[PPT] - Optimization of the magnetic field Optimization of the magnetic PowerPoint Presentation

SLIDE 1

F.Maviglia 1 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Optimization of the magnetic field Optimization of the magnetic field configuration for JET breakdown configuration for JET breakdown

F. Maviglia
F. Maviglia

with contribution from

R. Albanese, P.J.
R. Albanese, P.J. Lomas

Lomas, A. , A. Manzanares Manzanares, M. , M. Mattei Mattei, A. , A. Neto Neto, F.G. , F.G. Rimini Rimini, P.C. de , P.C. de Vries Vries and JET EFDA Contributors and JET EFDA Contributors

SLIDE 2

F.Maviglia 2 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Introduction
Modelling activity
Static and dynamic simulations
Intensified Visible Fast Camera KL8A
Experimental results
Conclusions

Outline Outline

SLIDE 3

F.Maviglia 3 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

filling pressure;
toroidal electric field (E0

≈1V/m in present Tokamaks, E0 ≈0.33V/m ITER);

magnetic field configuration (null

position and extension of the low field region);

Breakdown optimization parameters:

1 2 3 4 5 6 −4 −3 −2 −1 1 2 3 4

∆Ψ = 0.01 Vs

R (m) Z (m) IPRIM −15 (kA) IP4T 180 (A) IERFA 0(A) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

P3ML P3MU P1 P2RU P2RL P3RU P3RL P4U P4L

Static simulation IPRIM transformer (open loop):

mode D,C initial value(=premag)[-7,-40] kA, iron fully saturated;
mode B premag ≈

0, iron with residual magnetization; IP4 Vertical field (open loop) IERFA radial field (feedback)

Circuits used for breakdown at JET:

Introduction Introduction

Hexapolar field

SLIDE 4

F.Maviglia 4 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Introduction Introduction

; 4 1   

z null c

B B a L 



 

z null

B B a



= size of the field min. P = pressure; E = electric field; = avg. stray field over the null;

; 10 3 . 1 exp 10 2

4 3

         



E P P

i



Avalanche successful if Lc /λi >> 1 minimize <δBz >, maximize anull

Connection length Lc

[1]:

Ionization length

λi :

= toroidal field;

1 2 3 4 5 −3 −2 −1 1 2 3 R [m] Z [m]

∆Ψ = 0.01 Vs iso B= [0.5;1;1.5;2;2.5] mT

IP1 −15 [kA] IP4 180 [A]

with: with:

Magnetic field configuration

[1] B. LLoyd, et al., Nucl. Fus. 31 (1991) 2031.

SLIDE 5

F.Maviglia 5 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

JET poloidal field (PF) coils

JET iron

8 Coil sets, named P1 to P4,D1 to D4 (D not shown).

P4 P3 Limbs P1 P2 Central column Collar

(Shoes not shown)

Introduction Introduction

SLIDE 6

F.Maviglia 6 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

CREATE Model [1] based on 2D FEM:
Passive structures: MS, VV, RR, MK2. Resistivity value
f MS has been refined via best fit of the simulations with

the experimental data. Resistivity of VV, RR, MK2 in agreement with [2-3].

References:

[1] R. Albanese et al., Nucl. Fusion, 38, 1998, pp. 723–738. [2] R. Albanese, et.al. Nucl. Fusion 44 (2004) 999–1007. [3] S.Gerasimov, ‘ JET_PassiveSimpleModel.pdf ‘.

Passive structures

#78021 (Ipre = -15kA)

Dyn. Sim. v.s. exp. estimations of mk2 current

Modelling Modelling

SLIDE 7

F.Maviglia 7 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

2 4 6 −6 −4 −2 2 4 6 CREATE−NL/CREATE−L model R [m] Z [m]

JET Iron Model

Centre limb pad Outer limb pad

3 mm gap present in the magnetic circuit at z=4.5m, and not at z=-4.5m, present at JET and considered in the model: up-down asymmetry.

Outer limb pad:

thickness 3mm

Centre limb pad:

thickness 3mm

CREATE iron model

Modelling Modelling

SLIDE 8

F.Maviglia 8 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

E

Mode D breakdown

Static simulation using as input primary and vertical field current (other noisy

curr. set to 0).
Hexapole null splits in two

quadrupole nulls in the direction

f

the radial field given by the upper iron gaps.

Inboard-low null expected

to be preferred for plasma formation for the higher electrical field.

|Br|≈0.35mT

Critical points for JET breakdown magnetic reconstruction:  Residual iron magnetization, not included in present exp. measurements;  Perturbing effect of vessel and in-vessel passive currents.

Static Static simulations simulations

SLIDE 9

F.Maviglia 9 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Configurations used for breakdown:

Configuration 1 Configuration 2 Improved

2 2.5 3 3.5 −1.5 −1 −0.5 0.5 1 1.5 2 R [m] Z [m] Z [m]

Intensified Intensified Visible Visible Fast Camera KL8A Fast Camera KL8A

Specifications Region

f

interest 272 x 384 px Freq. 1.0 kHz Exposur e Time 142.9 s Filter none Intensifi er Gain 700 v Specifications Region

f

interest 176 x 256 px Freq. 7.5 kHz Exposur e Time 125.0 s Filter none Intensifi er Gain 750 v

SLIDE 10

F.Maviglia 10 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Comparison between dynamic simulation and fast camera [1] (carbon wall):

Inner field null preferred for plasma formation.
Plasma pushed against the inboard and included in the region between the two

lines where Btg = 0.

The ionization cloud appears to be pushed down in the divertor region.
For a successful breakdown the plasma is pushed up toward the outer wall.
0.5

0.5 1.0 1.5 2.0 Pulse No: 78369 t = 0.006028s Pulse No: 78369 CREATE L flux-map 2.5

1.0
1.5
2.0
2.5

2.0 1.5 2.5 3.0 3.5 4.0 4.5 Z (m)

IPRIM = -19.4(kA) IP4T = 123(A) IFRFA = 138(A) Other 0(A) 6ms Force ∆ψ = 0.02Vs

R (m)

JG10.248-4c

1.5 2.0 1.0 0.5

0.5
1.0
1.5

2.5 2.0 3.0 3.5 Z (m) R (m) 1.5 2.0 1.0 0.5

0.5
1.0
1.5

2.5 2.0 3.0 3.5 Z (m) R (m) 1.5 2.0 1.0 0.5

0.5
1.0
1.5

2.5 2.0 3.0 3.5 Z (m) R (m)

JG10.248-5c

Pulse No: 78369 t = 0.014028s Pulse No: 78369 t = 0.028028s Pulse No: 78369 t = 0.038028s

time increasing

[1] F.Mavilgia et al.,Fus. Eng. and Des. 86 (2011) 675–679.

Standard non optimized breakdown dynamic evolution

H L

SLIDE 11

F.Maviglia 11 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Multi-pole field (2D plane geometry):

Hexapole splits in 2 quadrupole nulls in the direction of the perturbation field with vertical and radial components.

Hexapole

null.

Perturbed hexapolar field, represents iron-gaps Symmetric multipolar field: np=6

filaments added represent JET irongaps

Static Static simulations simulations

SLIDE 12

F.Maviglia 12 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Radial field added to rotate perturbed field: Zoom in The two quadrupole nulls are rotated in the direction given by the correction radial field.

* filaments added represent radial field correction

* *

Static Static simulations simulations

SLIDE 13

F.Maviglia 13 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Optimization of field null position during breakdown

Aims: Place the null field position far from the inner divertor region. Solution: Rotate the 2 quadrupole field nulls by applying an offset radial field bias.

2 3 4 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5

0 ms

R [m] Z [m]

IP1 −20.0 [kA] IP4 258 [A] IFRFA −2 [A] Other PF 0 [A]

∆Ψ = 0.01 Vs iso B= [0.5;1;1.5;2;2.5;] mT

Flux 2 3 4 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5

0 ms

R [m]

IP1 −20.0 [kA] IP4 258 [A] IFRFA 118 [A] Other PF 0 [A]

∆Ψ = 0.01 Vs iso B= [0.5;1;1.5;2;2.5;] mT

Flux 2 3 4 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5

0 ms

R [m]

IP1 −20.0 [kA] IP4 258 [A] IFRFA 358 [A] Other PF 0 [A]

∆Ψ = 0.01 Vs iso B= [0.5;1;1.5;2;2.5;] mT

Flux 2 3 4 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5

0 ms

R [m]

IP1 −20.0 [kA] IP4 258 [A] IFRFA 238 [A] Other PF 0 [A]

∆Ψ = 0.01 Vs iso B= [0.5;1;1.5;2;2.5;] mT

Flux

Standard radial bias =0A radial bias =+120A radial bias =+240A radial bias =+360A

In the cases with radial

bias ≠0 the upper inner field null expected to be preferred for plasma formation due to higher electrical field.

Static sim all at 40s:

Increasing radial field bias → quadrupoles rotation

Static Static simulations simulations

SLIDE 14

F.Maviglia 14 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

kl1-o4wb-raw @40.04:

Pulse #80400, std. radial bias =0

kl1-o4wb slow camera (40ms) first visible frame: plasma starts in an upper inner wall position with radial bias correction, far from divertor region.

kl1-o4wb-raw @40.03:

Pulse #80402, radial bias =+240A

Experimental Experimental results results mode D mode D -

20kA

20kA premag premag

SLIDE 15

F.Maviglia 15 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

20.0
19.0
18.0
17.0

103 A

0.2 0.4 0.6 0.8 1.0 1.2

103 A

0.0 1.0 2.0

103 A

4.0
3.0
2.0
1.0

0.0

105V

0.0 0.5 1.0 1.5 2.0 2.5

a.u.

40.00 40.02 40.04 40.06 SEC.

2.0
1.5
1.0
0.5

0.0

105 A 80400 80401 80402 80404

IP1 (primary) IP4 (vertical) IERFA (radial)

VFB (vertical velocity

control request)

HALPHA

Plasma current

0,std 120A 240A* 360A

Higher velocity loop

control for st. pulses

Higher radial

field current peak for standard pulse.

Plasma curr.>47kA,

velocity loop “ON”

Radial bias:

*radial bias =+240A

ptimum predicted

by sim.: exp. verified.

Experimental Experimental results results mode D mode D -

20kA

20kA premag premag

SLIDE 16

F.Maviglia 16 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

0.0 1.0 2.0

103 A

4.0
3.0
2.0
1.0

0.0

105 A

40.00 40.02 40.04 40.06 SEC.

1.0
0.5

0.0 0.5 1.0 1.5

104 A 80400 80401 80402 80404

IERFA (radial)

VFB (vertical velocity

control request)

std pulse

VFB (vertical velocity

control request)

ptimized

Radial bias:

0,std 120A 240A 360A Velocity loop control peaks (≈105V) when

plasma pulled up

from divertor region for std pulses. Moderate control action for optimized pulses. ∆IERFA in time interval of interest decrease form 3.2kA to 0.7kA: factor ≈4. 2.5kA peak std. pulse: amplifier limit =5kA

standard (radial bias=0)

ptimized

(radial bias=+240A)

time interval of interest

Experimental Experimental results results mode D mode D -

20kA

20kA premag premag

V V

SLIDE 17

F.Maviglia 17 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

20.0
19.0
18.0
17.0

103 A

0.2 0.4 0.6 0.8 1.0 1.2

103 A

1.0

0.0 1.0 2.0 3.0 4.0 5.0

103 A

1.5
1.0
0.5

0.0

106 V

2.5
2.0
1.5
1.0
0.5

0.0

105 A

40.00 40.02 40.04 40.06 SEC. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

a.u.

81755 81756 81757 81758

IERFA (radial) IP4 (vertical) IP1 (primary)

VFB (vertical velocity control request) plasma current

HALPHA

Radial bias (A): 0,marginal 0,NSB 0,NSB +360, ok

∆IERFA (≈4kA) for

0 bias pulses, limits (±5kA) upper lim.

Larger vertical

velocity request for zero radial bias pulses

NSB I0Z limit scan for -20kA breakdown

Non

Sustained Breakdown (NSB) or marginal for radial bias=0.

Experimental Experimental results results mode D mode D -

20kA

20kA premag premag

SLIDE 18

F.Maviglia 18 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

−100 100 200 300 400 500 1 2 3 4 5

Mode D: Premag −20kA IERFA @40s (A) (radial feld bias) ∆IERFA max (kA) in range of interest

radial bias =0 (std.)

min

Statistic all mode D breakdown.

ΔIERFA reduction factor ≈4 from std.

Statistic on all successful mode D breakdown with new VS system (named V5) measurements tuned with a precision lower then ±50A. radial bias scan

figure of merit:

min ( min (∆ ∆IERFA) IERFA)

for t  time interval:

|Plasma current| > 47

kA (velocity loop on) up to

40.065s (plasma in outer

limiter).

Experimental Experimental results results mode D mode D -

20kA

20kA premag premag

SLIDE 19

F.Maviglia 19 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

1000 2000 40 40.01 40.02 40.03 40.04 40.05 40.06 −4 −3 −2 −1

x 10

5

a.u.

sec

IERFA VFB (velocity loop control request)

A

1) 40.018s time increase

Fast visible camera for mode D Fast visible camera for mode D -

10kA premag

10kA premag

1

#82401 no radial bias #82400 radial bias = +150A

(radial) Larger bd area

SLIDE 20

F.Maviglia 20 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

1000 2000 40 40.01 40.02 40.03 40.04 40.05 40.06 −4 −3 −2 −1

x 10

5

a.u.

sec

IERFA VFB (velocity loop control request)

A

1) 40.018s time increase deeper in divertor 2) 40.045s

Fast visible camera for mode D Fast visible camera for mode D -

10kA premag

10kA premag

1

#82401 no radial bias #82400 radial bias = +150A

2

(radial) Larger bd area

SLIDE 21

F.Maviglia 21 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

1000 2000 40 40.01 40.02 40.03 40.04 40.05 40.06 −4 −3 −2 −1

x 10

5

a.u.

sec

IERFA VFB (velocity loop control request)

A

3

1) 40.018s time increase Plasma touches upper wall deeper in divertor 2) 40.045s 3) 40.058s

Fast visible camera for mode D Fast visible camera for mode D -

10kA premag

10kA premag

1

#82401 no radial bias #82400 radial bias = +150A

2

(radial) Larger bd area

SLIDE 22

F.Maviglia 22 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

40.445s

1.5 2 2.5 3 3.5 4 4.5 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5 R [m] Z [m] 1.5 2 2.5 3 3.5 4 4.5 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5 R [m] Z [m] 1.5 2 2.5 3 3.5 4 4.5 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5 R [m] Z [m] 1.5 2 2.5 3 3.5 4 4.5 −2.5 −2 −1.5 −1 −0.5 0.5 1 1.5 2 2.5 R [m] Z [m]

P4 (vertical field) bias

0.40 0.42 0.44 0.46 0.48 0.50 0.0 0.2 0.4 0.6 0.8 1.0

103 A

200 300 400 500 600 0.0 0.5 1.0 1.5

a.u.

40.40 40.45 40.50 40.55 SEC.

1.5
1.0
0.5

0.0

105 A 81627 low IP4 bias 81628 high IP4 bias

VLOOP (primary) IP4 (vertical) IERFA (radial) HALPHA IPLA

A V/m

Different IP4 (vertical field) dynamic induces different eddy currents dynamics: magnetic null enters in the chamber* later -> delayed breakdown.

81627 81628

*

dynamic simulations E @ r = 2.95m E0 ≈0.48V/m

Experimental Experimental results results for for mode B high E mode B high E0

40.475s

*

SLIDE 23

F.Maviglia 23 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

40.512 s

Mode B low Electric field Mode B low Electric field “ “ITER like ITER like” ”

82076 radial bias= -50A 82074 radial bias= 0 82077 radial bias= +50A 82078 radial bias= +100A

82076 82074 82077 82078

Radial bias:

50A

0A +50A +100A

centered

Same initial iron magnetization as mode D

E0 ≈0.3V/m

SLIDE 24

F.Maviglia 24 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

40.512 s 40.566 s

divertor

Mode B low Electric field Mode B low Electric field “ “ITER like ITER like” ”

82076 radial bias= -50A 82074 radial bias= 0 82077 radial bias= +50A 82078 radial bias= +100A

82076 82074 82077 82078

Radial bias:

50A

0A +50A +100A

centered E0 ≈0.3V/m

SLIDE 25

F.Maviglia 25 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

40.512 s 40.566 s 40.574s

failed E0 ≈0.3V/m divertor

E0 <0.3V/m not achievable without error

field optimization (i.e. radial bias).

Min. electric field E0

≈0.25V/m (pulse #82081),with radial bias = +100A.

Mode B low Electric field Mode B low Electric field “ “ITER like ITER like” ”

82076 radial bias= -50A 82074 radial bias= 0 82077 radial bias= +50A 82078 radial bias= +100A centered

SLIDE 26

F.Maviglia 26 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Up-down asymmetric iron gap modelling activity. Radial field ≈0.5mT at r=2.95m, for fully saturated iron. This leads to split the ideal hexapolar null in two quadrupolar nulls. Static and dynamic optimization to optimize the plasma formation region in the upper inner wall, far from divertor zone. Accurate model, with a precision of a fraction of mT, has been employed for magnetic null position scan. Fast visible camera used to validate optimized plasma starting position and dynamic evolution.

Conclusions Conclusions

SLIDE 27

F.Maviglia 27 (27) 7th Workshop on Fusion Data Processing Validation and Analysis. Frascati 26 - 28 March 2012

Optimized breakdown avoids the plasma to be pushed in the divertor region as for the past pulses. Non optimized breakdowns are more fragile with a number of NSB. VS system improved behavior, with smaller radial field current excursion (factor ~4), farther from amplifier limits. Re-established at JET low electric field “ITER like” breakdown using I0Z bias:  mode D min E0 ≈0.27V/m with premag = -7kA;  mode B min E0 ≈0.25V/m. Low electric field pulses breakdown success particularly sensitive to error field optimization: ITER E0 = 0,33V/m.