RFX Program on Active RFX Program on Active Control Control at the 9TH WORKSHOP ON MHD STABILITY CONTROL: "CONTROL OF MHD at the 9TH WORKSHOP ON MHD STABILITY CONTROL: "CONTROL OF MHD STABILITY: BACK TO THE BASICS": NOVEMBER 21-23, 2004, PPPL STABILITY: BACK TO THE BASICS": NOVEMBER 21-23, 2004, PPPL Consorzio RFX -Associazione Euratom-ENEA sulla fusione - Padova, Italy Consorzio RFX -Associazione Euratom-ENEA sulla fusione - Padova, Italy Presented by Stefano Martini Presented by Stefano Martini
Transcript
RFX Program on Active RFX Program on Active Control Control
at the 9TH WORKSHOP ON MHD STABILITY CONTROL: "CONTROL OF MHD at the 9TH WORKSHOP ON MHD STABILITY CONTROL: "CONTROL OF MHD STABILITY: BACK TO THE BASICS": NOVEMBER 21-23, 2004, PPPLSTABILITY: BACK TO THE BASICS": NOVEMBER 21-23, 2004, PPPL
Consorzio RFX -Associazione Euratom-ENEA sulla fusione - Padova, Consorzio RFX -Associazione Euratom-ENEA sulla fusione - Padova, ItalyItaly
Presented by Stefano MartiniPresented by Stefano Martini
Hopefully this is the last Hopefully this is the last workshop…workshop…
……. without RFX. without RFX ! !
……. RFX reconstruction is in the final phase . RFX reconstruction is in the final phase and first plasmas are expected in Sept. 2004and first plasmas are expected in Sept. 2004
The new RFX will be a “state of the art” The new RFX will be a “state of the art” MHD MHD MODE CONTROL FACILITY:MODE CONTROL FACILITY:
192 ACTIVE COILS, INDEPENDENTLY DRIVEN, 192 ACTIVE COILS, INDEPENDENTLY DRIVEN, COVERING THE WHOLE PLASMA SURFACECOVERING THE WHOLE PLASMA SURFACE
OPS…OPS…
I FORGOT A SLIDE FROM I FORGOT A SLIDE FROM LAST YEAR WORKSHOP!LAST YEAR WORKSHOP!
• Introduction: RFP and MHD dynamo
• Previous mode control experiments on RFX
• What has been learned from other RFPs
• RFX reloaded
• What we expect to do on the new RFX
OUTLINE OF TALKOUTLINE OF TALK
The RFP dynamo The RFP dynamo
The current profile in a RFP cannot be driven in steady state The current profile in a RFP cannot be driven in steady state by a constant inductive electric field by a constant inductive electric field EEoo
…….but RFP plasmas .but RFP plasmas last much longer than last much longer than the resistive diffusion the resistive diffusion time! (actually, as time! (actually, as long as long as EEo o is applied)is applied)
An additional An additional “dynamo”“dynamo” electric field electric field EEdd is is necessary to maintain the toroidal magnetic necessary to maintain the toroidal magnetic flux.flux.
A wide experimental and numerical database A wide experimental and numerical database supports the MHD turbulent dynamo theory:supports the MHD turbulent dynamo theory:
EEd d is produced by the coherent (non-linear) is produced by the coherent (non-linear) interaction of many MHD modes => interaction of many MHD modes => Multiple Multiple Helicity (MH) dynamoHelicity (MH) dynamo
bvEd~~×=
r
-6-7
-8-9-10
Mode ClassificationMode Classification
RFPRFP
““DYNAMO MODES”DYNAMO MODES”
The standard Multiple Helicity The standard Multiple Helicity RFPRFP
Magnetic stochasticity Magnetic stochasticity allover the plasma !allover the plasma !
The The Phase LockingPhase Locking of many modes results in a non-axisymmetric of many modes results in a non-axisymmetric deformation, the so-called “slinkydeformation, the so-called “slinky””
The slinkyThe slinky
Each mode is associated to an Each mode is associated to an helical perturbation of the helical perturbation of the plasmaplasma
IφJ n
Localised plasma-wall interaction 100 MW/m2
Braking torques by the vessel and field Braking torques by the vessel and field errors causeerrors cause Wall LockingWall Locking of the slinkyof the slinky
804020090180270360t [ms]toroidal angle Φ( )deg#12461601001 cm
Previous MHD control Previous MHD control experimentsexperiments
Previously MHD mode control on RFX based on :
• Reduction of field errors
• Control of modes via the Bφ coils:
• Active control of poloidal current:
Pulsed => PPCD
Oscillating => OPCD
• Active rotation of the locked modes (RTFM)
plasmaVθJ θExternal poloidal current drive External poloidal current drive (first tested on MST) (first tested on MST) transiently quenches the transiently quenches the spontaneous dynamo.spontaneous dynamo.
Pulsed Poloidal Current DrivePulsed Poloidal Current Drive
0
48
12
35 40
(a)PPCD
b /B
θ(10
-4)
~2
2
0.5
1
1.5
25 30 35 40 45 50
τ (m
s) (g)
t (ms)
200
300
Te(
0) (
eV)
(b) Strong reduction of Strong reduction of magnetic fluctuations and magnetic fluctuations and improved confinementimproved confinement
Oscillating Poloidal Current DriveOscillating Poloidal Current Drive
Oscillating poloidal electric field imposed by a GTO switch system
During co-drive phase confinement improves as for PPCD
Edge transport barrier not affected during counter-drive phase
Enforcing the proper phasing during the start-up phase: continuous rotation for the whole discharge
80
40
20
0 90 180 270 360
t [ms]
toroidal angle Φ (deg)
#12350
60
1 cm0
Three mode interactionThree mode interaction
for sufficiently high external field:
ω 1,n+1 − ω1,n = ω 0,1
• high n modes will co-rotate with (0,1) ext. perturbation
• low n modes will counter-rotate
• in general =>
• m=1 modes experience a non linear torque:
Tz1,n ∝ Cnbr
1,nbr1,n+1br
0,1sin(φ1,n+1 −φ1,n −φ0,1)+
Cn−1br1,nbr
1,n−1br0,1sin(φ1,n−1 −φ1,n +φ0,1)
090180270360φ00,1
090180270φ0,LHD
090180φ00,1-φ0,LHD
090180270φ01,7
090180270φ01,8
0901802702040 6080100φ01,11
( )Time ms#12350
090180270φ01,10090180270φ01,9
t (ms)
Dynamics of m=1 modesDynamics of m=1 modes
phase locking of m=0 mode
low n m=1 modes travel backward
harmonic generation
0ext
0LM
0ext0
LM
1,7LM
1,8LM
1,9LM
1,10LM
1,11LM
most restrained m=1 is stationary
high n m=1 modes travel forward
harmonic generation
Te (eV)
Prad (W)
ne (m-3)
Ip (A)
LM (°)
CVI (au)
with rotating with rotating modesmodes
NO rotationNO rotation
t (s) t (s)
1MA Pulses with & without rotation1MA Pulses with & without rotation
The Single Helicity (SH) dynamoThe Single Helicity (SH) dynamo
a theoretically predicted state with a unique a theoretically predicted state with a unique m m = 1 saturated resistive = 1 saturated resistive kink (kink (a pure helix wound on a torusa pure helix wound on a torus), ),
Stationary Stationary LAMINARLAMINAR dynamo mechanism with good helical flux surfaces dynamo mechanism with good helical flux surfaces
MAC-302
Escande et al., PRL 85 (2000)Escande et al., PRL 85 (2000)
Magnetic order with SH dynamoMagnetic order with SH dynamo
Good magnetic flux surfaces in SHGood magnetic flux surfaces in SH
SHSH Turbulent (MH)Turbulent (MH)
Overlapping of Overlapping of many modes !many modes !
Helical states in the experimentHelical states in the experiment
Quasi Single Helicity (QSH) spectra have been observed in Quasi Single Helicity (QSH) spectra have been observed in all RFP devices, under a variety of boundary conditions all RFP devices, under a variety of boundary conditions (Martin, NF 2003).(Martin, NF 2003).
The mode spectrum is dominated by one geometrical helicityThe mode spectrum is dominated by one geometrical helicity
The other modes have still non-zero amplitudeThe other modes have still non-zero amplitude
Summary of old RFX resultsSummary of old RFX results
In a thick and relatively distant shell machine like In a thick and relatively distant shell machine like RFXRFX
Long and high current RFP pulses sustained with a Long and high current RFP pulses sustained with a constant toroidal voltageconstant toroidal voltage
Only internally resonant mode are seen which provide Only internally resonant mode are seen which provide the dynamo action either in a MH or in QSH spectrumthe dynamo action either in a MH or in QSH spectrum
Dynamo modes are always (nearly) locked in phase Dynamo modes are always (nearly) locked in phase and to the wall, resulting in enhanced and spatially and to the wall, resulting in enhanced and spatially localized transport and severe plasma-wall interactionlocalized transport and severe plasma-wall interaction
Advanced operational modes such as OPCD , RTFM Advanced operational modes such as OPCD , RTFM and QSH alleviate the problems, but are not sufficient and QSH alleviate the problems, but are not sufficient to reach the design target 2MA regime.to reach the design target 2MA regime.
Therefore, with the aim of Therefore, with the aim of
making a new step (hopefully) making a new step (hopefully)
forward along the path to the forward along the path to the
“good RFP” we now have…“good RFP” we now have…
RFX reloadedRFX reloaded
toroidal coil new SS toroidalsupport structure
shell clampingbands
shell equatorialgap shortcircuits
vacuum vessel3 mm
copper shell
active coilsystem
48x4 active coil system 100% surface coverageFirst plasma with the new assembly mid December 2004
Major radius 2 m
Minor radius 0.46 m
Plasma current 0.3-1.1 MA
τpulse 100-200 ms
τshell 50 ms
Mode dynamics in RFPsMode dynamics in RFPs
Experimental evidence in several RFPs shows that Experimental evidence in several RFPs shows that the evolution of MHD modes, including the the evolution of MHD modes, including the dynamo modes, depends on dynamo modes, depends on the magnetic the magnetic boundaryboundary, and in particular on the , and in particular on the shellshell::
- Rotating (10-30 kHz)- small Br component- eventually wall locked
Time (s)
N.B.: Br component #16874
Am
plit
ude
(a.u
.)
0
50
100
0 0.01 0.02Time (s)
Plasma current (kA)
Conclusions on mode rotationsConclusions on mode rotations
Dynamo modes are spontaneously rotating in RFP Dynamo modes are spontaneously rotating in RFP devices with close fitting shell (even if a threshold devices with close fitting shell (even if a threshold in current might exist)in current might exist)
Externally resonant RWM are seen in RFPs. Their Externally resonant RWM are seen in RFPs. Their growth time agrees with the shell time constant growth time agrees with the shell time constant and they do no rotateand they do no rotate
There is a reasonable basis for spontaneous There is a reasonable basis for spontaneous rotation of dynamo modes in the new RFXrotation of dynamo modes in the new RFX
Mode control experience in T2RMode control experience in T2R
As shown in previous talks byAs shown in previous talks by
During the last year a fruitful collaboration between During the last year a fruitful collaboration between
the T2R and the RFX groups permitted to perform the T2R and the RFX groups permitted to perform
very interesting experiments on active mode control very interesting experiments on active mode control
on T2Ron T2R
Feedback experimentsFeedback experiments on T2Ron T2R
Tearing mode rotation is Tearing mode rotation is maintained with feedbackmaintained with feedback
From: P. R. Brunsell, et al., “First results from intelligent shell experiments with partial coil coverage in the EXTRAP T2R reversed field pinch”, 31st EPS, ECA Vol.28G, P-5.190 (2004)
See also: P. Brunsell et al., Feedback stabilization of multiple resistive wall modes, to be published on PRL
No feedback Feedback all modes
(intelligent shell)
Feedback all modes except -2 ≤n ≤+2 (wise shell)
Amplitude and angular phase velocity for resonant tearing mode m=1, n=-14.
plasma current
angular phase velocity
amplitude
Open loop control of RWM with Open loop control of RWM with rotating perturbationrotating perturbation
With sufficient With sufficient amplitude of amplitude of external external perturbation, RWM perturbation, RWM can be rotated can be rotated
Amplitude no longer Amplitude no longer increase with increase with rotationrotation
Rotating n=6 perturbation:. Freq.=100 Hz, 0=0, different amplitudes.
Ampl
Phase
Open loop experimentsOpen loop experiments11 clearly showed intrinsic error field clearly showed intrinsic error field
reduction or amplification, depending on amplitude and phase reduction or amplification, depending on amplitude and phase
of the applied external field. of the applied external field.
Feedback operations have shown clear reduction of MHD mode Feedback operations have shown clear reduction of MHD mode
amplitudes and beneficial effects on plasma-wall interaction. amplitudes and beneficial effects on plasma-wall interaction.
Intelligent shell and individual/ multiple mode control schemes Intelligent shell and individual/ multiple mode control schemes
successfully implemented and comparedsuccessfully implemented and compared
Active rotation of single mode demonstratedActive rotation of single mode demonstrated
T2R experience with mode T2R experience with mode controlcontrol
1 J.R. Drake - Open loop control experiments in EXTRAP T2R RFP, yesterday talk
RFX reloadedRFX reloaded
Main new components:Main new components:
1.1. new toroidal field power supplynew toroidal field power supply
2.2. first wall with higher power handling capabilitiesfirst wall with higher power handling capabilities
3.3. smoother and thinner shellsmoother and thinner shell
4.4. 192 saddle coils, covering the whole plasma 192 saddle coils, covering the whole plasma boundary, each independently powered and boundary, each independently powered and feedback controlledfeedback controlled
5.5. in-vessel system of magnetic and electrostatic in-vessel system of magnetic and electrostatic probesprobes
m=0 x RTFM & PPCD/OPCDm=0 x RTFM & PPCD/OPCD
Btor new ≈ 6 x Btor old
at 25 Hz with 3kA
Compared to RFX 92 RTFM capability greatly enhanced,which is of utmost importance since
•RTFM was not possible at high density in RFX92
ThThe new shelle new shell
One 3 mm Cu layer:One 3 mm Cu layer: 1 overlapped poloidal 1 overlapped poloidal
gap: 23° toroidal overlapgap: 23° toroidal overlap 1 toroidal gap on high 1 toroidal gap on high
field sidefield side
Shortcircuited gap: 50 bolted copper plates
Welded gap
New sNew saddaddle coil systemle coil system
each independently powered24 kAt: 400 A x 60 turns
Wide spectrum of Fourier Wide spectrum of Fourier components :components :
Ex-vessel magnetic and thermal Ex-vessel magnetic and thermal probesprobes
BT BP
PProbes between vessel and shellrobes between vessel and shell Integral probes:Integral probes:
4 Rogowski coils for I4 Rogowski coils for Ipp 8 toroidal voltage loops8 toroidal voltage loops 6 poloidal voltage loops6 poloidal voltage loops 32 partial V32 partial VPOLPOL probes for halo probes for halo
currentscurrents
Pick up probesPick up probes ( (40x36x4 mm bi-axial 40x36x4 mm bi-axial probe for Bprobe for BTT and B and BPP enamelled wire coils enamelled wire coils
wound around single core):wound around single core): 48 toroidal x 4 poloidal distribution48 toroidal x 4 poloidal distribution 2 higher resolution poloidal arrays2 higher resolution poloidal arrays
Magnetic probes total number ≈ 650Magnetic probes total number ≈ 650
48x4 saddle probes for B48x4 saddle probes for Brr 48 saddle coils for B48 saddle coils for Brr ( on ( on
eq.plane)eq.plane)
Improve the Improve the design of the plasma front enddesign of the plasma front end First wall power handling capability;First wall power handling capability;
Poloidal gap field error;Poloidal gap field error;
Toroidal gap field error.Toroidal gap field error.
Improve the active control of the MHD modesImprove the active control of the MHD modes Increase the torque applied to the plasma through the m=0 mode Increase the torque applied to the plasma through the m=0 mode
for RTFMfor RTFM
Produce Produce m=1 , n= 1-20m=1 , n= 1-20 single o multiple mode to induce: mode single o multiple mode to induce: mode
rotation , rotation , “single helicity”“single helicity” and to actively control ext.& and to actively control ext.& int. modesint. modes
Minimise the radial field Minimise the radial field at at plasmaplasma boundary => “boundary => “smart shell”smart shell”
Low current scenarioLow current scenario Theory (Guo, Fitzpatrick et al) and experiments Theory (Guo, Fitzpatrick et al) and experiments
(TPE-RX, EXTRAP T2R, MST) suggest that at low (TPE-RX, EXTRAP T2R, MST) suggest that at low current RFX should see spontaneous dynamo current RFX should see spontaneous dynamo mode rotation.mode rotation.
This is suitable to concentrate efforts on RWM This is suitable to concentrate efforts on RWM controlcontrol
High current scenario (> 1 MA)High current scenario (> 1 MA) Better for confinement improvement techniques Better for confinement improvement techniques
(OPCD) and for interaction with “dynamo” modes (OPCD) and for interaction with “dynamo” modes (but higher wall-locking probability).(but higher wall-locking probability).
Passive shell (and EXTRAP T2R experience) might Passive shell (and EXTRAP T2R experience) might postpone RWM issue up to ≈50-100 mspostpone RWM issue up to ≈50-100 ms
Active actions through 192 saddle coils:Active actions through 192 saddle coils:
Apply Apply m m =1 magnetic perturbations=1 magnetic perturbations
Work on individual modes: one at the time or several Work on individual modes: one at the time or several simultaneouslysimultaneously
Realize an intelligent shellRealize an intelligent shell
Zeroing of radial field at the edge to maintain an effective Zeroing of radial field at the edge to maintain an effective close fitting shell.close fitting shell.
Interesting also for QSH studies, since a smooth magnetic Interesting also for QSH studies, since a smooth magnetic boundary facilitates their onset.boundary facilitates their onset.
Drive of mDrive of m=1 magnetic perturbations=1 magnetic perturbations
Apply a monochromatic perturbation to affect one individual Apply a monochromatic perturbation to affect one individual mode:mode:
““pumping” the mode to drive QSH states through helical fields at pumping” the mode to drive QSH states through helical fields at the plasma boundarythe plasma boundary
Feedback stabilization of individual modesFeedback stabilization of individual modes inducing rotation of a single modeinducing rotation of a single mode
Apply several simultaneous geometrical helicities (various Apply several simultaneous geometrical helicities (various nn’s):’s): damping of main “dynamo modes”damping of main “dynamo modes” feedback stabilization of RWM feedback stabilization of RWM breaking phase locking among “dynamo modes” with induction of breaking phase locking among “dynamo modes” with induction of
Plasma pulses start mid December 2004. The Plasma pulses start mid December 2004. The program for 2005 envisages 45 weeks of program for 2005 envisages 45 weeks of operation including 11 weeks of machine operation including 11 weeks of machine commissioning and 34 with plasma.commissioning and 34 with plasma.
Aim of first year is to test most of the Aim of first year is to test most of the increased flexibility of RFX covering many increased flexibility of RFX covering many programs, including a first assessment of programs, including a first assessment of enhanced confinement regimes and active enhanced confinement regimes and active control of MHD instabilities.control of MHD instabilities.
Main objectives of first year Main objectives of first year
Maximise the parameter range for Maximise the parameter range for spontaneous fast rotation of tearing spontaneous fast rotation of tearing modes.modes.
Establish a clear comparison with the Establish a clear comparison with the reference passive operation of the old RFX.reference passive operation of the old RFX.
Explore scenarios for enhanced Explore scenarios for enhanced confinement.confinement.
Active control of MHD instabilities:Active control of MHD instabilities: feedback stabilisation of RWM, feedback stabilisation of RWM, control of single and multiple m=0,1 tearing control of single and multiple m=0,1 tearing
P2 e: feedback control of m=0 modesP2 e: feedback control of m=0 modes
Simulations show that the transport in the q=0 region is Simulations show that the transport in the q=0 region is
determined by amplitude and phases of m=0 and m=1 modesdetermined by amplitude and phases of m=0 and m=1 modes
Acting on the phases of m=0 modes can influence closed Acting on the phases of m=0 modes can influence closed
structures in the reversal regionstructures in the reversal region
Minimising m=0 amplitudes is important also for favouring QSH Minimising m=0 amplitudes is important also for favouring QSH
regimes and weakening m=1 mode non-linear coupling.regimes and weakening m=1 mode non-linear coupling.
P2 g: experiments on QSH statesP2 g: experiments on QSH states
Investigate conditions which help Investigate conditions which help onset of QSH statesonset of QSH states Parameter scansParameter scans Magnetic boundary optimizationMagnetic boundary optimization
Diagnostic improvements Diagnostic improvements will allowwill allow Particle and energy Particle and energy
confinement studiesconfinement studies Ion dynamics measurementsIon dynamics measurements Determination of the plasma Determination of the plasma
flow correlated with QSHflow correlated with QSH
SXR isoemissive surfaceduring a rotating n=6 QSH in
MST
Based on a sequence of 2D tomographic reconstructions
MAC-302
• Aim: suppressing the dynamo and establishing the
confinement and limits for the RFP configuration
• Rationale: 3-D MHD simulations show that a stochastic
plasma reaches a condition of low amplitude modes by
applying proper time-dependent magnetic boundary
conditions (R.A.Nebel et al, PoP 2002)
• The decay rate is about --10/τR
• In RFX a first test of this prediction will be performedIn RFX a first test of this prediction will be performed
P2 h: Self-similar current decayP2 h: Self-similar current decay
Motivations:• study of steady state current drive in RFP• perturbative study of dynamo mechanism
PLANS:PLANS: Use new power supply system to test the concept on RFX Optimize of plasma and power supply parameters Coordinate experimental plans with MST (different
frequencies and plasma regimes)
Range of parameters:Range of parameters:• frequency of the 100-150 frequency of the 100-150
HzHz• amplitude 20 V toroidal amplitude 20 V toroidal
and 25 V poloidaland 25 V poloidal
P2 i: Oscillating Field Current Drive
P3 1b: Active Control of tearing P3 1b: Active Control of tearing modesmodes
Open loop and feedback control on Open loop and feedback control on single/multiple dynamo modes single/multiple dynamo modes including:including:
induction of QSH regimesinduction of QSH regimes drag of locked tearing modes drag of locked tearing modes
into slow rotation and control of into slow rotation and control of their relative phases.their relative phases.
Encouraging results from T2R:Encouraging results from T2R: Pre-programmed helical Pre-programmed helical
perturbations with resonant perturbations with resonant helicities (open-loop)helicities (open-loop)
Feedback suppression of other Feedback suppression of other resonant modes (closed loop)resonant modes (closed loop)
Open loop
Closed loop
5 10 15 20 25
n=12
Time [ms]5 10 15 20 25
n=13
RFX-Tokamak operationRFX-Tokamak operation
With its highly flexible diagnostic and control With its highly flexible diagnostic and control system, RFX could contribute also to studying system, RFX could contribute also to studying MHD mode control on tokamak plasmasMHD mode control on tokamak plasmas
A maximum BA maximum Bφφ = 0.6 T allows us to set up a = 0.6 T allows us to set up a Tokamak with:Tokamak with:
I = 100 kA @ q(a)=3 I = 100 kA @ q(a)=3 A total flux swing of 6-7 Vs permits to sustain the A total flux swing of 6-7 Vs permits to sustain the
plasma for times much longer than the 50 ms shell plasma for times much longer than the 50 ms shell time constant time constant
We are open to suggestions/proposal for We are open to suggestions/proposal for collaborations also in this area! collaborations also in this area!
Hopefully this is the last Hopefully this is the last workshop…workshop…
RFX reconstruction RFX reconstruction is is completed and first completed and first plasmas are expected in plasmas are expected in Dec.Dec. 20042004
The new RFX will The new RFX will isis a “state of the art” a “state of the art” MHD MHD MODE CONTROL FACILITYMODE CONTROL FACILITY::
192 active coils, independently driven, cover 192 active coils, independently driven, cover the whole torusthe whole torus
An extensive experimental program is plannedAn extensive experimental program is planned
We look forward to contributions & We look forward to contributions & collaboration proposals from Youcollaboration proposals from You
OPS…OPS…
I FORGOT A SLIDE FROM I FORGOT A SLIDE FROM LAST YEAR WORKSHOP!LAST YEAR WORKSHOP!
This is the last workshop without This is the last workshop without RFXRFX
