Edge radiation control in stochastic magnetic field and with RMP application in LHD

M. Kobayashi1,2, S. Masuzaki1,2, S. Morita1,2, H. Tanaka3, B.J. Peterson1,2, Y. Narushima1, K. Mukai1,2, T. Kobayashi2, and the LHD experimental group

1 National Institute for Fusion Science, 322-6 Oroshi-cho, Toki city, Gifu-ken 509-5292, Japan2 Department of Fusion Science, Graduate University for Advanced Studies, Oroshi-cho, Toki-city, Japan

3 Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8601, Japan

2nd Technical Meeting on Divertor Concepts13 to 16 November 2017, Suzhou, China

1

2

Introduction: Helical DEMO design & divertor power load

J. Miyazawa, IAEA DPW4 2016.

Psep~400 MW, Psep/R~25 MW/m

FFHR-d1(LHD based)Rc=15.7 mBc=4.7 TVp=1500 m3

Pfusion= 3 GWHelical divertor

HELIAS 5-B(W7-X based)Rc=22 mBc=6 TVp=1400 m3

Pfusion= 3 GWIsland divertor

0=⋅∇ qHeat flux in SOL for stellarator (LHD) cases

0~//

//

lqq

stq

−⊥

λAt upstream,

RqlaR

Pq SOL ππ

2~,4 //2=⊥with

pstq

SOL

BB

RPq

λπ2// =∴

It may follow the similar scaling as tokamaks Needs detachment operation

(Scaling of λqst?)

Introduction : Control of enhanced radiation at divertor region

Necessity of divertor heat load reduction for future devices:Divertor detachment, radiative divertor is prerequisite to meet the engineering limit of PFC heat load (< several MW/m2).Control of enhanced radiation region in its location & intensity is one challenging issue.

Involved issues in the physics:A/M processes with strong non-linearity in cold plasmaEnergy transport in parallel/perpendicular to field linesImpurity transport, plasma recycling with volume recombination, momentum loss process

3D magnetic field configuration:Seeking divertor optimization in helical devices, RMP application to tokamaks Effects of three dimensionality/symmetry breaking on radiating edge plasma are not yet

fully understood.

This contribution reports experimental results of detachment control with RMP application to the edge stochastic layer of LHD.

3

1. Edge magnetic field structure of LHD2. Time traces of plasma parameters of the controlled detachment discharge3. Radiation enhancement and modification of radiation distribution by RMP

Density dependence of radiationEMC3-EIRENE prediction for radiation modification, AXUV measurementsEUV measurements of CIII – CVI emission profiles

4. Divertor flux distribution with RMPToroidal asymmetryReduction during detached phaseRotation of the asymmetry by RMP phase shift

5. Operation space and simple model analysis6. Core confinement of the RMP assisted detachment7. Summary

Contents of the talk

4

Magnetic field structure of LHD: RMP (m/n=1/1) application remnant island in stochastic region

Helical coils

Plasma shape

RMP coils(m/n=1/1)

0

1

2

3 3.5 4 4.5 5R(m)

With RMP

Without RMP

Resonance value

%1.0~

0≈

Bb coil

r

Rot

atio

nal t

rans

form

ι

2.5 3.0 3.5 4.0 4.5 5.0

-1.0

-0.5

0

0.5

1.0

R (m)

Z (m

) Without RMPWith RMP (m/n=1/1)

Stochastic region

Edge surface layers

Divertor legs

Magnetic island O-point

104103102101100Connection length (m)

R = 3.9 m, a ~ 0.7 m, 10 field periods (toroidal)Divertor : carbon, Fist wall : Stainless steel

5

With RMP Stable sustainment of radiative divertor operation (RMP assisted RD)Without RMP Radiation collapse due to thermal instability

Reduction of divertor power load by a factor of 3 ~ 10

Radiation increase by a factor of ~ 3

Stable operation around density limit

No significant degradation of main plasma confinement

Plasma shrinks at RD phase due to radiative energy loss and RMP penetration

No noticeable high Z impurity (Fe) emission at high density range.

4 102

6 102

8 102103

3 103

5 103

0 200 400 600 800 1000

Inte

nsity

(a.u

.)

Pixel number

Fe V

III 1

67.4

Fe IX

171

.07

Fe X

174

.52

Fe X

177

.24

Fe X

184

.52

Fe X

II 18

6.85

Fe X

I 188

.22

Fe X

XIII

132.

87Fe

XX 1

32.8

5

Fe X

XI 1

28.7

3Fe

XXI

121

.21

Fe X

XII 1

17.1

7Fe

XXI

I 114

.41

Fe X

IX 1

08.3

5Fe

XXI

102

.22

Detach (6.6x1019 m-3)

Attach(ne = 2.0x1019 m-3)

Radiative divertor

024

Div

erto

r pow

erlo

ad (M

W/m

2 )

0

10 )m (10n -319e

With RMPWithout RMP

01020 Radiation

(a.u.)

0

200400

1 2 3 4 5time (s)

Wp (kJ)0.5

0.6 (m)a99

M. Kobayashi et al., Nucl. Fusion 53 (2013) 093032.6

0

1

2

3

0 2 4 6 8 10

Increased volume of low Te region (~10 eV) at remnant island with RMP leads to enhanced carbon radiation

)m (10n -319e

Radiation collapse

Without RMP

With RMP

Rad

iatio

n Po

wer

(a.u

.)

ne dependence of radiation power (bolometer)

100

101

102

103

100

101

Without RMP

With RMP

Resonance layer

Te flattening at island

Te, ne profiles at outboard midplane(Thomson scattering)

10-1

100

101

4.4 4.6 4.8R (m)

(Estimated with ncarbon=0.01neneτ = 1017 m-3 s)

Carbon radiation

DetachRadiation enhanced

M. Kobayashi et al., Nucl. Fusion 53 (2013) 093032.7

Modification of 3D edge radiation structure by RMP : 3D numerical simulation

R

Z

Without RMP

With RMP

Carbon radiation distribution by EMC3-EIRENE

Inboardside

Outboardside

2.0x100

2.0x10-1

2.0x10-2

2.0x10-3

MW/m3

Without RMP

With RMP

Polo

idal

ang

le (d

eg.)

0

100

200

300

Inboard

Outboard

Outboard

Radiation peak at inboard side

Without RMP Radiation peak appears at inboard side. With RMP X-point of m/n=1/1 island is selectively cooled. 8

Polo

idal

ang

le (d

eg.)

Toroidal angle (deg.)0 100 200 300

0

100

200

300

Inboard

Outboard

Outboard

Radiation profile : Comparison between experiments & simulation

With RMP

Split to two peaks

Carbon radiation distribution by EMC3-EIRENE

Without RMP

Inboard side

LOS

of

mea

sure

men

ts

Ch1

Ch16 Experiments

Simulation

Intensity (mW/m2)

Cha

nnel

Intensity (a.u.)

Cha

nnel

The qualitative change of profiles due to RMP application roughly agrees between experiments & simulation.

Results implies selective cooling at X-point of m/n=1/1 island in experiments.Imaging bolometer shows similar effect.

The well-structured magnetic field “catches” the radiation and prevents it from penetrating inward?!

9

Poloidal asymmetry of radiation is enhanced.

Vertical profiles of impurity emission with EUV spectrometers

O-point

X-point3

R (m)4 5

Z (m

)

0.5

0

-0.5

0

0.51

CIII

024C

IV

0

0.5

CV

05C

VI

-0.6

-0.4

-0.2

0

0.6

0.4

0.2

Z (m

)

Intensity (1015 phs./cm2/s)

CIII (386.2 Å) CIV (384.2 Å) CV (40.3 Å) CVI (33.7 Å)

With RMP

W/O RMP

H. Zhang et al., POP 24 (2017) 022510.

CIII, CIV, CV are enhanced by RMP application Profiles of CIII & CIV affected significantly by RMP Up-down asymmetric

CVI is slightly reduced by RMP good indication of impurity screening (?!)

10

Polo

idal

ang

le (d

eg.)

Toroidal angle (deg.)0 100 200 300

0

100

200

300

Inboard

Outboard

OutboardEMC3-EIRENE

Location of measurement

Island

Div. plate

fr25, 6.10 sfr18, 5.05 s

fr15, 4.60 sDivertor leg

Island

LCFS

LCFS LCFS

CII emission and Hγ/Hβ : Comparison with connection length (LC) plotDetach.

Island

Divertor leg

Div. plate

Stable detachment(with RMP)

fr25, 6.10 sfr18, 5.05 sfr15, 4.60 s

Divertor leg

Island

Connection length (LC)

10

100

1000

1

LC (m)

CII (426.7 nm)

CII radiates at LC~10 m region & along divertor legs. Clear upstream shift of emission at detachment transition. With RMP: In detach. phase, CII is stabilized outside of island (LCFS).

LCFS LCFS

M. Kobayashi et al., NME 12 (2017) 1043.11

Hγ/Hβ

Hγ/Hβ (index for volume recombination) : increase after detachment transition along LCFS. With RMP: Formation of very cold plasma around (outside) of island

Comparison with 3D modeling is underway …..

Magnetic field configuration & divertor probe arrays in LHDOpen field lines

Helical coils

One field period(Δφ=36°)

Top view of torus

Divertor probe arrays at inboard midplane

Midplane

RMP coils(m/n=1/1)

Helical coils

Plasma shape

12

0

0.2

0.4

0.6

0

50

100

1 2 3 4 5 6 7 8 9 10Toroidal section

0

0.2

0.4

0.6

0

50

100

1 2 3 4 5 6 7 8 9 10Toroidal section

L div.3D model

Exp.

R div.

101

102

103

85 90 95 100 105s (mm)

4L

101

102

103

95 100 105 110 115s (mm)

8R

101

102

103

100 105 110 115 120s (mm)

6L

Toroidal asymmetry of divertor flux caused by RMP application

RMP

No RMP

Wetted area increases with RMP.

Wetted area remains same.

RMPNo RMP

RMPNo RMP

Wetted area increases with RMP.

Higher flux is attributed to increased wetted area induced by RMP

Analysis of power load is underway (mostly same trend as the particle flux)13

The toroidal asymmetry can be rotated by toroidal phase shift of RMP

0

0.2

0.4

0.6

0

50

100

1 2 3 4 5 6 7 8 9 10Toroidal section

L div.

Shift by 36°

0

0.2

0.4

0.6

0

50

100

1 2 3 4 5 6 7 8 9 10Toroidal section

L div.3D model

Exp.

0

0.2

0.4

0.6

0

50

100

1 2 3 4 5 6 7 8 9 10Toroidal section

L div.

0

0.2

0.4

0.6

0

50

100

1 2 3 4 5 6 7 8 9 10Toroidal section

3D model

Exp.

Shift by 180°

14

0

0.2

0.4

0.6

1 2 3 4 5 6 7 8 9 10Toroidal section

Detach

AttachL div.

0

0.2

0.4

0.6

1 2 3 4 5 6 7 8 9 10Toroidal section

R div.

Detach

Attach

Toroidal asymmetry of divertor particle flux changes at detachment phaseAt certain sections, the flux even increases.The asymmetry during detachment phase can also be shifted by RMP phase shift.

0

0.2

0.4

0.6

1 2 3 4 5 6 7 8 9 10Toroidal section

L div.

0

0.2

0.4

0.6

1 2 3 4 5 6 7 8 9 10Toroidal section

R div.

RMP toroidal phase shift by 36°

15

0

2

4

ΔΦrin

pla

sma

(10-

4W

b)

0

10

20

30

3 4 5 6 7 8Time (s)

Isla

nd p

oloi

dal p

hase

Shift

(deg

ree)

ΔΦ ex

#115085

Detach Reattach

Shielded

Amplified

Relative to vacuum RMP

Significant plasma response to external RMP

Attached phase: RMP is shielded, island poloidal phase shift ~ 30 deg. Detached phase: RMP is amplified, island poloidal phase shift ~ 15 deg. Hysteresis in plasma response at detach reattach transition

Attach

Effect on divertorparticle/power deposition

16

Quantitative comparison of temporal evolution of radiated power with EMC3-EIRENE

Indication of the study:1. Larger Dimp is closer to experiments. Needs drifts (electric field)?, or turbulence?2. Impurity source should be reduced after detachment. 3. The modeling still overestimates radiated power as compared to experiments.

Sputtering coefficient or atomic data base?

S. Pandya et al., NF 56 (2016) 046002.

Experiments

EMC3-EIRENE

17

Comparison of radiation profiles with EMC3-EIRENE

Radiation distribution by bolometer from top view port

Experimental results are much broader than the simulations Dimp should be greater than the bulk plasma at least by a factor of 4

S. Pandya et al., NF 56 (2016) 046002.18

19

n e(1019m-3)

Radiative collapse

Detachment transition

Lower threshold of perturbation strength for sustained detachment

Density limit of radiative collapse seems independent of perturbation field strength. Upper bounds for density operation range

Detachment transition density decreases with increasing IRMP. Easy access to detachment & enough

margin for operation (but at the expense of Wp.)

No sustained detachment has been realized so far at Icoil < 1.9 kA ( <0.06 %)

Upper bounds for density

0

5

10

0 1 2 3

0 0.05 0.1

IRMP

(kA)

(%)/~ 0Bb

0/~ Bb

Sustained detach

Operation domain of stable detachment in LHD

M. Kobayashi et al., Nucl. Fusion 53 (2013) 093032.

Remnant island of m/n=1/1

LCFSConfinement region

islandLCFSx −Δ

0.00

0.05

0.10

0.15

Healing

vac0coilr )B/b(~

Stable detach.Radiation collapse

2.0x10-4 1.0x10-3 2.0x10-30

(m)

Key geometric parameters: ,islandLCFSx −Δ ,/~0Bbr

Separation between radiation region (island) & confinement region is important factor for stable detachment

Threshold for RMP amplitude to overcome plasma screening.

100

101

102

4.4 4.6 4.8 5R(m)

100

101

102

0.1%)B/b( 0coilr ≈~

0.05%)B/b( 0coilr ≈~

vacw

vacw

Stable detach

Radiation collapseNo Te flattening due to plasma healing

20

Main plasma confinement : Recovery of energy confinement after detachment transition due to pressure profile peaking

0.6

0.8

1

0.2 0.4 0.6 0.8 1Sudoce n/n

Without RMP

With RMP (attach)

With RMP(detach)

τ Eexp

/ (f re

n)

τ EISS0

4

Confinement enhancement factor vs ne

RD transition

Increase of ne leads to confinement degradation without RMP. Significant degradation in RMP attached phase due to large magnetic island in the edge. Energy confinement recovers after RD transition with RMP due to pressure peaking. The cause of the pressure peaking is under investigation.

0

5

10

15

3 3.5 4 4.5R(m)

With RMP(detach)

With RMP (attach)

Without RMP

pe profiles with & without RMP

21

2210-9

10-8

10-7

10-6

10-5

10-4

10 100 1000 10000

6I10.5U

Frequency(Hz)10 100 1000 10000

Frequency(Hz)

Attach Detach

Fluctuation of Isat

0

2

4

6

8

0

0.2

0.4

2.98 3 3.02 3.04 3.06 3.08Time(s)

Radiation

Isat

Isat

Fluctuation of magnetic probe & Isat

Magnetic probeHigh frequency components (several tens kHz) disappears after detach transition Low frequency (~5kHz) component

Div probePeaked at 60~90Hz after detach transition. Strong correlation with radiation oscillation.

With RMPDetach

Fluctuation of magnetic probe

Time (s)1 2 3 4 5 6 7

Freq

uenc

y (k

Hz)

0

20

40

60

80

100

SummaryEffects of RMP application on the detachment is being investigated in LHD.

1. The RMP (m/n=1/1) application leads to stable sustainment of detached plasma.2. Radiation is enhancement by RMP application

EMC3-EIRENE prediction: Poloidal asymmetric radiation due to islandAXUV line integrated profiles consistent with the modelingEUV measurements of CIII – CVI : Enhancement at X and O-pointsVisible CII and Hγ/Hβ : Enhancement outside/around of LCFS (signature of volume recomb.)

4. Divertor flux distribution with RMPToroidal asymmetry according to toroidal mode numberReduction during detached phase: asymmetric pattern changes from the attached phaseThe asymmetry can be rotated by RMP phase shift

5. Operation space Key geometric parameters : RMP amplitude and separation between edge island and confinement region

6. Geometrical effect on energy transport and radiation distributionPoloidally asymmetric radiation Energy flow from O to X-point can help to sustain the localized radiation

7. Confinement of the RMP assisted detachmentConfinement is recovered after detachment transition due to pressure peaking

23

Issues to be investigated furtherEffects of different impurity species: Ne, N, Ar etc in relation to the island TeMode number of RMPDecoupling effect between SOL & confinement plasma in terms of neutral penetration (RMP thicker SOL decoupling)