LHD Project,National Institute for Fusion Science

Introduction of the pellet injection system and relevant experiments in LHD

Ryuichi Sakamoto National Institute for Fusion Science

14th Coordinated Working Group Meeting (CWGM)

/10Pellet injectors in LHD

❖ Two types of solid hydrogen pellet injection systems are installed in one place • Pneumatic pipe-gun type injector

- Simplest concept of all pellet injector - Independently controlled 20 pellets injection

• Screw extruder type repetitive pellet injector - Consecutive pellet injection up to 11 Hz without time limits

• Common facilities - Differential pumping, Diagnostics (Velocity meter, Ablation observation)

2

Differential pumping system

First

expansion

chamber

Second

expansion

chamber

Final

expansion

chamber

Pipe-gun

Pellet InjectorLHDRepetitive

Pellet Injector

Valve set

Cryo-

chamber

Gap &Light gate

/10Pneumatic pipe-gun type injector

❖ Conventional in-situ pipe gun concept3

✓Reliable operation - No movable parts in cryo - Fixed pellet diameter at inner

diameter of barrel ✓Adequately high velocity

(~1,300 m/s) for LHD ✓Single pellet injection per barrel

- Multiple barrels are required - 20 barrels for LHD (30 m3) • 3.0 mm (~1.0×1021) ×10 pellet • 3.4 mm (~1.5×1021) ×6 pellet • 3.8 mm (~2.0×1021) ×4 pellet

High pressure He gas Pressure: 2.5 - 5.0 MPa

4)

H2 gas feed flow rate: 20-100 scc/min Pressure: 5 - 100 kPa

Vacuum

2)

Occluded by solid-hydrogen

Vacuum Vacuum

3)

Solid-hydrogen pellet

H2 gas feed flow rate: 20-100 scc/min Pressure: 5 - 100 kPa

Vacuum

1)

Solidified hydrogen

~10 K

/10Screw extruder type repetitive pellet injector

❖ Continuous production of solid hydrogen by screw extruder • Liquefaction and solidification

are simultaneously processed • Solid hydrogen is continuously-

extruded by screw - Essentially steady state operation

❖ Pneumatic acceleration

4

Fast ValveCutter

Heat

Exchanger

Thermal

Anchor

40 K

Cryostat

Cold Head

Pumping

Cold Head

Heater

Hydrogen

HeliumPellet

Motor

Schematic of Screw Extruder

2.5 mm

Solid Hydrogen rod

Barrel & Cutting Device

Extrusion Nozzle

Extrusion

• Pellet size: 1.4 mm (~1.0×1020) • Pellet velocity: 200 — 500 m/s • Injection frequency: up to 11 Hz

/10Pellet penetration depth

❖ Typical pellet penetration depth: half radius ~ core ❖ Pellet penetration depth is strongly depend on target

plasma conditions • Pellet ablation affected by fast ion originate from NBI • Insensitive to pellet size and velocity

5

1.0

0.8

0.6

0.4

0.2

0

λ/a

3210Te(0) [keV]

1.4 mm, 250 m/s

3.0 mm, 1100 m/s

3.4 mm, 1100 m/s

3.8 mm, 1100 m/s

NGSwith fast ion ablation

Plasma and Fusion Research: Regular Articles Volume 1, 033 (2006)

Fig. 6 New scaling using the stored energy of fast ions Wf0 in-stead of the electron density ne0 compared to the mea-sured pellet penetration depth in LHD.

the difference between measured and predicted pellet pen-etration is large for deep penetration although for shallowpenetration there is no difference between the measuredand predicted results in the two cases. The discrepancy ofexperimental results from NGS model suggests the effectof fast ions on pellet ablation since NGS scaling considersonly thermal electrons. This tendency is pronounced whenthe pellet penetrates deeply and Wf0 is high. It is concludedthat the experimental data produced from LHD cannot beexpressed by NGS scaling alone.

In order to derive an empirical expression of the pene-tration depth, regression analysis was applied. The energyof fast ions Wf0 is used as a parameter in the statistical anal-ysis since it has a potential to express the effect of fast ionson pellet ablation. The obtained regression expression isthe following formula:

λ/a = 0.269T−0.256±0.069e0 W−0.136±0.024

f0

× m0.263±0.047pel v0.144±0.030

pel . (7)

As shown in Fig. 6 (RMSE = 0.026), the regression ex-pression for the penetration depth accounts for the mea-sured pellet penetration. The energy of fast ions Wf0 asa parameter is requisite to predict the measured penetra-tion depth since the estimate value (i.e., the error bar of themultiplier factor) is smaller when the electron density ne0

instead of Wf0 is used.As described in Sec. 2, the new ABLATE code [19]

is employed to analyze the pellet penetration depth in theLHD experiments. For this calculation, the fitting elec-tron temperature and density profile based on experimentalmeasurements using the Thomson scattering and the FIRinterferometer, and the density profile of fast ions calcu-lated by the FIT code [23] are used. The ABLATE codeconsidering not only thermal electrons’ but also fast ions’contribution to the ablation can calculate the ablation rate

Fig. 7 Hα profile compared to the calculated ablation rate profilein the case of high Wf0 (mpel = 5.30 × 1020 atoms, vpel =

1129.70 m/s, and Wf0 = 42.12 kJ).

Fig. 8 Hα profile compared to the calculated ablation rate profilein the case of low Wf0 (mpel = 6.19 × 1020 atoms, vpel =

1148.40 m/s, and Wf0 = 16.14 kJ).

profile in Figs. 7 (for the high Wf0 case) and 8 (for thelow Wf0 case), where the time-dependent profiles of theelectron temperature and density during pellet ablation areconsidered in the calculation. The profile of Hα emission(solid lines), the model of electrons only (dashed lines),and the model of electrons and fast ions (filled circles) arecompared. The measured pellet penetration depth is com-pared with the ABLATE penetration depth, i.e., the widthof the calculated ablation rate profiles. Figure 9 shows acomparison between the ABLATE penetration depth con-sidering only the ablation of thermal electrons and the mea-sured penetration depth. This comparison duplicates theresult from the NGS scaling shown in Fig. 5. A differencein the trends of the cases of high and low Wf0 is also ob-served. The comparison of the ABLATE penetration depthconsidering the ablation of thermal electrons and fast ionsis shown in Fig. 10. The NGS scaling systematically un-

033-6

/10Density control with PI timing

❖Real-time PI timing control with simple comparison between set-value and density signal (interferometer or bremsstrahlung) enables density control

❖ Identical confinement property is shown in transient- and long-pulse discharges (except high recycling shots)

6

0

10

20

30

0

0.5

1.0

0

1

2

3

4

0.4×1020 m-3,1.2 MW

0.7×1020 m-3, 1.7 MW

1.0×1020 m-3, 3.3 MW

#52762#52605

n e 1

020 [m

-3]

Time [s]

#63336

PN

BI

[MW

]

#63336

#52605#52762#56112

#56112ramp-up

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0

0.1

0.2

0.3

0.4

0.5

0.6

Wp/

Pab

s0.41

[A

.U.]

�ne [1020 m-3]

Large Pellet #60783

#63336

Gas puff #39611

ISS95

Repetitive Pellet

#52605

#56122

/10Ablation light observation (Fast camera)

❖ Fast camera observation with bifurcated imaging optical fiber scope • Stereoscopic measurement: observe from different positions

- 3-D pellet ablation phenomena

• Imaging spectroscopy: observe at same position with filters - Temperature and density measurement

7

Helical Coil

Helical Coil

Pellet Injection18°

line of sight

line of sight

object lens

object lens

15°

15°

X

YZ

Z [m]

01.02.03.04.05.0

Magnetic

axis

Plasma

boundaryR [m]

2.0 3.0 4.0 5.0 6.0 7.0

15,000 mm

1,2

40 m

m

Fast imaging camera

Bifurcated imaging

optical fiberscope

#55639 @0.555881 s, tint=50 μs, texp=3 μs

Em

ission intensity [a.u.]

L image R image

Pellet ablatant

Breakawayplasmoid

Vpel =380 m

/s

Hα Inte

nsity [a.u

.]

Abla

tion r

ate

[a.u

.]

4.6

4.4

4.2

4.0

3.8

3.6

R

[m

]

0.5580.5570.5560.555

Time [s]

Pellet ablatant

Breakaway plasmoid

NGS calculated

ablation rate

Pellet ablation

light (Hα) intensity

10 cm

LCFS

Ablation duration

#55639

Core Ablation

(a)

(b)

/10Ablation light observation (Photo diode matrix)

❖ Complementary observation with fast camera • Lower spatial resolution • Faster time resolution

❖ Plasmoid drift Tracking

8

10

1

20

91

100

11

② Breakaway plasmoid

ch. 25

ch. 36

ch. 47

ch. 58

2.666561 s

2.666547 s

2.666542 s

2.666538 s

10 km/s

30 k

m/s

5

4

3

2

1

0

ΔC

hann

el

50403020100ΔTime [μs]

ch. 60

ch. 59

ch. 58

ch. 57

ch. 56

ch. 55

ch. 54

ch. 53

① Pellet trajectory

0 100 200 300 400

ΔC

hann

el

ΔTime [μs]

1.25 km/s

/10PI synchronized TS measurements

❖ Real time pellet position measurement using TOF • TS laser is triggered by real time

calculation with FPGA

❖ Multiple TS profile measurements during pellet ablation • Detect of pellet plasmid before

homogenization

9

6

5

4

3

2

1

0

n e

[1020

m-3

]

4.54.03.53.02.5R [m]

3.733580 s 3.746636 s 3.746836 s 3.766912 s

Abl

atio

n Li

ght

[a.u

.]

3.74723.74683.74643.7460Time [s]

3.746636 s 3.746836 s

4

3

2

1

0

T e

[keV

]

4.54.03.53.02.5R [m]

3.733580 s 3.746636 s 3.746836 s 3.766912 s

First

expansion

chamber

Second

expansion

chamber

Final

expansion

chamber

LHD

D(2)

D(3)

DLCFS

DTS

t=0tLG

tD+t

TS

LGU= 2, LGD= 3

の時

LG1LG2LG3

= + － ( )( ) － ( ) －

/10Summary

❖ Two types of solid hydrogen pellet injection systems is installed in LHD • Pneumatic pipe-gun type injector

- 1 - 2×1021 H/pellet - 1000 - 1300 m/s - 20 barrels

• Screw extruder type repetitive pellet injector - 1×1020 H/pellet - 200 - 500 m/s - 11 Hz, Steady state

❖ Detailed pellet ablation phenomena observations • Fast camera

- Three-dimensional imaging with stereoscopic measurement - Imaging spectroscopy for plasmoid parameter measurement

• Fast imaging with photo-diode matrix • PI timing synchronized TS measurements

10