Studies of Improved Confinement in Heliotron J

Heliotron J

IAE, Kyoto University

Studies of Improved Confinement

in Heliotron J

Presented by

MIZUUCHI Tohrufor Heliotron J Team

Heliotron J IAE, Kyoto University

Studies of Improved Confinement in Heliotron J

Contents

Introduction

Transition to Improved Confinement─ Configuration Effects─ Effects of Heating Scenario

Fueling Control

Summary


Due to the time limitation, this talk will focused on “Transition”. For other topics, please contact me or refer the proceeding


<r> (m)

/2

4/7

4/8

8/15

8/17

8/13

12/22

Heliotron J (vacuum)

12/23

0.00 0.10 0.20

0.50

0.55

0.60

The edge magnetic surfaces are strongly modified by the resonant conditions.

STD

Ad. Ip

Sub. Ip

Vacuum Rotational Transform

i(a)/2p=0.493 i(a)/2p=0.498

i(a)/2p=0.56 i(a)/2p=0.62

The effects of rational surfaces have been studied is not only from the MHD activities but also from the appearance of transport barriers viewpoints.

HFC


Bumpiness Control B drift can be suppressed with bumpiness control.

1

1.1

1.2

1.3

1.4

1.5

1.6

0 45 90

|B|

Toroidal angle (deg.)

Magnetic field strength at plasma axis

High eb

Medium

Low

StraightCorner Corner

High eb : Higher mirror rippleLow eb : Mirror field reversal (at axis) Constant parameters :- Rax/<ap> = 1.2 m/0.17 m, Vp ~ 0.7 m3

- Edge rotational transform- Magnetic well in entire region

-0.2

-0.1

0

High b

MediumLow

B1

4/B0

0

Helical component

-0.2

-0.1

0

B1

0/B0

0

Toroidal component

-0.1

0

0.1

0.2

B0

4/B0

0

Bumpy component

0.50

0.55

0.60

0 0.2 0.4 0.6 0.8 1

2

(=s1/2)

7/4

15/8

m/n=13/8

Rotational transform


The Heliotron J experiments have revealed the existence of the spontaneous transition to improved confinement mode.

00.05

0.10.15

P (

A.U

.)

PECH

time (ms)200 300

00.10.2

Wb

(A.U

.)

Diamag

0.10.20.30.4

I (A

.U.)

Dalpha

#28407 H-modeDithering phase

#28407

0

0.5

1

1.5

time (ms)

I.S.

(a.

u.) Fast Langmuir Probe

200 300-1

0

1

I (a

.u.)

Magnetic Probe

ECH-only plasma

Observation with fast cameras revealed the existence of filament structure parallel to B in the peripheral plasma turbulence.

Phase image analysis suggests The CCW poloidal rotation of the filament

structure during the L-mode. During the L-H transition, this rotation

speed decreases. After the transition, it restarts the

movement but in CW direction.

Ref. Nishino, C09 (Tue)


The transition phenomena were observed almost all (a)/2 cases for the input power range in the experiment (ECH+NBI).

Some configurations show a remarkable improvement but the others show a slight improvement.

P = PECH ( 0.29 MW) + PNBI ( 0.57 MW)

Vp (m3), R (m)

(a)vac/2

Wp (

kJ)

70GHz ECH+NBI (-Sacn) Max. Wp

before

0.0

1.0

2.0

3.0

8/17 8/15 8/13

4/8 4/7 4/6n/m

W

p/W

pb

efo

re

HJ14990-HJ15456 ECH+NBI

0.45 0.50 0.55 0.60 0.650.0

1.0

2.0

3.0

Before the transition, Wp strongly depends on i(a)/2p|vac, but it is moderated after the transition.


One robust condition to the transition is the core density.The value of the threshold line-averaged density is not sensitive to

the heating method, Pinj level, i(a)/2p|vac.

The threshold density is a robust condition of the transition. The threshold density (lower density limit for the transition) is ~ 1-21019 m-3. Only for higher values of /2, some exceptional low threshold density cases

are observed. MHD activities?

STD configuration

Pinj (MW)

ne-b

ar

(10

19 m

-3)

ECHNBIECH+NBI

HJ09472-HJ12298 STD

0.2 0.4 0.6 0.8 1 1.20.0

1.0

2.0

3.0

ne

thr (

10

19 m

-3)

(a)/2

HJ ECH+NBI

0.45 0.5 0.55 0.6 0.650

0.5

1

1.5

2

2.5

38/17

4/8 4/7 4/6

8/15 8/13

n/m

PECH (~ 0.29 MW)+PNBI (~ 0.57 MW)


the “quality of improvement” based on ISS04 seems to depend on the configuration.

Iota-windows for high quality H-mode close to the low-mode rationals

All configurations basically have “stellarator shear” (vacuum) hard to get edge flux surfaces on the “right-side” of the rationals.

Difference in the poloidal viscous damping among the configurations can cause the difference in the “quality”?

─ Still open question.─ Bias experiment will give us

some information.

tex

p/(

ftIS

S0

4)

F. Sano, et al., Nucl. Fusion 45 (2005) 1557.

i(a)/2p|vac


3

Electrode of LaB6

(Hot Cathode)

Electrode Bias Exp. in Heliotron J under a Low Field ConditionS. Kitajima, et al., Tohoku Univ.

Externally controlled torque could give us a lot of (but NOT sufficient) information on the viscous damping.

Note: This experiment was performed under a special configuration for 2.45GHz ECH discharge.


Rapid increase of IE and ne.

Vs changed from positive to negative.

Drop in the fluctuation level.

Clear hysteresis of IE.

Triangular voltage was applied in order to give rise to forward/ back transition.

Time Evolution of a Biased Discharge


R (m)

Z (

m)

2.45GHz ECDCHeliotron J

0.8 1 1.2 1.4 1.6

-0.2

0

0.2

0.4

Bias Exp. suggests plasma rotation can modify the edge field topology.

Bias

time (s)1 2

0

2

4

0

1

2

HJ-SH365/090305

CH8

Divertor Probe Array

0

0.02

0.04

CH9

0

0.02

0.04

CH11

0

0.02

0.04

CH12

0

0.02

0.04

CH14

0

0.02

0.04

CH13

0

0.02

0.04 Time traces of ion-saturation currents at divertor probe array during biasing:Increase for some channels, but decrease for other channels. Modification of edge field topology!

Probe array


The “quality of improvement” based on ISS04 seems to depend on the configuration.

All configurations basically have “stellarator shear” (vacuum) hard to get edge flux surfaces on the “right-side” of the rationals.

Difference in the poloidal viscous damping among the configurations?

But, is that all? How about the “plasma effects”,

especially effects of (non-induction) plasma current.

Experiments suggest the change of the iota and/or the size of the confinement region. 　

tex

p/(

ftIS

S0

4)


i(a)/2p|vac


ad. 4kAad. 2kAad. 1kA

sub. 1kA sub. 2kA sub. 4kA

HINT2 predicts the change of the field topology caused by plasma current:

b(s) = 0.5·(1-s)2 [%], jp = jp0·(1-s)2

vacuum

0kA

“Additive” plasma current can modify not only the i/2p-profile but also the “shape” of magnetic surfaces.

Asymmetric effect of plasma current due to the “proximity” to a low-mode rational number.

<r> (m)

/2

(s)=b0(1-s)2, 0=0.5 %jp(s)=j0(1-s)2HINT2

Ip=4kA

2kA

-4kA

-2kA

1kA

-1kA0kA

4/812/23

8/15

4/7

8/13

12/22

0 0.1 0.2

0.50

0.60


Experiments suggest the change of the edge field topology during a plasma shot.

ECH ( 0.3 MW) + NBI ( 0.7 MW, co-injection) @ STD

The observed non-inductive toroidal plasma current is gradually increases as increase of the stored energy. – In the density range shown in the figure, the observed

plasma current is considered to mainly consist of the bootstrap current and the NB induced current.

Position of Divertor Leg

R (m

)

time (ms)

Lc PeakDensity Peak

200 250

1.40

1.45

NBIECH

nel

n el (

1019

m-2

)

HJ15449

H

H

(a.u

.)

0

0.5

1

1.5

0

0.5

Wp

Wp

(kJ)

Ip

Ip (k

A)

0

1

2

0

1

2

3

n e, d

iv (1

018 m

-3)

R=1.407 m

0

2

4

(1018

m-3

) (b) t = 277 ms

R (m)1.4 1.45

0

1

2

3

(1018

m-3

) (a) t = 211 ms

HJ15449

0

1

2

The profiles of the diverted plasma density at t = 211 and 277 ms.

–At 211 ms, the density distribution is consistent with that expected from the vacuum field topology.

–At 277 ms, the density peak position shifts inward about 4 cm compared to the position at 211 ms.

The overall trend of the density-peak-position shift seems to be well synchronized with the change of Wp or Ip.


Comparison between Co/CTR NBI-only plasmas

- No transition in CTR NBI -

In Co-NBI, the transition were observed, but no transition in CTR-NBI.

Difference in the transition condition between these NBI-only plasmas due to

Deformation of i/2p Position of rational surface, Shear, Shape of LCFS,

Direction of momentum input ?

0

2

W(k

J)

NBI

ECH

Co-NB

CTR-NB

#25892(Co-NB), #25989(CTR-NB)Config. ; (a)/2=0.54

0

2

n e(x

1019

m-3

)

-2

0

2

I p(k

A)

Co

CTR

0

2

4

6

180 200 220 240

I H(A

.U.)

time (ms)


The transition is observed when the plasma current reaches a critical value.

The critical value of the plasma current depends on the configuration; 0.7±0.1kA in middle eb

1.3±0.2kA in high eb cases. The observed time delay would be related

to the growing-up time of the current.

− In high eb case, higher Ip was

required.

I p @

Ha

dro

p (k

A)

Dependence of IP at transition on PNB

2

1

00 200 400 600

High b

Medium b

PPortthrough

(kW)

High eb

Medium eb

Dependence of delay time on PNB

dela

y tim

e (

ms)

0

10

20

30

40

50

0 200 400 600

High b

Medium b

PPortThrough

(kW)

The effects of the plasma current on the field configuration should depend on the vacuum rotational transform. i(a)/2p|vac-scan experiment.


2

1

00.45 0.5 0.55 0.6 0.65

(0)/2 (=1/q)

I p (

kA)

I p@

tra

nsi

tion

(kA

)

i(a)/2p|vac

4/8 4/78/17 8/15 8/1312/2212/23

(High eb)

The critical current for NBI-only plasma strongly depends on i/2pvac.

So far, the transition was observed only in Co-NBI (ad. NBCD) plasmas (Pinj < 0.6 MW).

A “critical current” for the transition exists for all config.

Does the change in the rotational transform caused by non-inductive plasma current trigger the transition? Check other heating cases.

Radial profile of iota

<r> (m)

/2

4/7

4/8

8/15

8/17

8/13

12/22

Heliotron J (vacuum)

12/23

0.00 0.10 0.20

0.50

0.55

0.60

0.25 MW < Pinj < 0.6 MW, ne ~ 1.5-21019 m-3

0.45 0.50 0.55 0.60 0.65

0.50

0.55

0.60

12/22

8/13

4/7

(a)/

2

Vacuum Iota

4/8

8/15

HJ Close to Rationals?


Comparison of the plasma current at the transition for ECH-, NBI-, and ECH+NBI-plasmas

2

1

00.45 0.5 0.55 0.6 0.65

(0)/2 (=1/q)

I p (

kA)

I p@

tran

sitio

n (k

A)

i(a)/2p|vac

4/8 4/78/17 8/15 8/1312/2212/23

(High eb)

((a)/2)VAC

I p (

kA)

ECHdensity ramp-up

HJ15567-15587mid.-b

high-b

low-b

low-b

HJ28256-28275(ECCD, low density)

HJ18494-18860

HJ18365-18483

HJ18413-18929

0.45 0.5 0.55 0.6 0.65

0

1

2

3

NBI-only

Ip,NET itself seems not a key factor? Need profile data!

((a)/2)VAC

I p (

kA)

ECH+NBIdensity ramp-up HJ14990-15453

0.45 0.5 0.55 0.6 0.650

1

2

3ECH-only

ECH+NBI


Even for the same net current, the deformation of i/2p and the shape of the magnetic surface strongly depend the current profile.

- MODEL CALCULATION -

<b> ~ 0.32 %, p=p0(1-s2)2

(a) j = j0(1-s)8, Inet = +2.0 kA(b) j = j0(1-s)8, Inet = -2.0 kA(c) j = j0(1-s4)2, Inet = +2.0 kA(d) j = j0(1-s4)2, Inet = -2.0 kA(e) j = j0(60(1-s2)s2-5(1-s2)),

Inet = +2.0 kA(f) j = j0(60(1-s2)s2-5(1-s2)),

Inet = -2.0 kA

r/a (=s1/2)

current profile model

curr

ent/

unit

are

a (a), (b)

(c), (d)

(e), (f)

0 0.2 0.4 0.6 0.8 1-5

0

5

10

15

20


Comparison of the plasma current at the transition for ECH-, NBI-, and ECH+NBI-plasmas

2

1

00.45 0.5 0.55 0.6 0.65

(0)/2 (=1/q)

I p (

kA)

I p@

tran

sitio

n (k

A)

i(a)/2p|vac

4/8 4/78/17 8/15 8/1312/2212/23

(High eb)

((a)/2)VAC

I p (

kA)

ECHdensity ramp-up

HJ15567-15587mid.-b

high-b

low-b

low-b

HJ28256-28275(ECCD, low density)

HJ18494-18860

HJ18365-18483

HJ18413-18929

0.45 0.5 0.55 0.6 0.65

0

1

2

3

NBI-only

Ip,NET itself seems not a key factor? Need profile data!■ Even in low density ECH-only plasma lower than the threshold

density, the transition can be observed when the EC driven “additive” current is high enough.

■ Transition is observed in ECH+NBI plasmas even in the low eb case.

((a)/2)VAC

I p (

kA)

ECH+NBIdensity ramp-up HJ14990-15453

0.45 0.5 0.55 0.6 0.650

1

2

3ECH-only

ECH+NBI


Transition to a better confinement mode was observed in a high ECCD current discharges.

Low-eb, i/2p(a) ~ 0.56, ECH-only with PECH > 0.3 MW

The line-averaged density is much lower than the critical low-density limit observed in the previous experiments.

No enough experiments, from this point of view, in sub. high ECCD condition.

HJ28275

ECH

Gas Puff(a.u

.)

0.0

1.0

2.0

H#11.5

(a.u

.)

0

0.01

0.02

1019

m-2

mW

b

0

0.2

0.4

0

0.1

kA

Ip

time (ms)

ad.

200 250 300-3.0

-2.0

-1.0

0.0


Summary (1) - Transition to improved confinement -

No transition in CTR-NBI-only plasmaExistence of the “critical current” in Co-NBI-only plasma Change in the rotational transform (and/or its radial profile) caused by plasma current and/or the direction of momentum input can trigger the transition in NBI plasma.

No clear current dependence in ECH-only plasma Difference in the transition condition between NBI-only and ECH-only plasmas. ECH should have some effect on the transition condition in NBI plasma.

In low density (lower than the threshold density) ECH-only plasma, the transition can be observed when the EC driven “additive” current is high enough. Change in the rotational transform (and/or its radial profile) is effective even for ECH-only plasma.


ECH-onlyNBI-only

(Co)NBI-only

(CTR)

NBI (Balance)ECH+NBI(Balance)

ECH+NBI (Co)

ECH+NBI(CTR)

i(a)/2p|vacX only for

0.493 OK X No Data OK No Data

eb* △

for high-eb

X for low-eb

X No Data OK No Data

Ip,th** No Yes - Yes

remarks

Transition is observed in low ne,

high Ip

Summary (2) - Transition to improved confinement -

for ne > ne,th

* Data for different i(a)/2p|vac are necessary.** Data @ Ip=0 are necessary.


Thank you for your attention! &

Thank you All Contributors for Nice Collaborations!

F. Sano, K. Nagasaki, H. Okada, T. Minami, S. Kobayashi, S. Yamamoto, S. Ohshima, K. Hanatani, S. KonoshimaInstitute of Advanced Energy, Kyoto University

Y. Nakamura, K. Mukai, S. Kishi, H.Y. Lee, K. Minami, Y. TakabatakeGraduate School of Energy Science, Kyoto University

T. Mutoh, S. Okamura, Y. Takeiri, K. Y. Watanabe, M. Yokoyama, K. Nagaoka, Y. Yoshimura, Y. Suzuki, H. Takahashi

National Institute for Fusion Science

N. NishinoGraduate School of Engineering, Hiroshima University

S. MurakamiGraduate School of Engineering, Kyoto University

S. Kitajima, H. UtohGraduate School of Engineering, Tohoku University

Y. NakashimaPlasma Research Center, University of Tsukuba

I. PankratovNational Science Center, Kharikov Institute of Physics and Technology, Ukrina

Q. Yang, L. Yao, W. ChenSouthwestern Institute of Physics, China


No clear threshold power for Phase-I

No clear/simple relation between nth and Pabs.

At the (a)/2 0.493, ECH-only discharges cannot make a transition. configuration effects on the power threshold.

Different power dependence for Phase-I and II?


Studies of Improved Confinement in Heliotron J

Documents

Transcript of Studies of Improved Confinement in Heliotron J