Prsentation PowerPoint

1
Modelling of ELMs control by RMPs for ITER
M. Bécoulet1, E. Nardon1,G. Huysmans1, W.
Zwingmann1, R.Moyer2, T. Evans3, V. Chuyanov4,,
Y. Gribov4 , A. Polevoi4, G. Vayakis4 , G.
Federici5 G. Saibene5, A. Portone5, A. Loarte5,
V. Parail6, J. Hastie6, C. Gimblett6.   1Associati
on Euratom-CEA, CEA Cadarache, F-13108, St.
Paul-lez-Durance, France 2University of
California, San Diego, La Jolla, CA 92093, USA
3General Atomics, P.O. Box 85608, San Diego, CA
92186-5688, USA 4ITER JWS Cadarache, F-13108,
St. Paul-lez-Durance, France 5 EFDA close
Support Unit,2 Boltzmannstrasse,D-85748
Garching,Germany 6Euratom/UKAEA Fusion
Association, Culham Science Centre,
Abingdon,OX143DB, U.K.

• Outline
• Design work of Resonant Magnetic Perturbation
  (RMP) coils for ITER
• -36 external coils (200-300kAt)
• -in 9 mid-ports (150kAt)
• -blanket in-vessel coils(20kAt)
• 2. Influence of the equilibrium on RMP spectrum
  scan li, magnetic shear, beta poloidal.
• 3. Screening of central perturbations by toroidal
  rotation.

2
Modelling tools

• Biot-Savart in cylindrical coordinates gtBRn,BZn.
  Toroidal (n) spectrum.
• Normalised magnetic perturbation perpendicular to
  the equilibrium flux surfaces in flux
  coordinates. Poloidal (m) spectrum.
• 3) Islands size, Chirikov parameter, Poincare
  plots.

Spectrum of the radial magnetic perturbation
from external coils for ITER,JET, DIIID

Non-linear 3D MHD code JOREK geometry with
X-pointSOL, convective perpendicular EXB
transport, reconnections, toroidal rotation
effects

• Plasma response on external RMPs
• 1) Convective density transport.
• Screening by plasma rotation.
• 3) Amplification of RMPs.

Heat transport,3D code TELM
1)Heat transport with ELMs and RMPs
3
Projection on equilibrium mesh and other formulas.
1) Biot-Savart for vacuum fieldsgtFourier for
cylindrical BR,BZ
2) Flux coordinates component perpendicular to
the flux surface(normalized)
RM,BM on the magnetic axis
3) Physical perpendicular magnetic component
normalised to the equilibrium field gt
4) Cylindrical approximation for ,
poloidal spectrum (in flux coordinates!) island
width and Chirikov
½ island width
Chirikov
4
Reference case of ELM suppression DIIID 125923
q953.51.55MA/1.9T ltdgt0.5,k1.78, Icoil4kA
(even)
5
DIIID reference case
Normalised to the equilibrium radial
(perpendicular to flux surfaces) magnetic field.
Poloidal spectrum and position of rational
surfaces (diamonds)
6
ELM suppression islands overlapping for y
1/2gt0.95
Cylindrical approximation for islands width
Chirikov parameter 1 on y1/20.95
7
On JET? Large pol.spectrum. Larger islands for
n1.
IEFCC, max 2.3kA 16 turns n1 or 2
8
EFCC on JET(n1) larger window for higher q95
Locked mode
ELM mitigation starts
Y. Liang et al
time(s)
9
Seems that effect on ELMsgt when edge islands
overlap(and not their size!)
Islands at threshold current

• I EFCC threshold when islands overlap (and not
  islands size!).
• I EFCC threshold lower for higer q95 larger
  magnetic shear, favourable for overlapping.
• Shots with low q95 more sensitive to locked modes
  because q2 closer to the edge
• (? less screening by rotation)

Shot 67959 q954.8 IEFCC0.75kA
Shot 67954 q954.0 IEFCC1kA
q
Vacuum islands
Screened islands
Shot 68211 q953.5 IEFCC1.2kA
Shot 68212 q953.0 IEFCC1.4kA
q
Y 1/2
Y 1/2
10
RMP coils designs tested for ITER. Toroidal n3.
11
Best designs n3 36 external coils 9 mid-ports
blanket
Optimum physics,low current,but in-vessel !
Larger current,but better accessibility!
36 external coils 200-300kA
Blanket coils26(55) 20kA/11turns
9 mid-ports coils 150kA/11turns
12
Rejected designs (larger currents, bad
spectrum, accessibility)
External around 18 ports and below (2m tor.width)
18 ports (heating systems6ports )
External66
On 18 TF coils (0.8m tor.width)
6 large on PF
9 behind vv, (rad. width up0.3m down0.7m, tor.
3.6m)
13
ITER reference scenarios (ASTRA,HELENA).
14
36 (18 up,18 down) external coils
15
Radial perturbation and poloidal spectrum.
Poloidal spectrum and position of rational
surfaces H-mode,Hybrid,ITB
Normalised to equilibrium (li0.8,bp0.6) field
radial magnetic field.
16
Field lines integration for 36external RMP coils.
7cm
17
Ergodic region y 1/2gt0.95 at 300kA for all
scenarios.
Islands width estimation for H-mode
(li0.8,bp0.6),Hybrid and ITB q-profiles for
higher q95 (m/n) islands are smaller (higher m)
, but magnetic shear is largergt overlapping
18
Toroidal phasing optimisation.
Best for H-mode (q953) and Hybrid (q954) 0
(even parity) , 20(one coil), 40(2 coils). More
resonant is 20
19
300kA (0)gt200kA(20) for H-mode and
Hybrid. (4/3) island7cm
not for ITB (300kA)!
20
Equilibrium fields near coils
21
I
z(m)
y(m)
x(m)
22
18 mid-ports gt12(3NB,IC,LH,EC). 9 port
coilsgtn3.
Geometry 11 turns on 0.5m from R8.7m
23
150kA/11turns all scenarios central island
(4/3) 7cm
24
Blanket coilsclose to plasma (!), but
in-vessel (- - !)
6(55) up 6(55) low possible poloidal and
toroidal phasing optimisation.
25
20kA/11turns all scenarios central island
(4/3) 6cm
26
Limit for Cu cable lt20kAt! ( estimated by
Portone A.)
Only plasma w/o current
Plasma 20kAt current
20kAt!gt300-600
27
Comparison of 3 designs 36ext(300kA),
blanket(20kA/11turns) 9 ports(150kA/11turns)
28
Changes in equilibrium and RMPs spectrum? li
scan.
li scan at bp0.6 for 36ext coils,300kA,n3(0)gt
small changes in pedestal
29
Magnetic shear scan for q953, li 1, bp0.6.
Boundary of ergodic region (Chirikov s1) moves
slightly inwards for low shear and outwards for
high shear, but changes are small.
1/2
30
Beta poloidal scan (0.65-1.2) increased shear gt
islands smaller,but distance is
smallergtChirikovthe same
31
Analytical estimations of screening by rotation
in ITER
Analogy with induction motor. Eddy currents in
the vicinity of qm/n in rotating plasmas can
prevent penetration of the magnetic perturbation
and hence reconnections. Shielding of error
fields on qm/n surfaces. Visco-resistive linear
layer theory (A. Cole and R. Fitzpatrick, Phys.
Plasmas 13, 032503 (2006)).
(J. Hastie, C.Gimblett,V.Parail)
Characteristic times
viscous time
resistive time
Alfven time
visco-resistive layer time
32
Screening by toroidal rotation.
Rotation scan (1)from ASTRA H-mode scenario (only
torque from beamsgtf) and (2) f/2) for q953,
li 0.8, bp0.6
33
Pedestal screening is small slow rotation
larger resistivity
Central 4/3 island is 1.5-2cm with rotation
instead of 7cm (for vacuum fields).
Y
34

• Conclusions
• New 36 external coils (1818) design is proposed.
  
• Toroidal phasing can be changed between upper and
  lower coils. Shift 20 is optimum for H-mode
  (q953)and Hybrid(q954)gt currents needed are
  200-300kA.
• Li scan showed little changes in spectrum and
  Chirikov parameter.
• Edge magnetic shear scan showed small changes in
  the position of ergodic boundary (Chirikov1).
  Lower sheargt inwards. Higher gt outwards.
• Rotation screening small in pedestal,large in
  the centre island in vacuum without
  rotation(4/3) 7cm gt2-1.5cm with rotation.
• bp scan (0.6-1.2) gt negligible changes in RMP
  spectrum.