Non-solenoidal Startup in PEGASUS Discharges A.J. Redd, D.J. - - PowerPoint PPT Presentation
Poster NP6.00135 Non-solenoidal Startup in PEGASUS Discharges A.J. Redd, D.J. Battaglia, M.W. Bongard, R.J. Fonck, E.T. Hinson, B.A. Kujak-Ford, B.T. Lewicki, A.C. Sontag, and G.R. Winz University of Wisconsin - Madison 1500 Engineering Drive
Poster NP6.00135
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 2
solenoidal startup techniques for ST and tokamak applications.
(plasma guns) produces discharges with toroidal current Ip up to 50 kA, using
for radial force balance, also providing inductive current drive, and have achieved Ip=80 kA using less than 2 kA of injected current.
shutoff, while the plasmas relax into typical tokamak equilibria with well- defined edges.
determined by the helicity injection rate, radial force balance, kink stability, and the Taylor relaxation criterion.
allowing higher Ip and normalized current IN, and enabling both flux amplification studies and predictive testing of the Ip model.
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 3
– Formation of the startup plasma saves limited Ohmic transformer flux – May enable high-IN, high-β studies on PEGASUS – Enables completely solenoid-free operation
– Biased plasma guns produce low-impurity plasma – Gun assemblies can be placed at any experimentally convenient location – Power supplies, gun design, operating scenarios, and the underlying theory are presently being studied on PEGASUS
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 4
Ohmic Trim Coils RF Heating Anntenna Vacuum Vessel Centerstack:
Exposing Ohmic Heating Solenoid (NHMFL)
Equilibrium Field Coils Plasma Limiters Toroidal Field Coils
Experimental Parameters Parameter Achieved Goals A 1.15-1.3 1.12-1.3 R (m) 0.2-0.45 0.2-0.45 Ip (MA) 0.18 0.30 IN (MA/m-T) 6-12 6-20 RBt (T-m) 0.06 0.1
1.43.7 shot (s) 0.02 0.05 t (%) 25 > 40 PHHFW (MW) 0.2 1.0
– High-IN, high-β operating regimes – ELM-like edge MHD activity (see Bongard, poster NP6.00136)
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 5
Magnetic helicity is a measure of the linkage between magnetic fluxes (or, equivalently, the currents that generate those fluxes). The general definition of magnetic helicity is an integral over a volume that encompasses the linked fluxes: Magnetic helicity is the best-conserved constant of motion in magnetized plasma, decaying on resistive timescales. In the case of two linked but distinct fluxes φ and ψ, similar to the rings shown, the total magnetic helicity of the volume is K=2φψ. In a tokamak, the magnetic helicity K is proportional to the product ITFIp, with ITF determined by the TF coil power supply. Increases in the helicity K correspond to increases in the toroidal plasma current Ip.
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 6
V
A
– E = ηJ → much slower than energy dissipation (ηJ2) – Turbulent relaxation processes dissipate energy and conserve helicity
A
V
Total helicity in a tokamak geometry:
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 7
two divertor- mounted guns are shown.
filaments can relax to form a tokamak plasma.
formation and sustainment of a tokamak plasma.
Shot #37460 Shot #37222
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 8
– Plasma current persists after Ibias=0 – Coil currents static (no PF ramps) – BT=11 mT at plasma magnetic axis – Vacuum vertical field is 7 mT
hallmark of significant relaxation:
– Coincident with the current multiplication ratio (defined M=Ip/Ibias) exceeding the vacuum-field current windup factor G
– Consistent with flux amplification – Vacuum-field windup in #32606 was 5
Shot #32606
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 9
– Washers stabilize arc while limiting surface contact [1] – Sputtered high-Z impurities mostly trapped in gun cavity
space charge limitations [2]
power supplies.
– Iarc = 2 kA using a pulse forming network – Varc = 100 - 500 V – Ibias < Iarc for impurity sputtering – Ibias feedback-controlled in real time – Vbias up to 2000 V
[1] Den Hartog, D.J., Plasma Sources Sci. & Tech. 6 (1997) [2] Fiksel, G, et. al., Plasma Sources Sci. & Tech. 5 (1996)
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 10
– 2D filament calculations illustrate the formation of a field null. – Experimentally, variations in vertical field and/or gun-driven current (equivalently, variations in the total toroidal current in the gun-driven streams) lead to sharp boundaries between relaxation and no-relaxation cases.
– The relaxation process is not stopped by changes in the vertical field (e.g., to maintain radial force balance) – Plasma position (R0), size (a), shape (ε)… can be modelled – The plasma current simultaneously satisfies four conditions: (1) Radial force balance (2) Tokamak stability (vs kinks, etc) (3) Helicity balance (injection vs dissipation) (4) Taylor relaxation requirement
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 11
ITF = 300 kA, IPF = 1.2 kA (PF1-3, 6-8) Calculations by D.J.Battaglia Iinj = 0 A Iinj = 2 kA
M > G correlates with inboard B field reversal
– Treat the discrete filaments as a toroidally averaged current sheet tied to a flux surface
reversal when IPF ~ 1.2kA – Bv ~ 0 at inner sheet edge Force-free plasma filaments perturb the vacuum magnetic field
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 12
Helicity balance in a tokamak geometry:
dK dt = 2 J B d3x
V
t 2 B ds
A
confinement via the η term
Taylor relaxation of a force-free equilibrium:
B = µ0J = B p edge µ0Ip µ0Iinj 2RinjwB
,inj
Assumptions:
Ap Plasma area Cp Plasma circumference Ψ Plasma toroidal flux w Edge width
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 13
Estimated plasma evolution Plasma guns Anode Relaxation limit Helicity limit Ip max Time
ITF = 288 kA Vbias = 1kV Vind = 1.5 V Iinj = 4 kA w = dinj L-mode τe
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 14
– Each point represents one discharge – Includes one and two gun operations – Max Ip for given conditions achieved at max Bv that allowed flux reversal
roughly approximates average Te
– Assume Spitzer η & Zeff = 2.5 – Calculated average Te = 35 - 65 eV, assuming L-mode confinement
achieve helicity equilibrium
– Vsurf estimated using a flux measurement at center column – G-S solver provides plasma geometry for Veff calculation
65 eV 45 eV 35 eV 13 eV Te = ˙ K
DC /ITF = µ0NinjVinjAinj /Rinj
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 15
120 V 900 V 1200 V Relaxation limit R = 47 cm Set by gun limit Varied via neutral fueling
Vbias = Zinj Iinj
Thus, As a general trend, the maximum Ip
increases with bias voltage Vbias
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 16
(e.g., due to the lack of a field null), there is no significant MHD activity measured in the magnetic probe arrays.
– Typically n=1, with frequencies varying from 20 kHz to 60 kHz – Observed mode can be either continuous or intermittent – Mode frequency and amplitude may vary with magnetic field
relaxation in CHI-driven spheromaks and STs:
– Computational work by X.Tang, D.Brennan, and A.Bayliss (see Refs) each provide detailed studies of current-profile relaxation through the action of low-n modes – A similar mechanism may be responsible for relaxation in PEGASUS
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 17
divertor
field line connecting gun & anode
centerstack limited
easily retrofitted into PEGASUS
Zero current plasma filaments in vacuum magnetic field
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 18
require significant coil-current ramps
– To maintain radial magnetic stability, and – To reach typical Ohmic operating parameters (e.g., increased TF to maintain kink stability).
– Form with typical Ohmic parameters (relatively high TF). – More easily couple to outer-PF induction. – Be more accessible to diagnostics – Have longer L/R decay timescales
a more optimum injector configuration for PEGASUS.
– Further studies could indicate the optimum injector configuration for another device, such as NSTX.
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 19
Anode 3 plasma guns Side view Top view (rotated) Anode Plasma guns Limiter Plasma streams Anode Outer Limiter Plasma guns 40 cm
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 20
Pegasus shot #40458: two midplane guns, outer-PF ramp, typical discharge
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 21
no PF ramps.
above 50 at gun shutoff.
rampup correspond to “bursts” of low-n MHD activity, which may also correspond to low-order rational values for the edge-q.
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 22
Single gun, static B field discharges
Shot 39762: ITF = 300 kA, IPF = 1.3 kA Shot 39761: ITF = 300 kA, IPF = 1.2 kA
Compare two single-gun discharges:
while the other did not (#39762)
which prevented relaxation
– Vacuum-field windup G ~ 2 (as in #39762), but the relaxed discharge achieved current amplification M > 12 – Long Ip decay after gun shutoff at 28 ms – ∫ ne dl time trace shows improved particle confinement over unrelaxed case – Line averaged density in relaxed plasma is near the Greenwald density limit
Rtang = 8 cm
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 23
Three-gun discharge #42292:
bank to drive gun bias current
after relaxation begins
at gun shutoff
region at gun shutoff; afterwards, plasma is centerstack-limited
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 24
– Low Bv required for field reversal – Larger Bv required after field reversal to maintain radial force balance with larger Ip
toroidal loop voltage, comparable to the Veff from relaxation
– Two plasma guns in operation – Ramp begins after field reversal
– Veff calculated using R0 and plasma shape estimation from magnetic measurements – VPF calculated using 2-D vacuum field model that includes wall effects
Shot 40540
~ ~
R = 50 cm, z = 0 cm Calculated
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 25
Slower PF ramp Plasma detaches at 29.4 ms Faster PF ramp Plasma detaches at 25.3 ms After detachment, current drive is purely inductive and MHD activity is reduced.
Estimated Estimated
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 26
Both discharges had gun-driven startup, outer-PF ramps, and applied Ohmic drive. The peak current Ip in each discharge was approximately 90 kA.
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 27
6 4 2 vloop (V), Iinj (kA) 35ms 30 25 20 time 6 4 2 vloop (v) 0.15 0.10 0.05 0.00 Ip (MA) 0.15 0.10 0.05 0.00 Ip (MA)
Ip vloop Iinj plasma gun startup OH only 41708 41536
handoff to OH drive
single-swing reaches the same peak Ip as Ohmic double-swing with twice the flux
– Implies ~ 50% flux savings
gun-driven discharges for
– 0.8-1.0 MW PEGASUS HHFW
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 28
drive provides tool for modifying the current profile
using non-inductive startup
at high IN
– Discharges have been limited by available current drive
2
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 29
SPRED spectra from various times in gun-driven Pegasus discharges
#40458 at 27.04 ms Two guns + PF ramp OV dominated #41708 at 28.288 ms Three guns, PF ramp, OH OV dominated, some NIV #42292 at 28.288 ms Three guns + PF ramp OV and NIV dominant
The SPRED spectra for gun-driven discharges do not show significant metal contamination, and are dominated by OV (113.9 eV), implying electron temperatures of 50-70 eV. The NIV emission may result from the plasma limiting on the boron nitride gun casing. It is possible that operational improvements will reduce or eliminate this nitrogen impurity. NIV OV OV
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 30
has been demonstrated on the Pegasus Toroidal Experiment
– Formation of tokamak-like plasma correlates with inboard field reversal – Maximum Ip described by force and helicity balance – Observed in relaxed gun-driven discharges:
– Outer-PF induction provides current drive and maintains radial force balance with larger Ip plasma – Fast PF ramps can cause the tokamak-like plasma to detach from the gun – Handoff to OH induction is robust – Plasma-gun startup is equivalent to a Pegasus Ohmic half-swing
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 31
– Jedge broadening due to magnetic turbulence (edge and global), magnetic shear, gun characteristics, physical geometry, etc. – Plan to measure directly using probes and study dependence on ITF, Iinj & plasma properties
– χ⊥ versus χ in the presence of magnetic turbulence – Confinement will depend on degree of stochasticity in core plasma – Requires Te and Ti measurement
– Voltage drop across sheath and filament length – Measure with floating probe in SOL R = 47 cm
Terms may have Ip dependence ex: paramagnetism and shaping
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 32
– PRAD, nel and ne, Te, impurities, flows – Possibly implement Thomson scattering (for ne and Te) and IDS (Ti and flows)
– Langmuir and magnetic probes → measure λedge directly, study relaxation mechanism – Increase gun area → determine effect on w & increase Kinj – Decrease Rinj & maintain outboard injection → should increase both limits – Increase Lfilament → determine effect on Zinj – Possibly implement Thomson scattering and ion Doppler shift → Te and Ti – Increase TF by 20-30% → extend database to higher toroidal field
– Routine equilibrium fitting with KFIT Grad-Shafranov solver – Stability calculations using DCON – PF induction/compression modeling using TSC
A.J.Redd, 50th APS-DPP Meeting, Dallas, Nov 2008 33
in the HIT-II spherical tokamak,” in preparation.