Supported by
College W&MColorado Sch MinesColumbia UnivComp-XGeneral AtomicsINELJohns Hopkins UnivLANLLLNLLodestarMITNova PhotonicsNew York UnivOld Dominion UnivORNLPSIPrinceton UnivSNLThink Tank, IncUC DavisUC IrvineUCLAUCSDUniv ColoradoUniv MarylandUniv RochesterUniv WashingtonUniv Wisconsin
Culham Sci CtrUniv St. Andrews
York UnivChubu UnivFukui Univ
Hiroshima UnivHyogo UnivKyoto Univ
Kyushu UnivKyushu Tokai Univ
NIFSNiigata UnivUniv Tokyo
JAERIHebrew Univ
Ioffe InstRRC Kurchatov Inst
TRINITIKBSI
KAISTENEA, Frascati
CEA, CadaracheIPP, Jülich
IPP, GarchingASCR, Czech Rep
Univ Quebec
New Capabilities and Results for the
National Spherical Torus Experiment
M.G. BellPrinceton Plasma Physics Laboratoryon behalf of the NSTX Research Team
presented at theJoint Meeting of the
3rd IAEA Technical Meeting and the
11th International Workshop on
Spherical TorusSt. Petersburg, Russia
Oct. 3 – 6, 2005
Bell, M. / STW’05 / 051003
2
NSTX Just Completed 18 Weeks of Experiments Exploiting New Capabilities
and Regimes
• Good performance of TF coil joints: ~2500 pulses, 75% at 0.45T or higher
– Operated to 95% of design field (0.6T) at end of run
– Joint resistances remained below expectations
• Extensive investigation of Coaxial Helicity Injection
– Higher capacitor bank voltage (1.5kV) and localized gas injection
• New PF1A coils for simultaneous high triangularity and high elongation
• Three pairs of magnetic field perturbation coils
– Powered by fast Switching Power Amplifiers for Error-Field Correction and Resistive Wall Mode control
• Lithium Pellet Injector, Supersonic Gas Injector
– LPI used both for plasma perturbation experiments and Li coating
• Movable Glow Discharge electrode
– More uniform coverage and operation at lower gas pressure
Bell, M. / STW’05 / 051003
3
Coaxial Helicity Injection Experiments Benefited from New
Capabilities
•Inject gas and microwave power into chamber beneath injector gap
– Reduced total gas requirement for breakdown
•Capacitor bank with fast crowbar and higher voltage for coil producing initial flux
– Operated bank at up to 1.5kV
– Force reconnection to produce closed flux
•Fast cameras to observe plasma on closed flux surfaces
Gas, ECH Injection
Injectorcurrent
– )| +Capacitor bank
10 – 50 mF, 1 – 2 kV
Insulated gaps between inner, outer divertor
plates
Poloidal fieldlinking platesToroidal
fieldCL
Bell, M. / STW’05 / 051003
4
CHI Produced Persistent Current and Closed Flux For ~10ms After the
Injector Current Pulse
17 ms15 ms
13 ms11 ms
9 ms7.5 ms
• Need to develop transition to control of toroidal equilibrium R. Raman, B. Nelson (U. Wash.), R. Maqueda (Nova), B. LeBlanc
5101520010120100Time (ms)Injector Voltage (kV)Injector Current (kA)Toroidal Plasma Current (kA)Times of Thomsonscattering data
Time of capacitorbank crowbar
118340
Radius (m)0.51.01.500.51 t = 13 ms15 ms
Electron density (1019m-3)0.51.01.501020 t = 13 ms15 ms
118340 Electron temperature (eV)118340
Bell, M. / STW’05 / 051003
5
Motional Stark Effect Diagnostic Routinely Measures Magnetic Field Line
Pitch
• 8 channels span from outer edge past magnetic axis• Calibrated by NBI into neutral gas with known applied fields
• EFIT provides between-shots analysis for q-profile– Can include correction for Er using CHERS v profile
– Also applying LRDFIT, ESC equilibrium codes
RMS error 0.4°
Radius (m)
Pitch angle (degrees) +40
+30
+20
+10
0
-100.9 1.0 1.1 1.2 1.3 1.4 1.5
MSE data
EFIT analysis
F. Levinton, H. Yu ( Nova Photonics), S. Sabbagh (Columbia U.)
Bell, M. / STW’05 / 051003
6
0.51.01.5Radius (m)0.00.51.00.00.51.0neTiTenDkeV1020
m-30.75s1178140.75s
51-Channel Charge-Exchange Recombination Spectroscopy Reveals
Extreme Gradients
•Diagnostic measures Ti(R), v(R), nC(R)
– Infer nD(R) under assumption that carbon is dominant impurity
Radius (m)
Toroidal velocity (km/s)
1.3 1.4 1.5
150
100
0
50
MPTS
R. Bell, B. LeBlanc
Bell, M. / STW’05 / 051003
7
New Inboard Divertor Coils Increase Accessibility of High-Triangularity,
High-Elongation Shapes
• Highest = 2.7 now obtained at highest ≈ 0.8 S q95 IP/aBT = 35 MA/m·T
– Record stored energy = 430kJ at IP = 1.4MA, N = 5.3, T = 29%, q* = 2.3
• Small ELM regime recovered at high > 2.5 with new coils– Previously observed onset of large ELM-like events when > 2.2
2004 2005
New PF1A coilOld PF1A coil
All values at
MAX(T) > 20%
WMHD = 430kJ
at IP = 1.4MA
2004 2005
J. Menard, D. Gates
Bell, M. / STW’05 / 051003
8
Record Pulse-Lengths Achieved at Moderate Current by Operating with
Sustained H-mode
•H-mode with small ELMs lower flux consumption, slow density rise
•High P increases bootstrap fraction lower flux consumption
N
IP (MA)
VSURFACE (V)
T = 17%P = 1.5, li =
0.65
Current Relaxation
Time (s)
ne / nGW
= 2.3, L = 0.75
E
H89P = 2-2.2
H98(y,2) = 1-1.1
PNBI (MW) / 10
J. Menard
Bell, M. / STW’05 / 051003
9
Low Loop-Voltage Quiescent Phase Ends at Onset of Saturated n=1 Mode When qmin
1
•Saturated n=1 mode persists for 0.5s
•Central rotation drops by factor 3
•Edge f maintained
• N decreases
– N = 6 above no-wall limit
– N = 4 below no-wall limit?
•qmin sustained near 1
– No sawteeth observed
– Similar to hybrid mode
Time (s)
Core f (kHz)
Edge f
N
PNBI = 6MW
qMIN
qMIN without Er correction
J. Menard, F. Levinton, S. Sabbagh
Bell, M. / STW’05 / 051003
10
Long Duration Discharges Exceed 70% Non-Inductive Current During High-
Phase
•TRANSP model agrees with measured neutron rate during high-phase
•TRANSP over-predicts neutron rate early and late when low-f MHD present– Fast-ion diffusion and/or loss likely
– Assessing impact on JNBI profile and q-profile
•85% of non-inductive current is p-driven– Bootstrap + Diamagnetic + Pfirsch-Schlüter
0.00.51.01.50.00.50.01.02.00.00.5DD neutronrate (1014s-1)Fraction of
total currentFraction of total energy
IonsBeam ionsElectronsInductiveBootstrap (Sauter)Beam-driven∇pMeasuredModeledTRANSP116313 03A
0.01.02.0P< >EFIT( )Time s
Bell, M. / STW’05 / 051003
11
Also Extended Pulse Length at Higher Current
•At high Ip, contribution reduced from– bootstrap current (fBS 1/2P)
– NB-driven current (higher density)
0.51.01.5Radius (m)0.00.51.00.00.51.0neTiTenDkeV1020
m-30.75s1178140.75s00.51010100.51Time (s)keV1020
m-3V0101Ip (MA)VsurfPne(0)<ne>Ti(0)Te(0)117814PNBI( )/10MW-H mode
WMHD=430kJ
BT = 0.45T
T = 29%
J. Menard, D. Gates
Bell, M. / STW’05 / 051003
12
Lithium from Injected Pellets Deposited on Plasma-Facing Surfaces Provides Edge
Pumping
•Fired lithium pellets (1.7 – 5 mg) from multi-barrel pneumatic injector into sequences of ohmically-heated helium discharges– Limited on center-stack tiles or lower-single-null divertor
– After “pre-conditioning” surfaces with OH helium plasma
– 1 or 2 pellets per discharge, 24 – 30 mg total lithium in each sequence
•Dramatic reduction in density in 1st subsequent NB-heated deuterium gas-fueled discharge (~3.5mg D2) in both configurations
0.00.51.01.50.24 – 0.28 s0.51.01.5ne(1019m-3)
Ist shotafter LiBefore Li(after He)Radius (m)0.00.51.00.00.10.20.30.40.5Time (s)<ne>
(1019m-3)Before Li(after He)3rd after Li2nd after Li1st after Li (25mg)0.8MA, 0.45T, divertor discharges, deuterium gas fueled (~4mg/shot)Gas off,plasmadiverted
•Effect disappeared by 3rd similar discharge– Expected if most injected gas reacts with deposited lithium H. Kugel
Bell, M. / STW’05 / 051003
13
External Radial Field Coils Used to Counteract Effects of Intrinsic Error
Fields
• Investigated threshold for occurrence of locked-modes in low-density, low- discharges as magnitude and direction of radial field varied
VALEN (Columbia Univ.)
6 ex-vessel coils(1.6m2 2 turns each)48 in-vessel sensors
(B, B)
Inferred intrinsicerror fieldB 2,1 = 1.3G
EF = 140º
ne at q=2
~41018m-3
• IP=0.7MA• BT = 0.45T• q(0) =1.1-1.5 (no sawteeth)
3 SPAs 3.3kA3.5kHz
J. Menard
Bell, M. / STW’05 / 051003
14
Applying n=1 Correction Field with External Coils Extended Pulse Length at
High N
•With field in “non-correcting” directions, rotation is damped – Earlier island locking and/or RWM formation•In “correcting”
direction, near-edge rotation collapse is avoided– Extends pulselength at high-
117577
117571
f
(carbon)
J. Menard, S. Sabbagh, R. Bell
Bell, M. / STW’05 / 051003
15
Applied n = 3 Stationary Field Used to Control
Plasma Rotation Reversibly
•Rotation stops completely in vicinity of resonant q = 3 surface
– Neoclassical toroidal viscosity affects inner region
•Rotation can recover if RWM is not destabilized W. Zhu, S. Sabbagh, R. Bell
Bell, M. / STW’05 / 051003
16
Repetitive Bursts of MHD Activity Affect
Fast Ion Population During NBI Heating
• Bursts of MHD activity detected by Mirnov coils are kinetic instabilities driven by population of fast ions
– Instabilities chirp in frequency
– Toroidal mode numbers n = 2 – 6
• Bursts are coincident with rapid drops in DD neutron rate
– Mainly beam-target reactions
• NPA measures energy and pitch angle resolved effect on fast ions
– Depleted over wide energy range
– No D spikes to suggest changes in edge neutral density
E. Fredrickson, S. Medley
Bell, M. / STW’05 / 051003
17
Robust Reversed-Shear Startup by Varying
Ramp-Rate, NB Timing and Gas Injection
q profile (LRDFIT with MSE)0.00.10.20.30.4Time (s)100210100200Ip (MA)PNBI (MW)WMHD (kJ)115731115730
•q-profile very dependent on early MHD acitivity
– Events lead to monotonic profile
•Also combined RS with H-mode edge but not yet optimized F. Levinton (Nova), J. Menard
Bell, M. / STW’05 / 051003
18
Reversed-Shear Appears to Produce an Electron Transport Barrier
• e reduced by ~2 in confinement zone
• HHFW heating also raised central Te in RS
00.51√ΦNe(m2s-1)100101115730,0.296s115731,0.326s±0.025Average sVariation±0.025sTRANSP with classical NB thermalization
F. Levinton (Nova), R. Bell, B. LeBlanc, S. Kaye
3210Ti (keV)115730, 0.296s115731, 0.326s210Te (keV)115730, 0.293s115731, 0.327s420ne (1019m-3)115730, 0.293s115731, 0.327s0.51.01.5Major radius (m)
Bell, M. / STW’05 / 051003
19
Newly Installed High-k Scattering Diagnostic Will Probe Turbulence Related to Electron Transport
• = 1mm probe beam launched tangentially near midplane
•Inboard / outboard launch configurations– kr = 0 – +20 cm-1 (in), -20 cm-1 (out)
•Superheterodyne receivers (UC Davis)– Five channels with NF of ~3000oK
•First scattered signals at end of 2005 runOutboardLaunch
InboardLaunch
0 500 kHz-500Amplitude (arb.)
Fourier transform of scattered signalOutboard launch, kr = -15cm-1
Shot 118148t=200ms
H. Park, D. Smith, E. Mazzucato, C. Domier (UCD)
Bell, M. / STW’05 / 051003
20
MSE Shows Evidence for Formation of “Current Hole” in NSTX
•At 0.12 s current profile is hollow but central current density is finite
•Small region of almost zero current density forms at 0.13 s
•Expands to about 15 cm diameter by 0.20s
•Central current density becomes positive again by 0.24 s
F. Levinton, H. Yu (Nova)
Bell, M. / STW’05 / 051003
21
Investigating Aspect-Ratio Dependence of
H-mode Pedestal in Experiment with MAST, DIII-D
•Also studying H-mode threshold in MAST/NSTX
ne [1019m-3] Te [keV]
tot [%]
N
Ti [keV]
e
NSTX (117723@427ms) DIII-D (121516@1175ms) MAST (12899@270ms)
Cross-sections projected onto NSTX equilibrium
R. Maingi (ORNL), A. Field (MAST), H. Meyer (MAST), T. Osborne (DIII-D)
Bell, M. / STW’05 / 051003
22
Puffing Deuterium into Private Flux Region Reduces Outer Divertor Heat Flux
by Factor 2–4
•No change in inner divertor heat flux - inner leg already detached
•Evidence for volume recombination from increase in D/D ratio– Divertor radiation increases but not spatially resolved
•Outer divertor detachment not achieved by midplane injection of D2 or Ne– Ne did reduce divertor heat flux by factor 4 by plasma and SOL radiation
QuickTime™ and aTIFF (LZW) decompressor
are needed to see this picture.0
-1
-2
1(m)
2
0 2 (m)1D2
D2 fueling rate (1021/s)Peak heat flux on outer divertor
(MWm-2)
Ip = 0.7MA, BT = 0.45T, PNBI = 3–4MW
V. Soukhanovskii (LLNL)
Bell, M. / STW’05 / 051003
23
NSTX Made Significant Progress in 2005
•Clear demonstration of persistent current initiated by CHI
•Extension of operating regime at high and •First experience with full set of Error Field Correction and RWM Control coils– Cancellation of intrinsic error fields– Control of rotation and effect on mode growth
•Significant extension of pulse-length at moderate current
•Exploration of reversed-shear regime– Beneficial effect on electron confinement
•Demonstration of particle control with lithium coating in both limiter and divertor plasmas
•Contributions to physics of H-mode, ELMs, pedestal, confinement– Collaborative experiments and participation in ITPA
Bell, M. / STW’05 / 051003
24
Status and Plans
•Now entering an outage planned to last until December 2005
– Install in-vessel components for “poloidal CHERS” diagnostic
•Opposing vertical views across NB path to measure vpol and Ti
– Install lithium evaporator for coating areas of divertor and wall
•Build on experience in LTX (former CDX-U) and with LPI in NSTX
– Examine all TF coil joints and refurbish if necessary
•NSTX 2005 Results Review and Research Forum for planning experiments in 2006 will take place December 12 – 16
– Participation by our collaborators is encouraged
– Length of 2006 experimental run is not yet known
•Awaiting outcome of budget negotiations in US Congress