1. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 1 Physics and control of Resistive Wall Mode Valentin Igochine Max-Planck-Institut für Plasmaphysik, Euratom-Association, Garching, Germany

2. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 2 Introduction Motivation Simple dispersion relation Physics of the RWM Electromagnetic part Kinetic part Control of RWM Control methods Triggering of RWMs Outline

3. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 3 This scenario has to have higher β N compared to conventional scenario to be attractive. Advanced Tokamak Scenario [C. M. Greenfield et. al. PoP 2004] Stationary operations are possible

4. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 4 Resistive wall mode is an external kink mode which interacts with the resistive wall. The mode will be stable in case of an perfectly conducting wall. Finite resistivity of the wall leads to mode growth. Resistive Wall Mode (RWM) [M.Okabayashi, NF, 2009] [T. Luce, PoP, 2011] RWM has global structure. This is important for “RWM ↔ plasma” interaction. DIII-D [I.T.Chapman, PPCF, 2009] JET

5. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 5 Simplest RWM dispersion relation rotation frequency of the plasma perturbed energy in vacuum growth rate

6. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 6 rotation frequency of the plasma perturbed energy in vacuum growth rate Simplest RWM dispersion relation

7. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 7 rotation frequency of the plasma perturbed energy in vacuum growth rate Simplest RWM dispersion relation

8. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 8 Feedback stabilizes from no-wall limit rotation frequency of the plasma perturbed energy in vacuum growth rate Simplest RWM dispersion relation

9. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 9 Feedback stabilize from no-wall limit rotation frequency of the plasma perturbed energy in vacuum growth rate Simplest RWM dispersion relation

10. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 10 Feedback stabilizes from no-wall limit Rotation stabilizes from ideal-wall limit rotation frequency of the plasma perturbed energy in vacuum growth rate Simplest RWM dispersion relation

11. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 11 Resistive Wall Mode (RWM) Stabilization is really important only in the advanced scenario. RWM stabilization allows to gain this region. Conventional scenario (moderate improvements) Advanced scenario (crucial improvements)

12. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 12 Physics of RWMs

13. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 13 RWM physics in tokamak RWM physics in tokamaks RWM interaction with externally produced magnetic fields resistive wall error fields control coils RWM interaction with plasma plasma rotation fast particles thermal particles +

14. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 14 RWM physics in tokamak RWM physics in tokamaks RWM interaction with externally produced magnetic fields resistive wall error fields control coils Mode can be represented as a surface currents Physics: electromagnetism RWM interaction with plasma plasma rotation fast particles Wave-particle interaction Physics: kinetic description of the plasma-wave interaction +

15. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 15 RWM physics in tokamak RWM physics in tokamaks RWM interaction with externally produced magnetic fields resistive wall error fields control coils Mode can be represented as a surface currents Physics: electromagnetism RWM interaction with plasma plasma rotation fast particles thermal particles Wave-particle interaction Physics: kinetic description of the plasma-wave interaction +

16. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 16 Interaction of RWM with external perturbations Resistive wall (currents in the wall) Inertial layer Main plasma Current layer (RWM) Simple models [R. Fitzpatrick, PoP, 2002; V.D.Pustovitov, Plasma Physics Reports, 2003; A.H.Boozer, PRL, 2001] Real vessel [F.Vilone, NF, 2010; E.Strumberger, PoP, 2008] currents in the wall

17. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 17 RWM physics in tokamak RWM physics in tokamaks RWM interaction with externally produced magnetic fields resistive wall error fields control coils Mode can be represented as a surface currents Physics: electromagnetism RWM interaction with plasma plasma rotation fast particles Wave-particle interaction Physics: kinetic description of the plasma-wave interaction + Reversed Field Pinch: small plasma rotation no fast particles

18. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 18 What one can achieve in RFPs with external coils ? [T. Bolzonella, V. Igochine, et.al., PRL, 2008; V. Igochine, T. Bolzonella, et.al., PPCF, 2009] [P. R. Brunsell et.al., PRL, 2004] Feedback Stabilization of Multiple Resistive Wall Modes Decoupling and active rotation of a particular RWM EXTRAP T2R RFX

19. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 19 RWM physics in tokamak RWM physics in tokamaks RWM interaction with externally produced magnetic fields resistive wall error fields control coils Mode can be represented as a surface currents Physics: electromagnetism RWM interaction with plasma plasma rotation fast particles thermal particles Wave-particle interaction Physics: kinetic description of the plasma-wave interaction +

20. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 20 RWM interaction with plasma [M. Okabayashi et. al. PPCF 2002] RWM is unstable if rotation drops below critical value Plasma rotation try to decouple RWM from the wall and plays stabilizing role How strong the rotation stabilization? What is the critical rotation which is necessary to stabilize RWM? Is only the plasma rotation important? The answers depend on the type of interaction between RWM and plasma rotation which is considered by the model and/or accuracy of the model for such interaction

21. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 21 Implementation of the rotation effects in linear MHD Linearized MHD equations with influence of rotation is the (non-uniform) rotation frequency of the plasma at equilibrium is the toroidal mode number

22. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 22 Implementation of the rotation effects in linear MHD is the (non-uniform) rotation frequency of the plasma at equilibrium is the toroidal mode number momentum equation Linearized MHD equations with influence of rotation

23. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 23 Implementation of the rotation effects in linear MHD Dissipation, viscous stress tensor One has to make approximations for kinetic description of the dissipation Linearized MHD equations with influence of rotation momentum equation is the (non-uniform) rotation frequency of the plasma at equilibrium is the toroidal mode number

24. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 24 Simplest variant of the viscous stress tensor Fluid approximation of the Landau damping: ‘sound wave damping’ [Hammet G W and Perkins F W 1990 Phys. Rev. Lett]

25. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 25 Simplest variant of the viscous stress tensor Fluid approximation of the Landau damping: free parameter between 0.1 and 1.5 ‘sound wave damping’ [Hammet G W and Perkins F W 1990 Phys. Rev. Lett]

26. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 26 Simplest variant of the viscous stress tensor Fluid approximation of the Landau damping: ‘sound wave damping’ [Hammet G W and Perkins F W 1990 Phys. Rev. Lett] free parameter between 0.1 and 1.5 ‘kinetic damping’model uses kinetic energy principle with [Bondeson A and Chu M S 1996 Phys. Plasmas] No free parameters magnetic drift frequency diamagnetic frequency

27. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 27 Comparison of the critical rotation with experiments [LaHaye, NF, 2004] no-wallideal-wall sound wave damping overestimate „kinetic damping“ model underestimate

28. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 28 Comparison of the critical rotation with experiments [LaHaye, NF, 2004] no-wallideal-wall sound wave damping overestimate „kinetic damping“ model underestimate …but rotation is sufficiently strong in these experiments due to NBI

29. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 29 Experiments with balanced NBI input in DIII-D [Reimerdes, PPCF, 2007] [Strait, PoP, 2007]

30. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 30 Experiments with balanced NBI input in DIII-D [Reimerdes, PPCF, 2007] [Strait, PoP, 2007] No stabilization is possible here

31. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 31 Comparison with almost zero rotation case in NSTX [S.A. Sabbagh, IAEA, 2010, PD/P6-2] Stabilization is achieved for almost zero rotation (not possible in ”kinetic model”) Thus, an important part is missing in the model! NSTX

32. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 32 Possible variants of modeling Self-consistent modeling (MARS,…) Linear MHD + approximation for damping term (+) rotation influence on the mode eigenfunction (-) damping model is an approximation Perturbative approach (Hagis,…) Fixed linear MHD eigenfunctions as an input for a kinetic code (-) rotation does not influence on the mode eigenfunctions (+) damping is correctly described in kinetic code

33. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 33 Possible variants of modeling Self-consistent modeling (MARS,…) Linear MHD + approximation for damping term (+) rotation influence on the mode eigenfunction (-) damping model is an approximation Perturbative approach (Hagis,…) Fixed linear MHD eigenfunctions as an input for a kinetic code (-) rotation does not influence on the mode eigenfunctions (+) damping is correctly described in kinetic code

34. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 34 Perturbative approach. Influence of the resonances δWKδWK Average  d Average  b [I.T.Chapman, PPCF, 2009] Transit Frequency Bounce Frequency Precession Drift Frequency stable unstable Change in mode energy has a term which contains several different resonances (denominator of equation tends to zero, get a large contribution to δW K ) JET

35. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 35 Perturbative approach. Influence of the resonances δWKδWK Average  d Average  b [I.T.Chapman, PPCF, 2009] Transit Frequency Bounce Frequency Precession Drift Frequency stable unstable Different resonances are important! & Low frequencies are important! Change in mode energy has a term which contains several different resonances (denominator of equation tends to zero, get a large contribution to δW K ) JET

36. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 36 New model for rotation influence on the MHD modes (MARS-K) full toroidal geometry in which the kinetic integrals are evaluated Kinetic effects are inside the pressure [Liu, PoP, 2008, Liu, IAEA, 2010]

37. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 37 full toroidal geometry in which the kinetic integrals are evaluated …but still some strong assumptions are made: neglects the perturbed electrostatic potential, zero banana width for trapped particles, no FLR corrections to the particle orbits. There is no guaranty that all important effects are inside. Kinetic effects are inside the pressure [Liu, PoP, 2008, Liu, IAEA, 2010] New model for rotation influence on the MHD modes (MARS-K)

38. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 38 Application of both models for ITER RWM is stable at low plasma rotation up to without feedback due to mode resonance with the precession drifts of trapped particles. [Liu, NF,2009, IAEA, 2010] perturbative self-consistent ideal wall ideal wall no wall no wall rotation Stable at low rotation black dots are stable RWM

39. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 39 Application of both models for ITER RWM is stable at low plasma rotation up to without feedback due to mode resonance with the precession drifts of trapped particles. … but some important factors are missing (for example alpha particles are not taken into account). [Liu, NF,2009, IAEA, 2010] perturbative self-consistent ideal wall ideal wall no wall no wall rotation Stable at low rotation black dots are stable RWM

40. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 40 Influence of α-particles on RWM in ITER (perturbative) [Liu, NF, 2010] Thermal particles only Thermal particles and α-particles Stabilization at higher rotation from α-particles ideal wall no-wall rotation ideal wall no-wall

41. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 41 Control of RWMs

42. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 42 Control by external magnetic coils DIII-D: two coil sets 6 C-coils, 12 I-coils usually connected in quartets. Controller Power Amplifiers Information about the mode Control algorithm Decided amplitudes for external currents Control currents

43. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 43 Feedback logic (inside the controller)

44. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 44  Looking more closely at feedback control system components  Sensors and mode identification  Control logic: intelligent shell-like or including plasma response? …  Power supply  Coil geometry, number and position Basic control logic There are multiple possibilities in each of the points and one has to find optimum solution.

45. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 45 How important low n control close to the pressure limit? NSTX

46. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 46 Why the RFPs are important for RWM study? RFX-mod “quantum” for active control:  =7.5 ° ;  =90°. Area covered=0.52% RFX-mod control system is routinely working on discharges with 30-40 MW Ohmic input power, where effective control is essential. RFX-mod control system is made by 192 active saddle coils, each independently fed. 100% coverage of the plasma surface.

47. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 47 Active system reconfiguration NB: it is a software reconfiguration, active on selected harmonics only! This will allow to understand how much coils we need for RWM stabilization. The main idea is to test on the same plasma and on the same device different active control configurations.

48. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 48 Resistive Wall Mode (RWM) [T. Luce, PoP, 2011] Installation of the high field side coils is much more easy then in the other places. And these coils are much more effective due to the ballooning mode structure! Number of the coils is restricted by: -Tokamak design - Required currents for active control Restricted number of coils could lead to the sideband and excitation of other modes (multiple modes control is necessary)

49. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 49 JT-60SA RWM stability simulations RWM growth rates versus achievable-  N calculated by VALEN. No-wall, ideal wall  N -limits and feedback with different proportional gains cases are shown. High gain is required for higher

50. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 50 E. Olofsson et al, RFX-mod programme workshop, 2011 E. Olofsson et al, “Closed loop direct parametric identification of magnetohydrodynamic normal modes spectra in EXTRAP T2R reversed-field pinch,” Proceedings of the 3rd IEEE Multi-conference on Systems and Control (MSC) July 2009 From intelligent shell (SISO) to real time DFT and to dithering. Close loop system identification (reversed field pinch) (plasma)

51. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 51 E. Olofsson et al, Plasma Physics and Controlled Fusion (53), (084003) Cylindrical stability vs. Dithering in reversed field pinch Modeled picture Experimental picture

52. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 52 Difference between current driven and pressure driven RWMs Interaction of RWM with plasma is different for current driven and pressure driven RWMs. One has to investigate pressure driven cases. RFPs expertise is not applicable here. Current driven kink-peeling Current driven kink Pressure driven kink

53. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 53 Triggers for RWM at low plasma rotation [M.Okabayashi et.al, NF, 2009][G. Matsunaga et.al., PRL, 2009 Okabayashi et.al., PoP, 2011] ELMs It is important that RWM could be triggered by core (off axis fishbones) and edge (ELMs) modes. This also shows global structure of RWM. Integrated control of different MHD modes is required to stabilize RWM. Energetic Particle Driven Wall Mode (EWM)

54. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 54 Review papers

55. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 55 Conclusions RWMs may limit our operations close to the pressure limit. RWM stability is affected by interactions with resistive wall, external fields, fast particles and plasma rotation. Stabilization at low rotations (comparable and less to what is expected for ITER) is achieved, but self consistent modeling is still a problem. Recent modeling shows that RWMs could be stable in ITER without external feedback well above the no wall limit (kinetic effects at low plasma rotation) Other modes are able to trigger RWMs at low plasma rotation. For RWM control we need further optimization of: sensors, actuators, control stratages and identification tools.

56. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 56

57. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 57 Open issues and discussion  Present and coming devices: - sensor optimization: field component, number, position, sampling frequency - actuator optimization: mode rigidity - controller optimization: test different control strategies, noise effect removal - develop advanced identification tools (dithering,...) - prove stabilization of multiple unstable RWMs (long pulses needed)  Future devices: ITER - compatibility of RWM active control with other control needs (e.g. ELM mitigation) still to be fully assessed  Future devices: DEMO class - Are active coils compatible with a DEMO/reactor environment? - Can we imagine other RWM control strategies for burning plasmas

58. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 58 Resistive Wall Mode (RWM) [DIII-D review 2002] Mode is always unstable Mode is always stable Mode stabilization is necessary Stabilization gives approximately factor 1.5-2 in β N, which is about factor 3-4 for fusion power.

59. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 59 Early suggestions for active control C.G. Gimblett, PPCF, 31 2183 (1989)  Intelligent shell A network of active coils which cover the surface of the resistive wall compensates for the dissipated flux on the resistive wall according to the measurements of an associated network of sensors.  Rotating secondary wall A rotating secondary wall can stabilize a thin shell mode (today a RWM). The rotation rate required is determined largely by the inverse time constant of the least conducting wall. C.M. Bishop, PPCF, 31 1179 (1989)

60. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 60  Two essential components in feedback (modeling point of view)  Plasma dynamics (Plant, P)  Controller (K) Liu, Phys. Plasmas, 7 3684 (2000) Basic control logic

61. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 61 Different configurations explored 48 (  ) x 4 (  ): 100% 48 (  ) x 1 (  ): 25% 16 (  ) x 1 (  ): 8.3% 3 (  ) x 4 (  ): 6.25% 8 (  ) x 1 (  ): 4.2% 12 (  ) x 1 (  ): 6.25%

62. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 62 Present facilities for RWM control studies DIII-D: two coil sets 6 C-coils, 12 I-coils usually connected in quartets. See also the MAST system (RFA studies). New HBT-EP (Columbia University, New York) setup: 6x20 control coils for modularity tests, 20 adjustable wall segments.

63. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 63 Facilities for RWM control studies JET Error Field Correction Coils (RFA studies). See also MAST system. New AUG Active in-vessel coils to be coupled with a new conducting wall for MHD stabilization experiments

64. Hefei, China/ August 2012 / Lecture 5 Valentin Igochine 64 Resonant FieldApproximation (RFA) A.M. Garofalo et al., Phys. Of plasmas 9 1997 (2002)H. Reimerdes et al., PRL 93 135002 (2006)