Activities at EXTRAP T2R and MHD control

1
Activities at EXTRAP T2R and MHD control
RUSA, May 5-6, 2009, Uppsala Univ.
Erik Olofsson (PhD stud), Waqas Khan (PhD stud),
Lorenzo Frassinetti, Per Brunsell, James Drake
(presented by Per Brunsell) KTH, EES/Fusion
Plasma Physics
2
Collaborations

• Max-Planck Institut für Plasmaphysik, Garching
• W. Suttrop and ASDEX Upgrade Team Development of
  RWM controller for ASDEX Upgrade
• Consorzio RFX, Padova
• G. Manduchi and RFX-Mod Team, Control system
  implementation at EXTRAP T2R
• R. Paccagnella and T. Bolzonella Mode
  identification experiments at RFX
• UKAEA, Culham/ Chalmers
• Y. Liu, D. Yadikin, RWM modeling with MARS-F code
• UJF-INPG/GIPSA-Lab, Grenoble
• E. Witrant, Control issues
• KTH, EES/Automatic Control
• H. Hjalmarsson, E. Jacobsen, Control issues

3
EFDA work programme

• The research on active control of MHD modes is
  part of the EFDA work programme
• Some parts of the research are carried out
  through Task Agreements coordinated by the
  Topical Group on Stability and Control.
• During 2008-2009, KTH contributes with manpower
  resources corresponding to 1.5 person-years for
  the task WP08-MHD-05-01
• Measurements on EXTRAP T2R of plasma rotation
  braking due to non-resonant external magnetic
  fields
• Experimental comparison of MIMO and SISO systems
  for RWM feedback on EXTRAP T2R

4
Outline

• Key issues for MHD control research
• Resistive Wall Mode (RWM) stabilization
• RWM feedback control studies at EXTRAP T2R
• Plans for MHD control at ASDEX Upgrade
• Summary and plans

5

• Key issues for MHD control research

6
ELM suppression

• Suppression of Edge Localized Modes (ELM) in
  high-confinement plasmas by externally applied
  Resonant Magnetic Perturbations (RMP).
• Suppression through stochastization of edge
  magnetic field?
• Understand transport in stochastic magnetic
  fields, effect on H-mode pedestal.
• Understand how to avoid magnetic braking of the
  plasma rotation by the RMP. (This is an unwanted
  side-effect, which may trigger other MHD activity)

7
Control of TM rotation

• Control of tearing mode rotation by external
  fields
• Understand interaction of TM with external
  fields, how modes phase-lock to externals fields.
• Force mode to rotate in order to avoid a pending
  locked-mode disruption, or delay it sufficiently
  for the control system to take mitigating
  actions.
• Force Neoclassical Tearing Mode (NTM) to a
  specified phase in order to allow stabilization
  by localized ECCD injection in the island O-point
  region. (Stabilization through replacement of the
  lacking bootstrap current in the island.)

8
Resistive wall modes

• Resitive Wall Mode (RWM) stability is important
  for the advanced tokamak concept (high-b,
  steady-state operation).
• The ideal MHD mode instability is limiting b for
  advanced tokamak equilibria that avoid
  neoclassical tearing mode (NTM) instability
  (Equilibria with negative central shear and high
  minimum q-value.)
• Stabilization of some external ideal MHD modes by
  conducting structures near the plasma is possible
  however it leads to the Resistive Wall Mode
  (RWM) instability.
• A current key issue is the kinetic effects on the
  RWM
• Understanding of kinetic stabilization physics.
• Experimental observations of new branches of RWM
  Energetic-particle-driven Wall Mode (EWM)
  (JT-60U), fishbone-driven RWM (DIII-D).

9

• RWM stabilization

10
RWM stabilization by plasma rotation

• Stabilization of RWM by sufficiently rapid plasma
  rotation is theoretically predicted and it has
  been seen in experiments at DIII-D, NSTX and
  JT-60U
• Recent high-b experiments on DIII-D (2007)
  indicate that the RWM remains stable also in low
  rotation plasmas.
• Stabilization by kinetic effects has been
  proposed to explain low-rotation experiments.
• Key issue Understanding physics of kinetic
  damping of the RWM.

11
RWM stabilization by feedback

• Active magnetic feedback control of RWM is
  probably required for reliable operation at very
  high b.
• In the US, feedback control of RWM has been
  demonstrated in experiments on several tokamaks
  HBT-EP, DIII-D and NSTX.
• In Europe, RWM feedback control experiments are
  currently carried out on the two RFP devices
  RFX-Mod and EXTRAP T2R. There is a plan to
  provide ASDEX-Upgrade with capabilities for RWM
  control.
• Various RWM stability codes are in use (VALEN,
  STARWALL, CARMA) to predict the requirements for
  RWM stabilization in ITER.
• Use of internal coil systems in ITER for RWM
  stabilization is foreseen. Calculations predict a
  50 increase of bN with feedback.

12
Priority issues for RWM control in ITER

• Priority issues for RWM feedback stabilization in
  ITER includes
• RWM feedback studies with realistic 3D wall
  geometry,
• multimodal feedback,
• specification of noise,
• effects of blankets,
• assessment of power supply requirements.

13

• RWM control studies at EXTRAP T2R

14
RWM control studies

• Our philosophy is to develop understanding and
  strategies for RWM control from the viewpoint of
  process control, gaining access to a number of
  tools already developed in this field.
• Scope is wide, including
• modeling,
• mode identification,
• controller design,
• real-world implementation,
• conduction of experiments.

15
RWM in RFP

• In RFP, the RWM instability is current-driven,
  and is therefore observed at all values of b.
• The unstable RWM spectrum consists of a range of
  m1 kink modes with different toroidal mode
  number n. RWM feedback in the RFP requires
  multi-mode control.
• Issues related to multi-mode control are
  side-band harmonics generation by the active coil
  array, and aliasing of higher harmonics in the
  sensor array.
• The main RWMs are non-resonant and as a
  consequence rotational stabilization is
  ineffective.

16
RWM growth rate spectrum
A range of m1 modes are unstable
17
Control coil array on EXTRAP T2R
18
RWM control strategies

• Intelligent shell
• External control coils minimize the total radial
  magnetic field everywhere at the resistive wall
  (measured by sensors coincident with the coils)
  allowing the resistive wall to be seen by the
  plasma as an ideally conducting wall.
• Conventional control problem, single-input-single-
  output (SISO) system.
• Controller can be implemented with traditional
  analog electronic circuits.
• Mode control scheme specifies an arbitrary
  action on the RWM (tracking a reference input)
• The controller algorithm combines data from an
  array of sensors in order to determine the output
  to an array of actuators.
• Complex multivariable control problem,
  multiple-input-multiple-output (MIMO) systems.
• Require the use of a modern digital, computer
  based controller with control algorithms
  implemented in software.

19
Mode control tracking reference signal

• Advanced mode control strategy for general RWM
  control with a non-zero reference input.
• Controller performs output tracking Reference
  for a specified mode is the radial magnetic field
  at the wall (a generalized intelligent shell
  allowing non-zero radial field).
• Implementation uses 64 individually tuned PID
  controllers - compensates for wall asymmetry
  (gaps), and amplifier individual characteristics
• In press E. Olofsson and P. Brunsell,
  Controlled magnetohydrodynamic mode sustainment
  in the reversed field pinch Theory, design and
  experiments, Fusion. Eng. Des. (2008),
  doi10.1016/j.fusengdes.2008.11.052

20
m1, n-12 sustained at different amplitudes
21
Spectrum sweep from n-15 to n15
Radial field
Coil current
22
Mode identification experiments

• Mode identification experiments is also known as
  MHD spectroscopy.
• Involves active probing of the plasma by applying
  external fields using the control coils.
• The present experiments on EXTRAP T2R use a
  closed-loop identification method
• The response to external perturbations of
  unstable RWM is measured while simultaneously
  maintaining stabilizing feedback.
• A psuedo-random dithering signal is applied to
  all coils.
• The advantage of the closed-loop operation is
  that the disturbance to the plasma is minimal.
• Accepted to IEEE MSC-09 E. Olofsson, P.
  Brunsell, J. Drake, Closed-loop parametric
  identification of magnetohydrodynamic normal
  modes spectra in EXTRAP T2R reversed-field pinch

23
Applied dithering signal
Radial fields (top) and coil currents (bottom)
24
Resulting identification of model parameters
Wall penetration time constant
RWM growth rate
Current-to-field constant
25

• Plans for MHD control at ASDEX Upgrade

26
Plans for MHD control at AUG (1)

• Enhancement of AUG for active MHD control
  experiments are in progress
• involves installation of 24 in-vessel coils and a
  conducting wall
• goals are ELM suppression, NTM control, RWM
  feedback stabilization
• Phase I application for preferential support was
  approved in 2007
• Work programme (and Phase II application) is
  divided in five stages
• 16 in-vessel saddle coils (Phase II granted in
  2008, work is underway)
• Additional 8 in-vessel coils (2009)
• 12 AC power supplies (2009)
• Conducting wall, enabling RWM studies (2010)
• Additional 12 AC power supplies (2011)
• Total investment cost is 6.9 MEuro.

27
ASDEX Upgrade
In-vessel coils
Conducting wall
28
Plans for MHD control at AUG (2)

• The project is a collaboration between 4
  Associations
• Max-Planck-Institut für Plasmaphysik (Assoziation
  IPP-EURATOM)
• Forschungszentrum Jülich (Assoziation
  FZJ-EUROATOM)
• Consorzio RFX (Associazione EURATOM-ENEA)
• KTH (Association EURATOM-VR)
• KTH participates in the development of the RWM
  controller
• Conceptual design,
• Hardware specification,
• Development of embedded software

29
Summary and plans

• Key goals for MHD control research are
  development and understanding of ELM suppression,
  NTM control and RWM stabilization. The planned
  enhancement of ASDEX Upgrade will address these
  issues in an ITER relevant geometry.
• Our basic philosophy is to develop understanding
  and strategies for RWM control from the viewpoint
  of process control, thereby gaining access to a
  number of tools already developed in this field.
• The capabilities of EXTRAP T2R and RFX are very
  useful for development of practical
  implementations and for experimental testing of
  new ideas.
• The plan is to devise general methods for RWM
  control that are useful for both RFP and
  tokamaks, and to contribute to the specific
  design of the RWM controller for ASDEX Upgrade.