Abstract Hoist Levitated Dipole Experiment (LDX) Significant Stored Energy in Thermal Plasma Towards Higher Density Dipole Confined Plasmas
No Levitation: 80515025 80515026 Flux Decay of “Ring Current” with ECRH Off Only hot trapped electrons Dipole Edge Power Balance Scaling LDX Phase III Target Plasmas The Levitated Dipole Experiment (LDX) investigates plasmas confined in 1.1 MA Floating dipole coil 6 Levitation Coil Total Heating Power 81001006 0.4 Supported 2.5 kW (2.45 GHz source)! the closed field line dipole magnetic geometry where the plasma stability 4 1.0! Levitated 2.5 kW (2.45 GHz source)! ‣ Nb3Sn superconductor 12 kW Levitated 15 kW! Steady state central chordal density Original proposal goal for thermal plasma studies
is provided by compressibility and where plasma convection leads to 2 0.2 0.8! ) -2 15 kW, He measurement versus neutral 10 13 3 ‣ Inductively charged by 10 MJ 0 cm 5 kW, He ~10 / cm density peaked profiles and may allow for τE > τp. In the past year, we have 13 ‣ 20 0.6! pressure for different conditions. 15 kW, D2 charging coil Neutral Pressure 0.0 8 continued to investigate the improved energy and particle confinement 15 5 kW, D2 ‣ peak β order unity (Flux (mWb))
10 0.4! 5 kW, D2, supported Torr 10 6 when the supports are removed from the plasma. ‣ Up to 2 hour levitation using -6 -0.2 Increasing power increases the core Log 10 5 Current plasmas ~ 15 kW Hot Electron Energy Fraction ! 0.2! active feedback on upper 0 density and there is large mass effect 4
Of most significance, we observe that: 5 -0.4 12 3 15 ‣ peak density ~ 0.8 x 10 / cm levitation coil > Chord average density 14.0 14.5 15.0 15.5 16.0 16.5 0.0! for the optimal fueling. This 2 -3 A large fraction of the stored energy in levitated plasmas is contained 10 Time (s) 0.0! 0.5! 1.0! 1.5! 2.0! 2.5! 3.0! 3.5! Line Integrated Density ( x10 • m observation is consistent with the ‣ edge temp 15-20 eV Diamagnetic Flux (mWb)! Two component plasma created 16 0 5 0SOL density5 10increases 15 with20 P . p25 is varying30 35 in the bulk electron population. 1 m dipole “natural profiles” (nV = rf 0 ‣ core temp 400 eV ? (see Davis UP8.00054)
by multi-frequency ECRH < 10 Neutral Pressure ( !Torr ) 0 The observation of signficant fraction of bulk plasma beta is determined from the 90723004-19 4 constant), where the edge • Eliminating plasma losses along field lines allows observations of a 2.45 GHz ECE (V-band) ‣ thermal peak β ~ 10 % ‣ 2.5 kW, 2.45 GHz 3 nature of the diamagnetic flux signals which show an initial energy decay of the parameters are set by power flow along strong particle pinch leading to a density profile with near equal 6.4 GHz Edge density proportional to power 2 bulk plasma followed by the long decay of the fast particles as also seen in the the scrape off layer (SOL). Here the FY2010 - University of Maryland 28 GHz gyrotron ‣ 2.5 kW, 6.4 GHz A.U. number of particles per flux tube. 1 supported shots. Furthermore, the radiometer, which sees primarily harmonic LDX experiment. The vacuum vessel is 5 m in diameter and the )
improved confinement of He is due to the -3 ‣ 10 kW, 10.5 GHz 0 ECE from the fast electrons, does not show the marked increase that the ‣ 10 kW addition power ( + 55 %) The previously observed low frequency fluctuations are consistent 560 kg superconducting dipole coil is 1.2 m in diameter. 3 slower sonic flow to the wall in the SOL. • ‣ 10 kW, 28 GHz (FY2010) Diamagnetic Flux diamagnetic loops demonstrate. Thus the increase in beta during levitation is ‣ peak density ~ 1.5 x 1012 / cm3 with the required turbulent convection to drive the observed particle 2 largely in the bulk warm electrons. Plasma Diagnostic Set Weber Indeed, probe measurements of density ‣ thermal peak β ~ 20 % -3 1 pinch. γ
10 The significant bulk beta is roughly consistent with a pV = constant profile
Improved Confinement with Levitation and temperature at the SOL indicate a Edge Density (cm Inductive 0 1 MW 3-28 MHz CW transmitter installation In future experiments, an optimal levitation control system will be 0 5 10 15 starting with the measured edge Te of ~20 eV. This profile would have core Te ~ nearly linear scaling of density with MagneticChargin equilibriumg Time (sec) 400 eV. (a) Side View Sept. 25, 2009 5 implemented, reflectometer and soft x-ray spectrometer diagnostics will power and a nearly constant edge T is approximatelyInput Power independent (kW) of P ‣ 200 kW ICRF heating ‣ flux loops, Bp coils, Hall effect sensors, levitation system trackers Similar shots Supported e rf be added, and a higher frequency (28GHz) electron cyclotron heating 2 m When levitated, a large increase in density and magnetically measured stored temperature, as predicted by the notion ‣ ~1013 / cm3 density Fast electrons and levitated energy is seen with only modest increase of ECE emission,Upper Hybr id which is sensitive that the edge power balance determines Edge temperature ≈ constant source will be added to increase the density of LDX plasmas. In Resonances ‣ peak β order unity ‣ Similar gas fueling mostly to the mirror trapped hot electron population.Cyclotron the edge parameters (see Kesner, et al. preparation for installing a 1MW ICRF transmitter, initial ion cyclotron ‣ X-ray PHA, x-ray detector, 60 GHz & 137 GHz radiometers Resonances ‣ Isotropic, thermal plasma 3-5 x line average density UP8.00052). heating experiments will also be performed. Core parameters Ti=Te, 500 eV core much more peaked interferometer, visible cameras, visible diode and array, survey spectrometer ‣ Thus, assuming the edge scales with ‣ (eV) Te profile with levitation Inward Particle Pinch Observed during Levitated Plasmas Columbia University Fluctuations power and the natural profiles, we can 2-3 x the diamagnetic extrapolate the power levels required to The Levitated Dipole Concept ‣ Edge Isat and Vf probes, Mirnov coils, visible diode arrays, interferometer, fast flux Dramatic Peaking of density profileCatcher seen when Levitated reach the LDX Phase III target plasma visible camera, floating probe array Catcher Open Significant stored energy Raised 4 Channel 28 GHz Gyrotron ECRH UpgradeInput Power (kW)forPSFC LDX MIT Columbia University Lowered Field-Lines Reconstruction of the density conditions. Sept. 25, 2009 6 Edge parameters in warm bulk plasma (b) Top View Interferometer 28 GHz Gyrotron ECRH Upgrade for LDX MHD stability from plasma 6 profile for supported P. C. Michael, P. P. Woskov, J. L. Ellsworth, J. Kesner,University of Maryland S71213003 R. F. Ellis, d compressibility when levitated Interferometer (Radian) (diamond) and levitated Columbia University P. C. Michael, P. P. Woskov,28 J. GHz L. Ellsworth, Vacuum WindowJ. Kesner, PSFC MIT V = ‣ swept, Isat, and floating potential probes B γ γ D. T. Garnier, M. E. Mauel, If p V = p V , then interchange does Observed over all 4 (square) discharges under University of Maryland28 GHz ECD. Resonances T. Garnier, 28 M.in LDXE. Mauel,GHz Columbia GyrotronFused University quartz brazed R. in F.standard Ellis, flangeUniversity of Maryland 1 1 2 2 hmm@ .2798 m In development Resonant Thickness > operating conditions similar conditions with 15 kW m =1, 2, 3, … not change pressure profile. (5.596 mm thick window reflectometer for density profile (peak density) corroboration and internal 28 GHz Gyrotron Installation at LDX m = 2 chosen) 28 GHz EC Resonances in LDX 28 GHz Vacuum Window ‣ 2 II. RESULTS Closed of ECRH. ABSTRACT -A 10 kW, CW, 28 GHz gyrotron is being implemented 28 GHz28 GHzGyrotron Power Installation at LDX Fused quartz brazed in standard flange Drift wave stability for interchange invariant Field-Lines on LDXWaveguide to increase the plasma density and to more fully explore the 44 mm diameter (2a) fluctuations ABSTRACT -A 10 kW, CW, 28more fullyGHz explore gyrotron the is being implemented 28 GHz heating Resonant Thickness potential of high beta plasma stability in a dipole magnetic Waveguide hmm> @ 2798 . m profile The shots analyzed are all deuterium plasmas, and the Supported mode has profile on LDX to increase the plasma density and to Outer closed Rupture Safety Factor (5.596 mm thick window ‣ soft x-ray spectral measurements 0 stability in a dipole magnetic configuration. Higher density increases the heating of ions by thermal 2 28 GHz Power m =1, 2, 3, … d ln T 2 flux‣ heatloop with broad profile acrossh plasma m = 2 chosen) 0 5 10 short-time15 “stripey” and correlation plots are all from consistent with particle potential of high beta plasma ses the heating of ions by thermal equilibration and allows for improved wave propagation in planned §· For η = = , density and time (s) Dipole Dipole S.F. 395¨¸ 24 44 mm diameter (2a) periods where all heating sources are on. The phase configuration. Higher density increa propagation in planned ICRF experiments. This represents over a 50% increase over the Innerin closedhigh field region ©¹a d ln n 3 sources and parallel losses. ws for improved wave Outer closed equilibration and allo flux loop flux loop Rupture Safety Factor 6 velocities were all calculated during periods where ev- over a 50% increase over the present 17 kW ECRH from sources at 2.45, 6.4, and 10.5 GHz. The F-Coil F-Coil 2 temperature profiles are also stationary. S71213004 ‣ antenna designed for direct beam to Interferometer (Radian) ICRF experiments. This represents higher frequency will also make possible access to plasma densities of 1st 28 GHz Absorption §·h ery heating source was on as well. Since the potential 2.45, 6.4, and 10.5 GHz. The Inner closed S.F. 395 24 Levitation System The levitated case has a 13 -3 2nd ¨¸ present 17 kW ECRH from sources at plasma densities of up to 10 cm . The 1 Tesla resonances are located above and below 28 GHz EC resonanceharmonic zones < 1%, or < 100 Watts at 10 kW beam power F-Coil F-Coil flux loop ©¹a is measured by the probes, the azimuthal/toroidal elec- ke possible access to 3rd the floating coil near the dipole axial region. The gyrotron beam will be resonances 1st Interchange mixing leads to constant 4 higher frequency will also ma th 01 Mode singly peaked profile with e located above and below 4 2nd tric field can be directly determined from the gradient -3 Gyrotron cabinet Waveguide Gap Losses, TE 28 GHz Absorption 13 cm . The 1 Tesla resonances ar transmitted in TE01 mode in 32.5 mm diameter guide using one 90q Gyrotron cabinet rd 28 GHz EC harmonic up to 10 Vacuum InstallationChamber underway 32 3 entropy peaked profiles 81002100 near constant number of l region. The gyrotron beam will be 1.4% or 140 Watts at 10 kW 4th resonances < 1%, or < 100 Watts at 10 kW beam power 300 of the potential as shown in equation 1 [6]. The electric q Vacuum Chamber bend and a short < 5 m straight waveguide run. A Vlasov launch Ph Well controlled flight the floating coil near the dipole axia er guide using one 90 45. 3 with fused quartz gap 1 γ L-coil Current field can be used to compute the E × B drift velocity, as ode in 32.5 mm diamet antenna in LDX will direct the beam to the upper 1 Tesla resonance n V − and p V − 200 2 particles per flux tube. This transmitted in TE01 m by‣ aCustom Vlasov antenna vacuum penetrationPao (with very Waveguide Gap Losses, TE01 Mode veguide run. A Vlasov launch region. A layout of the planned system is presented. beam will be launched implement to allow 32 ∝ ∝ ‣ After launch, < .2 mm z excursions 100 shown in equation 2 [7] . These drifts are directly related Water cooling will be kA bend and a short < 5 m straight wa •A diverging 28pole GHz plasmashallow region angle to vessel) Ph1.4% or 140 Watts at 10 kW “natural” profile is consistent to the upper 1 Tesla resonance cident on28 the GHz upper system di layout at LDX for ECRH in the high field plasma region •A diverging 28 GHz beam will be launched by a Vlasov antenna 45. to convection, and measuring their magnitude shows its in the high field plasma region ly at oblique continuous plasma operation > 1 min. 3 with fused quartz gap 0 antenna in LDX will direct the beam •The beam willpolarized be and most in Pao ‣ oscillation in x and y caused by 0 planned system is presented. 28 GHz system layout at LDX for ECRH •The beam will be incident on the upper dipole plasma region -100 strength in transporting particles to and from the core of with interchange mixing and Motivation •The incident beam‣ Vacuum will window, be beamline, linearly and dummy load fabrication at MIT. Water cooling will be implement to allow 1!! IC/P4-12 0 5 10 15 region. A layout of the c field direction •The incident beam will be linearly polarized and mostly at oblique slow rotation 10 angles to the magneti Gyrotron Valsov Launcher continuous plasma operation > 1 min. 8 Launcher Force/g time (s) the plasma. diffusion of total particles per Advance to more fusion relevant plasmas the donut hole should distribute angles to the magnetic field direction Uncommon Closed Field Line Plasma Topology TE Converter Complete turnow key the system dipole TE02 TE01 Converter 6 Gyrotron TE02 01 •Reflections and andabove/bel Initial propagation operation planned downfor Jan 2010. •Reflections and propagation down the donut hole should distribute TE01 Waveguide gravitational well in toroidal rotation Motivation the 28 GHz absorption around CPI HeatWave Model VIA-301 1 §·kcn Confinement Improvement with Magnetic Levitation kg 4 flux tube. key system • 10 kW ECRH power increase the 28 GHz absorption around and above/below the dipole Valsov Launcher (a) Short Half-Second Heating Pulse (b) Visible Light from Supported and Levitated Plasma Complete turn • Frequency: 28 GHz D sin ¨¸ caused by incomplete balance 2 – > 50 % increase over present ECRH (17 kW) ‣ See Woskov, UP8.00059 ©¹k of Superconducting Dipole CPI HeatWave Model VIA-301 • Power: 10 kW, min. Mode filter 0 #" ! Transmission Line 1 §·kcn TE01 Waveguide ECRH Power (kW)E = −∇Φ (1) Supported Advance to more fusionLevit arelevantted plasmas Mode filter Circular TE01• Duty: Continuous D sin 6 minute period with very small damping -2 • Frequency: 28 GHz Lacos 2 D 19.2 mm Circular TE01 Transmission Line ¨¸ L-Coil 10 S81002027 S81002020 10 kW, min. • Increase plasma density • Transmitted mode: Circular TE01 ©¹k 4500 A / 50 V Power supply 150 kW ! ! • Power: and LDX vacuum port entry have D.T. Garnier 1), A.C. Boxer 2), J.L. Ellsworth 2), J. Kesner 2), M.E. Mauel 1) Z position E × B 10 kW ECRH power increase 13 -3 ' Forward detector No magnetic shear Power Supply 5 • Continuous – 10 cm density cut off limit (tokamak regime) as much as possible • The gyrotron cabinet location and LDX vacuum port entry have Lacos 2 D wal present ECRH (17 kW) Duty: • The gyrotron cabinet location output Q n 2S #! !vE×B = (2) > 50 % increase over • 01 Forward detector the transmission line Directional0 k coupler been chosen to simplify the transmission line as much as possible 19.2 mm Control algorithm 0 2 – kcn O ‣ Resistive coil allows for rapid shutdown been chosen to simplify a ' B • Transmitted mode: Circular TE –LDX stray field at the gyrotron < 2 Gauss oriented vertically mm Reflection detector L Q ‣ ExB convective cells are -5 FIG. 1: A top-down view of LDX at the midplane. This view Enable newDirectional physics investigations coupler –LDX stray field at the gyrotron < ' 2 Gauss oriented vertically k 0n k 2S r 1) Department of Applied Physics, Columbia University, New York,Laser NY Position 10027, USA Low Frequency Fluctuations∂θ consistent with Turbulent Pinch output (Q cn a O Detectors ECRH Heating Pulse Increase plasma density Reflection detector Q 0n root of J’0 • The TE mode 28 GHz beam will be Realtime digital control computer ‣ P - I - D - A control of voltage -10 shows the extent of the probe array and its relative position • 01 ' possible 2) PSFC, Massachusetts Institute of Technology, Cambridge, MA 02139, USA " !vplasma = R (3) -3 mode 28 GHz beam will be Arc detector 24q L -15 13 cm density cut off limit (tokamak regime) • The TE01 ’ = 3.832 guided to LDX in 32.5 mm diameter Q 0n root of J’ (Q Launcher ∂t – 10 • Higher density will increase E and improve for TE01, Q 01 0 2 to the levitated coil and the vessel walls. 32.5 mm diameter Allows different control methods to be Catcher PS voltage feedback fast Arc detector guided to LDX in Vacuum window waveguide ’ ‣ 0 X position thermal equilibration with ions D for TE01, Q 01 = 3.832 24q Non-linear evolution of Magnetic waveguide a =16.25 mm – One TE01 corrugated bend coi implementede-mail contact of main author: [email protected] -2 ! II. RESULTS corrugated bend Loops Small operating space of reliable gains Vacuum window – One TE01 waveguide radius – Aprox. 5 m straight waveguide a =16.25 mm D interchange may lead to φ -4 Enable new physics investigations • Higher density will improve wave length to LDX port waveguide radius F-Coil mm Outer Flux Loop (mV sec) ∂N ∂ ∂N – Aprox. 5 m straight waveguide ‣ Matlab/Simulink Opal-RT development -6 and improve propagation1 MW for planned ICRF ICRH experiments Transmitterlength to LDX port (3-28 MHz) – Water cooled fused quartz window Abstract. We report the first production of high beta plasma confined in a fully levitated laboratory dipolePLC using Full state control under development A. Experimental Setup The shots analyzed are all deuterium plasmas, and the E Gyrotron tube convective cells Pastukhov, Chudin, Pl Physics 27 ‣ = S + D , (1) – Valsov antenna inside LDX -8 short-time “stripey” and correlation plots are all from • Higher density will increase – Water cooled fused quartz window in magnet environmentneutral gas fueling and electron cyclotron resonance heating. The pressureQNX results primarily from a population of -10 #$! ∂t ∂ψ ∂ψ Real Time Computer Low Speed I/O New magnetic pickup loops to improve periods where all heating sourcesInc arerea on.sed TheThe phasermal Energy thermal equilibration with ions – Valsov antenna inside LDX Beam Dump Water• Ideal Load straight copper waveguide Beam Dump Water Load energetic trapped electrons that is sustained for LDXmany Plasma seconds of microwave heating provided sufficientCoil and Bus neutral Temperature 2 CurrentGyrotron ECRH heating tube capability is a ‣ Explore non-linear (RTC) velocities were all calculated during periodsDurin whereg Lev ev-itation Probe Array Probe Array (@ R = 1 m) transmissiontesting loss andes ~0.005 dB/m 32.5 mm internal diameter and 10 mm ‣ 4 kHzgas feedback is supplied loop to the plasma. As compared DAQ to previous System studies in which the internal coiland was Joint supported,Voltage Montoring plasma isolation Y position inmaximum magnet of 17 kW emented for High Speed I/O 0 Early data from floating probes at the LDX edge in- ery heating source was on as well. Since the potential • Ideal straight copper waveguide 32.5 mm internal diameter and 10 mm wall copper bus bar tubing for wal Interlocks where S is the net particle source within the flux-tube, and the diffusion es ~0.005 dB/m A beam dump will be impl • Main transmission losses due to: A beam dump will be implemented for testing and levitation results in improved particle confinement that allows higher-density,Feedback Loop high-beta discharges to be Higher density will improve wave transmission loss wall copper bus bar tubing for straight waveguide run and Vlasov evolution of interchange is measured by the probes, the azimuthal/toroidal elec- • -2.45 GHz magnetrons, 2.5 & 1.9 kW – Tilt and offset errors antenna. calibration between LDX plasma campaigns ‣ Failsafe backup for upper fault (Extended) Kalman filter -2 dicated that low-frequency potential fluctuations occur !$" 2 2 2 straightNew waveguide Vacuum run and Port Vlasov Installationcalibration between LDX plasma campaigns
maintained at significantly reduced gas fueling. Elimination of parallel losses coupled with reduced gas leads to mm losses due to: tric field can be directly determined from the gradient coefficient is D = R E τcor in units of (V sec) /sec. propagation for planned ICRH experiments 1 MW, 3-28 MHz CW -6.4 GHz klystron, 2.5 kW • Main transmission antenna. – Waveguide bend like instabilities improved energy confinement and a dramatic change in the density profile. Improved particle confinement Hot Electron Energy ϕ 28 GHz ECRH system is being rapidly implemented on LDX r Coupling between coils affected by plasma -4 [5], but the probes were too distantly spaced to properly of the potential as shown in equation 1 [6]. The electric · -10.5 GHz klystron, 10 kW – Tilt and offset errors – Window assures stability of the hot electron component at reduced pressure. By eliminating supports used in previous Will be available for next plasma campaign this year Programmable Logic Controller LDX -6 field can be used to compute the E × B drift velocity,Su aspported Dipole Waveguide bend Operator capture the phenomenon with sufficient spatial resolu- !$! transmitter – Near marginal stability, studies, cross-field transport becomes the main loss channel for bothData the hot and the backgroundST species.OP beta! shown in equation 2 [7] . These drifts are directly related Current ECRH heating capability is a Interface 0 2000 4000 6000 8000 New Vacuum Port Installation – Window Tilt and Diameter Change ‣ SlowInterchange fault conditions stationary & density Interlocks profiles, corresponding to an equalSystem number of particles per flux tube, are Time (sec) tion. A vertically movable probe array with tight angu- to convection,137 andGH measuringz Radiom theireter magnitudeTemperatu showsre (e itsV) (c) Line Density from Supported and Levitated Plasma maximum of 17 kW From Archimedes isotope Density Increases with ECRH Power Drilling rig for Losses convective cells do not commonly observed in levitated plasmas. strength(! in transporting particles to and from the core of ‣ Slow wave (Stix) heating coi LDX Control Room lar spacing has been constructed to provide more spatial -2.45Levit GHzated magnetrons, 2.5 & 1.9 kW 28 GHz ECRH system is being rapidly implemented on LDX new vacuum port ‣ Coil and Bus temperature and resistance the plasma. plasma campaignData to date thisshows year electron density increasingTilt linearly and Diameter Change Waveguide overmoding is small, D/O = 3.2, necessarily transport energy φ -6.4 GHz klystron, 2.5 kW seperation projectWill be available for next with good tolerance for imperfections monitoring information in hopes to differentiate whether the poten- '! with ECRH power Losses ‣ bulk plasma ICRF heating -10.5 GHz klystron, 10 kW O = 3.2, Tilt mode conversion to TE , TE , TM Ricci, Rogers, Dorland, PRL 97, ' Edge Density Core Density 11 12 11 1. Introduction tial fluctuations are convection related or related to some E! = −∇Φ (1) Drilling rig for Waveguide overmoding is small, D/ Isolating the Plasma Mutual Inductance &! ‣ Donated to MIT by General Current peak estimate 0.5 – 0.8 x 1012 cm-3 P 2 (Eq. 1) new vacuum port with good tolerance for imperfections requires high field launch 063. o 76 cm other phenomenon. The electrostatic probe array is com- E! × B! LDX Vacuum ‣ 76 cm T 1 tilt corresponds to , chamberTE12, TM11 P !vE×B = (2) Atomics 10 -3 11 o 0.4% loss The dipole confinement concept [1, 2] was motivated by spacecraft observations of planetary B2 3 x 10 cm Tilt mode conversion to TE posed of twenty-four cylindrical, tungsten-tipped floating %! & [T in radians] Effect of Plasma on Levitation System FIG. 1: A top-down view of LDX at the midplane. This view ∂θ Density Increases with ECRH Power 2 magnetospheres that show centrally-peaked plasma pressure profiles forming naturally when 25 msec P o Beam trapping Possible Fusion Power Application shows the extent of the probe array and its relative position !vplasma = R (3) 063. T 1 tilt corresponds toMultiple antenna options Diameter change mode conversion to TE Testing a New Approach to Fusion and potential Langmuir probes.Plot The of array edge is floating positioned probe at a array, showing quasi-coherent structure of edge n density increasing linearly‣ Will be made available to LDX Beam trapping 12 Commercial (CPI) corrugated ! ∂t Gyrotron cabinet cavity with PTFE the solar wind drives plasma circulation and heating. Unlike most other approaches to to the levitated coil and the vessel walls. FIG. 2: Stripey plot showing the electric field in color and Data to date shows electro P 0.4% loss TE bend. A properly built bend Diamagnetic plasma increases effective mutual Electric Field (V/m RMS) with ECRH power o cavity with 2PTFE 01 water line radius of one meter and sweeps out a ninety degree arc at % with DOE ARRA stimulus Pr§·' 0.02 inch change without mode conversion would have the drift velocity as a vector field. The root of the arrows -3 [ in radians] ‣ Direct heating at plasma axis (21 MHz) turbulence. Dominant is a 2 kHz, m=1 mode, rotating in the electron diamagnetic Core Density 12 T water line corresponds to losses corresponding straight TE magnetic confinement in which stability requires average good curvature and magnetic shear, cm ¨¸157. 01 Hoist inductance of L and F coils Edge Density LDX Vacuum Views of recent12 (lastCommercial week) 28 (CPI) GHz corrugated ECRH system installation activity at LDX Pa0.06% loss waveguide. 33 cm the bottom of the LDX vacuum vesselA. Experimental as shown Setup in figure emanates from the point where the drift occurs and the length Current peak estimate 0.5 – 0.8 x 10 chamber Total Power Total Power bend. A properly built bend o ©¹ Laboratory Plasma Confinement MHD stability in a dipole derives from plasma compressibility [3–5]. At marginal stability drift direction. (Which is also the ExB#! !direction.) Assuming the plasma potential funding 15 kW Diameter change15 kW mode conversion to TE TE01 ! 2 without modesimilar conversion to would Phadreus have end cell 2 " S80515043 1. This gives a probe separation of about 7 centimeters, of the arrow shows the strength of the drift relative to the 0.02 inch change 01 33 cm Levitated Dipole Reactor !(pV ) = 0 (withFlux Loopp the#5 (mV plasma s) pressure, Use internal and external flux loops ' losses corresponding straight TE 0 Early data from floating probes at the LDX edge in- 10 cm-3 Supported Gyrotron cabinet Pr§·corresponds to giving the array a resolutionis a of constant 14 centimeters. offset This from res- the floatingpoints. potential gives a measure of the electric 3 x 10 Installation underway ¨¸157. waveguide. Internal ring Kesner, et al. Nucl. Fus. 2002 V= " dl /B dicated that low-frequency potential fluctuations occur ECRH Pa©¹0.06% loss ‣ Nagoya type III loops on upper launcher -2 Levitation Coil ‣ Feedforward on plasma energy from olution corresponds to a detectable[5], but the probes mode were too number distantly spaced range to properly "! o Internal flux field. Instantaneous radial ExB drift speeds of 35 km/s approach the sound ! allation activity at LDX -4 external loop capture the phenomenon with sufficient spatial resolu- ' ‣ Structural engineering of 28 GHz ECRH system inst of 1 to 46. The flow velocitiestion. A verticallyarising movable from probeE × arrayB withdrifts tight angu- An eight millisecond period from shot 81003019 is ‣ Other designs -6 Views of recent (last week) 40 loop pickups speed.lar spacing hasMeasured been constructed azimuthal to provide more electric spatial field fluctuation levels and correlation times Steady state Flux Loop #10 (mV s) can be measured by computing the electric field from the shown in! figure 2. It is a “stripey plot,” which is con- Total Power penthouse15 kW completed ! 30 ‣ Use coil position sensitive internal magnetic information in hopes to differentiate whether the poten- now used in plasma potential. The vertical height of the array is re- structed byInte plottingrferometer ( eachRadian) signal from the probe array & 3 msec Total Power 15 kW Antenna coupling studies (FY10) 20 loops for velocity feedback maytial fluctuations be used are convectionto estimate related or relatedturbulent to some diffusion( coefficient. OK to install 10 feedback motely adjustable, and canother be phenomenon. changed The during electrostatic or between probe array is com-side by' side and representing the signal amplitude with posed of twenty-four cylindrical, tungsten-tipped floating ‣ Inital experiments to study plasma Non-interlocking coils 0 545 kg 20 kg Increased Central Line Density Both require calibration for effective use shots forShots probe90312001−90312044 insertion scans.potential The Langmuir probe probes. modularity The array is positioned al- at a & % -10 color.FIG. 2: The Stripey stripey plot showing plot the electric shown field in in color Fig. and 2 is similar to ‣ Clearing out old TARA 290 Need Capacitors? coupling using low power ( ~ 1 kW) ham radio HF Levitation Control (kA Turns) lows other kinds of probesradius to of be one inserted meter and sweeps into out the a ninety array degree arc atmostthe of drift% the velocity shots as a vector taken field. and The displays root of the arrows the features set out 285 the bottom of the LDX vacuum vessel as shown in figure emanates from the point where the drift occurs and the length power supplies 12 Deuterium Shots ! at future date. 1. This gives a probe separation of about 7 centimeters, of the arrow shows the strength of the drift relative to the ! amplifier 280 5th degree polynomial to examine here. Therefore, it is the only one reported. Good field utilization givingHelium the Shots array a resolution of 14 centimeters. This res- points. 5.0 5.2 5.4 5.6 5.8 "$!! "$!% "$!& "$"! "$"% "$"& -10 kA 5 1.1 MA 10 6 kA The potential signals from the probes are read from The magnetic field is about 360 Gauss at the probe tips VECTORWORKS EDUCATIONAL VERSION 275 olution corresponds to a detectable mode number range time (s) )*+, -./ ‣ allow multiple designs to be analysed while 270 of 1 to 46. The flow velocities arising from E × B drifts An eight millisecond period from shot 81003019 is 100 Tara Tandem Mirror 6 the probe tips by inverting operational amplifiers with and this yields a computed average radial E × B drift Dipole Position (mm) 8 can be measured by computing the electric field from the shown in figure 2. It is a “stripey plot,” which is con- transmitter is installed Possibility for τE > τp 4 1.1 MΩ input resistors andplasma a potential. 10kΩ Thefeedback vertical height resistors. of the array is re-velocitystructed over by plotting the plottedeach signal periodfrom the probe of -0.11 array mm/s. The in- pulsed power supply 2 1 m motely adjustable, and can be changed during or between side by side and representing the signal amplitude with Turbulent startup of levitated plasma show pinch time consistent with expected + 5mm 6 This gives the amplifiers a gain of -0.0098, and the very stantaneous velocity, however, ranges from 34.7 km/s to capacitors. “Cadillacs” 0 shots for probe insertion scans. The probe modularity al- color. The stripey plot shown in Fig. 2 is similar to diffusion coefficient based on characteristics of edge fluctuations. See Mauel, -2 2.45 GHz Phase Velocity (km/s) high input resistance causeslows the other probes kinds of probes to float to be inserted (draw into al- the array-20.8most km/s, of the shots depending taken and displays on position the features and set out time. The sound Figure 4 -4 4 at future date. to examine here. Therefore, it is the only one reported. Advanced fuel cycle 20 ECRH Power (kW) most no current). The signals are digitized by two 16 speed for cold ions and for 25 eV electrons as measured UP8.00051 45 kV, 17 uF 15 6.4 GHz The potential signals from the probes are read from The magnetic field is about 360 Gauss at the probe tips km 60 m the probe tips by inverting operational amplifiers with and this yields a computed average radial Es × B drift e i 2 ∼ 10 γ 2 channel, 16-bit INCAA Computers TR10 bipolar digitiz- on a swept Langmuir probe is c = c T /m c = 35 s 5 p V and 1.1 MΩ input resistors and a 10kΩ feedback resistors. velocity over the plotted period of -0.11 mm/s. The! in- All must go! Get yours 500 MW 0 ∝ ers sampling at 80kHz. [8]. The E × B drift velocities have a near-zero mean are 1 This gives the amplifiers a gain of -0.0098, and the very stantaneous velocity, however, ranges from 34.7 km/s to 0 5 n 10 V − . 15 20 FIG. 1. Schematic of LDX device showing 2 3 4 5 6 7 8 9 10 today! 3 high input resistance−6 causes the probes to float (draw al- -20.8 km/s, depending on position and time. The sound D-D(He ) Fusion time∝ (s) electron cyclotron resonance zones configuration. Neutral Pressure (Torr) x 10 most no current). The signals are digitized by two 16 speed for cold ions and for 25 eV electrons as measured FIG. 7: The cross-correlation of probe channel 2 with all2 oth∼ erkm channel, 16-bit INCAA Computers TR10 bipolar digitiz- on a swept Langmuir probe is cs = c Te/mic = 35 s ers sampling at 80kHz. probe channels at 6 seconds[8]. The E in× B shotdrift velocities 81003019. have a! near-zero All heating mean are FIG. 5: The plasma rotational velocity plotted against the sources are on at this time, and the plasma phase velocity is 16 Inductive vessel neutral pressure. There is a maximum at 5µTorr m s .Theauto-correlationofchannel2(thelargestbluecurve) VECTORWORKS EDUCATIONAL VERSION where the plasma density is greatest. Phase velocity decreases shows the time it takes for a signal to decorrelate (go from 1 Charging with neutral pressure when the plasma density is constant ( to <.2) is about 15 data points or 190 µs. The correlation > µ µ 5 Torr). Below 5 Torr, the plasma density is not constant length can be seen by the number of probe lengths away the 2m with neutral pressure, and the velocity/pressure relationship signal takes to decorrelate. Here it is about 8 probe lengths,or is unclear. 54 cm. The second “bump” around 31 lag points corresponds to one rotation around the vessel and is very poorly correlated
to the original signal. “Reconstruction of Pressure Profile Evolution during Levitated Dipole Experiments”
waves. If the correlation velocity is greater than the phase velocity, the structures in the plasma decorrelate faster than a rotational lifetime and the structure can be re- garded as transient. If the correlation velocity is much See Poster (NOW!) CP6.00089: slower than the phase velocity, the structure can con- versely be regarded as static. As shown in figure 7, a typical correlation lengths and times in LDX are 52 cm and 150 µs, respectively. This corresponds to to a corre- km lation velocity of 347 s ,whichismuchfasterthanthe km FIG. 6: Plasma density vs. neutral pressure. Measured with phase velocity of 16 s at this time. Therefore, the edge the microwave interferometer on LDX. The plasma density of LDX is turbulent since the structures present are very levels out at 5 µTorr for deuterium plasmas.[9] transient and change configuration much faster than the bulk flow. Figure 8 shows the floating potential and electric field sure above 5µTorr, but there is another trend at lower values over the course of shots 81002020 and 81002027, pressures. The velocities lower than 5µTorr decrease and respectively. Shot 81002020 had the floating coil fully lev- then increase again at lower pressures. This may be be- itated, and shot 81002027 had the floating coil physically cause the plasma density is not constant below 5µTorr, supported. Results by Garnier, et al. have shown that as seen in figure 6, and may give rise to a more com- levitation causes the density profile in LDX to peak in plicated relation between neutral pressure and phase ve- the core revealing an inward pinch [9]. The particle loss locity. Above 5µTorr, the plasma density seems to be to the supports is fast since the plasma stays “hollow,” relatively constant, and the velocities decrease with neu- or with a density drop off in the center, during supported tral pressure as expected. mode. The sonic (collisional) loss time to the coil goes The auto and cross-correlations of the probe signals like L(Bmin/Bmax)/Cs,whereL is the radial distance at characterize the spatial and temporal dynamics of the the midplane. This time is about 100-200 µsnearcoil structures at the edge of the plasma. The probe array and 2.5 ms at L = 1.6 m. This is fast enough so that samples a slice of these structures, and if the slice has radial turbulent pinch cannot overcome the field-aligned very long correlation lengths and times, we can deduce losses. that the system is not turbulent since a static structure This indicates that, although both plasmas exhibit stays intact for many rotational lifetimes. A “velocity” similar edge turbulence, in the supported case the loss can be determined from the correlation length and time to the supports dominates over the pinch. This is an that can be compared to the phase velocity of individual interesting result because without these measurements