureselectron density fluctuations
RF ANTENNAPC
I Chords
Wide spectrum
of wavenum
bers are measured
simultaneously; spatial structure is probed over w
ideareaPoor spatial localization results from
line integration12 vertical chords:
•60 cm < R < 79 cm
(R0 =0.66 m
, a=0.22 m)
•0.5 < |kR | < 10 cm-1
•Effective spacing is set by magnification optics
Utilize 100 W
CO2 laser (λ=10.6 µm
)Sensitivity
•1012 m
-2 Hz -1/2
•2-500 kHz
Planned upgrade:•U
pgrade from 12 to 32 channels.
•Increase digitizing rate from 1 to 10 M
Hz w
ithincreased m
emory.
»May allow
direct measure of w
ave damping rate.
•Utilize field alignm
ent of scattered wave to
provide localization.
7
Detected m
ode converted Ion Cyclotron wave
Contour Plot of Fourier Analyzed PCI D
ata
-10-5
05
10k
R , cm-1
320
340
360
380
Df [kHz]
80.47
80.50
80.53
fRF [MHz]
ph
ase velocity
tow
ards
anten
na
ph
ase velocity
away fro
man
tenn
a
Dispersion curves near M
C Region
FW ICW
IBW
FW
-6 -4 -2 0 2 4 6
PCI chords
5< k <
10 cm-1
PCI Signal Structure
~
Noise Floor
MC
surface•M
easured wavenum
bers are ingood agreem
ent with dispersion
relation.•Fluctuations are m
easured to thelow
field side of the mode
conversion surface.E. N
elson-Melby et. al., PRL 90 155004-1 (2003).
8 Experimental observation of D
(H) m
ode conversion
.4
1.4
1.6
1.5
2 2.5
2 2.4
1.61.821.21 1.6
1.2
1.4
1.6
0.7
0.9
1.1
0.5
0.6
0.180.2
0.220.24
H/D
0.60.8
10 0.5
1 1.5
PRF (M
W)
1.62 2.42.8
1.6
2 2.4
1.61.71.81.9
1.4
1.5
1.6
1.21.3
1.4
0.80.91 1.1
0.5
0.55
0.6
0.22
0.24
0.26H
/D
0.970.98
0 0.5
1 1.5
PRF (M
W)
Te [keV]
R=0.68 m
1020911026 B
T=5.26 T
, Ip =1.0 MA
, ne0 =1.8x10 20 m-3
Tim
e (sec)
R=0.70 m
R=0.73 m
R=0.75 m
R=0.78 m
R=0.8 m
R=0.83 m
Experim
ent
TO
RIC
S (MW/m^3/MW_inc)
r/a
Com
parison of measured and predicted pow
erdeposition profiles during D
(H) M
Cexperim
ents.
First experimental observation of D
(H) m
ode conversion.Experim
ental fraction of absorbed power is 0.16 and TO
RIC
predicts 0.14.
Y. Lin et al., PPC
F (accepted) 2003.
9
Off-axis D
(H) m
ode conversion
D(H
) mode conversion is
an opportunity tocom
pare simulation
with experim
ent.•
Direct m
onitor of Hm
inority is routine.•
Experimental
absorbed power is 0.2
and TOR
IC predicts
0.18.Sim
ulation indicates acom
plicated mode
conversion region.•
IBW
and ICW
areboth present.
Y. Lin et al., PPCF (accepted) 2003.
-20 -10 0 10 20
Z (cm)40
200-20
-40Z (cm)
50-5
-10
-15
ICW
IBW
R - R
0 (cm)
R - R
0 (cm)
-14 -12 -10 -8 -6 -4 -2
0.19
0.85
3.74
72.2
16.4
MA
X: 144
S_eld
FW
Data
TO
RIC
r/a
S (MW/m3/MWinc)
10
On-axis current drive initiation
0 0.51 1.52 2.53
P_R
F (M
W) 1030710005
0.04
0.06
0.08W
_MH
D (M
J)
1.52 2.53
Te0
0 2 4H
E3 N
RA
TE
0.70.8
0.9
1nl_04
0.5
1 1.52P
I foil (MW
)
1.5
2 2.5Z
_eff (z_ave)
0.60.8
11.2
1.44 6
D_alpha m
ain (AU
)
0 0 .51 1.52 2.53
P_R
F (M
W) 1030710006
0.04
0.06
0.08W
_MH
D (M
J)
1.52 2.53
Te0
0 2 4H
E3 N
RA
TE
0.70.8
0.9
1nl_04
0.5
1 1.52P
I foil (MW
)
1 1.5
2 2.5Z
_eff (z_ave)
0.60.8
11.2
1.44 6
D_alpha m
ain (AU
)
co-currentcounter-current
heatingheating
•S
awtooth
period increased x3 (co vscounter phasing)
RF
pow
er
stored energy
electron tem
p.
neutron rate
density
radiated pow
er
Zeff
Tim
e (s)
11
Off-axis current drive initiation
•F
irst off-axis current drive experim
ents
•O
bserved both shortening and lengthening of the saw
tooth period depending upon phase, deposition location, and pow
er level.
•F
urther analysis required to determ
ine driven current.
3 3.4
3 3.4
3 3.4
2.4
2.8
0.9850.995
1.005
1 2
Tim
e [sec]
PRF (M
W)
Rm
aj =0.66 m
Rm
aj =0.69 m
Rm
aj =0.72 m
Rm
aj =0.75 m
1030716016 IP =0.8 M
A, B
T=8.0 T
Te [keV] Te [keV] Te [keV] Te [keV]
12
ICRF system technical achievem
ents
•O
perated ICRF antennas to maxim
um achievable voltage in their
12/2002 configurations and evaluated voltage and power lim
its.
•Evaluated and im
plemented changes to increase the voltage and pow
erhandling capabilities of all three antennas.
•Began high-pow
er heating experiments w
ith the modified antenna
systems and identified candidate target plasm
a conditions likely to leadto high tem
perature and low collisionality.
•Produced high tem
perature plasma w
ith 5 MW
ICRF power for 0.5
sec.
13
5 MW
ICRF H-m
ode discharges
•A
ntenna modifications significantly
improved pow
er handling of D, E
, and J-port antennas.
-D
and E-port pow
er limit
increased to 1.5 MW
in H-
mode.
-J-port pow
er limit increased to
3 MW
.
•F
ull assessment requires m
ore operation tim
e and an inspection (Jan 2004)
2 4
PRF (M
W)
WM
HD
(MJ)
2 4
Te0 (keV
)
2 4
RD
D (x10 13 s -1)
1 1.5
nline_04 (x10 20 m-2)
1.5
2 0.15
0.05
1030605030 IP =0.8 M
A, B
T=5.4 T
0.81.0
1.2T
ime [sec]
Zeff
14
Dem
onstrated flexible antenna phasing
•In L-m
ode discharge, 2.7 MW
is coupled through J-port in co-current phasing for ~
0.4 sec.
•C
omparable plasm
a response to D
and E-port antennas (heating
phasing)
-A
ntenna heating efficiencies are independent of phase.
-Im
purity and density production are sim
ilar.
•V
ery encouraging results for future F
WC
D and M
CC
D
experiments
-A
lso use for flow drive
1 20.04
0.06
2 32 40.8
10.5
1 1.5
Prad (M
W)
1.5
2 2.5
0.60.8
1.21.4
4 6D
main (A
U)
PRF (M
W)
WM
HD
(MJ)
Te0 (keV
)R
DD
(x10 13 s -1)
nline_04 (x10 20 m-2)
1030710005 IP =0.8 M
A, B
T=5.2 T
Zeff
1.0
J-portD
and E-port
Tim
e [sec]
15
Antenna perform
ance
•V
oltage limits in 12/2002 w
ere ~30, 30, 35 kV for D
, E, and J-port.
•Pow
er limits in 12/2002 w
ere ~1.0, 1.0, 1.5 MW
for D, E, and J-port.
•Further antenna upgrades w
ere carried out 12/2002 - 3/2003.
•H
ave now achieved 35 kV
on all antennas.
•Pow
er limits now
up to ~1.5, 1.5, 3.0 MW
for D, E, and J-port.
16
ICR
F antenna upgrades
•Perform
ance limits dom
inated by antenna voltageholding lim
its and antenna front surface - plasma
interactions.
•The em
pirical rf voltage holding limit of ~15 kV
/cmw
here E||B
has required redesign of internal antennaelem
ents to reduce local rf electric fields.
•A
ntenna front surface - plasma interactions have been
largely eliminated by covering exposed m
etal with
boron nitride.
17
D- and E-port fast w
ave ICRF antennas
D-port:
Arc dam
age on feedthrough centerconductors at 20 ohm
section on#2, #3, and #4.•
replaced feedthroughs #2, #3,and #4
•m
odified center conductordiam
eter to 50 ohm dim
ensionA
rc damage w
as found on the air sideof #4.
Upper left therm
ocouple was m
elted.
E-port:A
rc damage on #4 feedthrough center conductor at 20 ohm
section.•replaced feedthroughs #2 and #4•A
nd modified center conductor diam
eter to 50 ohm dim
ensionD
amaged feedthroughs on #4 cannot be repaired.
18
Arc dam
age in D and E feedthroughs
Dam
age suggests arcing is where E||B.
•D-port had dam
age on feedthrough #2, #3, and #4.•E-port had heavy dam
age localized to #4 with m
inor damage to #2.
Modified FS resulted in longer antenna electrical length.
•Moved higher voltage to this section of transm
ission line.Estim
ated impedance is ~20 ohm
s.•50 ohm
line has maxim
um voltage handling capability.
B
19
D and E feedthrough m
odifications
Modification w
ill lower electric field by ~30%
.•
Should require about 2 times the circulating pow
er to reachbreakdow
n.•
Coupled power should also double but w
ill be limited by
transmitter to a m
aximum
of 2 MW
.
20
J-port fast wave antenna
Installed new nose tiles replacing the
failed BN cover tiles.
Replaced septum spine because of
severe damage.
Refurbished feedthroughs:•
#4 vacuum side ceram
ic was
cleaned before re-installation.•
refurbished #2 and #4feedthrough center conductors.
Modified antenna m
ounting plate to improve contact and reduce E-fields
Center conductor connection to vacuum strip line and spool piece contact w
ith strip lineand current strap w
ere improved.
Did not install RF contact along side plate (capacitors).
Found arc damage across field lines on #4 near back plate feedthrough. V
acuumcoaxial line center conductors w
ere damaged.
21
BN
cover tiles failed
Tiles top #2 and #3 and bottom #7 failed
completely.
Tile bottom #5 partially failed.
Disruption force results in a torque
applied to the BN
tile.•
The plate the BN
tile rests uponand the rail experience thistorque.
•The B
N experiences both
compressive and tensile stress.
•B
N perform
s poorly undertensile stress.
Fastening system m
oved with the B
Ntile instead of being hard fastened.
Disruption induced torque
22
BN
tile redesign
Replace front and cover tiles w
ith singlefront tile.
•Previous nose tiles have not failed todate.
»Test indicates the tile fails ~1000 G
.»
Failed tiles suggests ~200 G loads.
•Shield fasteners from
plasma.
»Preload bevel w
ashers to ~ 200 G so
the BN
tile moves w
ith metal tile.
•M
aterial under the fastener increased to0.5”.
•Increase front m
aterial by removing
metal.»
Alleviate over heating.
Mo tile m
odified to accomm
odate thickerB
N tile.•
Tile strength is not comprom
ised.
23
Septum sustained significant dam
age
Observations:
•M
elt damage is m
ost severe where
there is no probe.•
Probes appeared to weld to the spine.
•Injections from
J-port ~ at cross overbegan 1020723.
•A
ntenna ran in (0,0,0,0) phasing from1020628-1020802.
»H
igh power for ~ 8 discharges.
Mechanism
:•
For (0,0,0,0) phasing, large fluxthrough slots results in voltage acrossslot.
•B
N tile facilitates breakdow
n.M
ade new spine w
ithout slots.
Top
Bottom
24
Mounting plate m
odifications
Address arc dam
age found invacuum
coax.B
asic problem is reliable contact.
•Possible to have poor contactdue to m
ultiple connections.To im
prove reliable contact andreduce E-fields.•
Increased diameter of opening
and added full radius.•
Increased the bolt pattern from4 bolts to 8 bolts.
•R
elieved the plate to allow the
plate to rest only around theopenings.
•Plated the flange face w
ithcopper to im
prove contact.
25
Modified all junctions to im
prove contact
Num
ber of junctions showed signs of series arcing.
Improved contact by relieving the flat section leaving ashoulder.
26
Antenna perform
ance summ
ary
•C
-Mod has presented a challenge to install high pow
er (~4 MW
) 4-strap and (~2 M
W) 2-strap IC
RF antennas in a tight space.
•plasm
a - front surface interactions•
empirical voltage holding lim
it•
high neutral edge pressure, 0.1-1 mTorr
•C
areful analysis of antenna fault modes together w
ith thoroughantenna inspection follow
ing each run period have allowed us to raise
antenna performance system
atically during use on C-M
od.
•W
e have been able to bring the total antenna power up to above 5
MW
, with good heating efficiency and no deleterious effects on the
plasma.
•Im
provements w
ill continue as we learn m
ore about antenna behavior.