A detailed study of carbon chemical erosion R C E O V D III D - - PowerPoint PPT Presentation

▶
a detailed study of carbon chemical erosion
SMART_READER_LITE
LIVE PREVIEW

A detailed study of carbon chemical erosion R C E O V D III D - - PowerPoint PPT Presentation

Feb 22, 2023 330 likes •561 views

O F Y W T I I S S A detailed study of carbon chemical erosion R C E O V D III D N I S N I U N NATIONAL FUSION FACILITY S A N D I E G O M N O A in L-mode plasmas in the DIII-D divertor D I S D. G. Whyte, University

slide-1
SLIDE 1

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

1

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

A detailed study of carbon chemical erosion in L-mode plasmas in the DIII-D divertor

  • D. G. Whyte, University of Madison - Wisconsin

J.N. Brooks, Argonne National Laboratory P.C. Stangeby, University of Toronto N.H. Brooks, General Atomics

10th international workshop on carbon materials for fusion application September 17-19, 2003 Jülich, Germany

slide-2
SLIDE 2

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

2

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Outline

  • L-mode plasmas for carbon erosion studies
  • Erosion modeling and interpretation
  • Erosion with attached divertor plasma

Divertor tile vs. main-wall tile Determination of Ychem at outer strikepoint Atomic carbon velocity distribution

  • Effect of plasma detachment on carbon erosion.
  • Discussion & Summary
slide-3
SLIDE 3

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

3

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

H-mode plasma studies showed unexpected reduction in carbon erosion in the DIII-D divertor

  • Encouraging results on the use of carbon:

In-situ Ychem reduction at lower divertor. Detachment, necessary for heat flux control, greatly reduces HC signals

  • …but tentative results:

Inconsistent plasma conditions over long-term study ELMs in H-mode complicate interpretation of erosion and spectroscopy.

80,000 90,000 100,000 Shot number 1 10 0.1

Ychem (%)

1992 2000 Year

ISP of new Upper divertor

OSP Lower divertor

Brightness D-α (ph s-1 m-2 sr-1)

D-β / D-α

0.1 10 1019 1020 time (ms)

2000 3000 4000

CD / D-α

10-3 10-2

Outer strikepoint (OSP) spectroscopy

C2 / D-α

(x10)

D2 gas injection (a.u.)

detached

slide-4
SLIDE 4

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

4

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

L-mode, simple-as-possible plasmas ideal for carbon studies

  • Low power leads to ~

constant Tsurf~375 K

  • No ELMs
  • Density control leads to

good detachment control

  • Multiple discharges

Improved DTS statistics Redundant divertor diagnosis Multiple C & HC emissions measured.

105500 105502 105503 105504 105505 105513 105506 105507 105508 105509

Ψn Is (A/m2)

0.0 4 104 8 104 1.2 105 1.6 105

0.90 1.00 1.10 1.20

+

1020 1021 1 10 100

1022 1023

Te (eV) ne (m-3)

Γi (m-2 s-1)

DTS Langmuir probe

line-average density (1019 m-3) 2.5 3.5 4.5 5.5

slide-5
SLIDE 5

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

5

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

High resolution spectroscopy and divertor sweeping diagnose erosion over wide variety of surfaces

  • Divertor tiles made from ATJ

graphite, an isostatically molded fine grain graphite

  • Multiple (>50) boronization layers

applied over +10 year lifetime.

  • Absolute wavelength

calibration from discharge lamps during plasma shot (+/- 0.001 nm ~ 300 m/s).

  • Can resolve TC < 0.5 eV.

#105505.2100 #105505.4900

baffle

V1 V2 V3 V4 V5 V6 V7

Langmuir probe clusters on lower divertor tiles

45 degree tile Row#1 Row#2 Row#3 Row#4

  • uter

ring High spectral resolution divertor spectroscopy (MDS) viewchords

DTS

slide-6
SLIDE 6

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

6

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

WBC Monte-Carlo code is used to interpret HC spectroscopy

  • Full dissociation chains of methane &

higher order HC’s

  • MOLDYN reflections vs. E
  • Full HC spectrum launched into OSP

plasma (DTS) with sonic flow to plate.

  • Particle followed until redeposition or

leave simulation zone (~5-10 cm)

  • Added C2 and C3 rates for C2

spectroscopic interpretation.

Close to C for ionization & diss. CX negligible in H plasma

  • Excitation rates of CD, C2, C I and

CII vs. Te,ne to calculate expected emission --> photon efficiency.

43.5% CD4, 3.7% C2D2, 24.8% C2D4, 11% C2D6, 16% C3D6

CI photon efficiency XB S      

CI ,910nm

= ne

i

∑

XCI ,910nm(Te) ∆ti,CI

C atom trajectories

slide-7
SLIDE 7

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

7

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Atomic carbon velocity distribution can be an indicator of erosion source

  • WBC computes emission

weighted f(vz) arising from HC dissociation into C I.

  • Thompson model with light-

ion energy cutoff/correction predicts direct CI f(vz) from D+ on C physical sputtering.

dfv(E) dv ∝ E3 / 2 (E + EB)3       1 − EB + E γ (1 − γ ) E      

1 / 2

       

γ = 4mCmD mC + mD

( )

2 ~ 0.49

vz (103 m/s) fCI(v)

1 10 20

  • 10

WBC calculation of emission-weighted atomic carbon velocity distribution Thompson velocity distribution for 100 eV D+ on carbon (normal incidence_

slide-8
SLIDE 8

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

8

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

The main/inner wall tiles has 5-6 times higher Ychem than the inner divertor tiles

  • V1 is a rare location for ISP,

small particle/energy fluence.

  • Spectroscopy verifies

~identical ISP plasmas at two locations: Te ~ 10 eV ne~1.5x1019 m-3

  • Boron (BD) higher from

inner wall.

2

Spectral Intensity at ISP (1019 ph s-1 m-2 sr-1 nm-1)

434 432 430 428 516 515 517

0.8 100

656 657 658 659

1 2

909 908 910

wavelength (nm) C II CD BD D-γ

V1 V2

C II C2 C I D-α C II

#105505.2100 #105505.4900

baffle

V1 V2 V3 V4 V5 V6 V7

Langmuir probe clusters on lower divertor tiles

45 degree tile Row#1 Row#2 Row#3 Row#4

  • uter

ring High spectral resolution divertor spectroscopy (MDS) viewchords DTS

slide-9
SLIDE 9

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

9

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Attached outer strikepoint is dominated by physical sputtering , Ychem= 0.3%

  • Incident plasma: Te=20 eV, Ei~5 T=100 eV, ne ~2.5x1019 m-3
  • Matches of CD/C2 ratio gives confidence in HC modeling.
  • Match of CII/CI ratio gives confidence in ion transport modeling.

CD (431 nm) C2 (516 nm) C I (910 nm) C II (514 nm) 4π x Brightness (ph s-1 m-2) 1017 1018 1019 1020

Experiment WBC, Ychem=0.3% WBC, Yphys=1.8%

slide-10
SLIDE 10

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

10

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Neither erosion model fits the CI spectral features.

  • Calculated f(vz) convoluted with

spectrometer instrumental function for comparison to measured CI spectra.

  • Discrepancy with sputtering models

unresolved.

Physical: Teff ~ 1 eV OK, shift too large Chemical - WBC: shift OK, but Teff~3 eV too large.

  • N.B. chemical erosion can actually

lead to higher Teff,CI than physical sputtering

λ - λ0 (nm)

  • 0.05

0.05 Normalized intensity

1

data Thompson Thompson, λ shifted chemical erosion, WBC

vz (103 m/s) fCI(v)

1 10 20

  • 10

WBC calculation of emission-weighted atomic carbon velocity distribution Thompson velocity distribution for 100 eV D+ on carbon (normal incidence_

slide-11
SLIDE 11

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

11

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

WBC modeling predicts increasing photon efficiency in detached plasmas

  • C2 is particularly interesting

case: in 1 eV plasma no e- impact ionization /dissociation but readily excited by e- impact (Eth~2.4 eV).

  • In qualitative agreement with

C2D4 injection on JET at high density…C2 most easily excited (Stamp et al.)

CASE WBC-20 WBC-21 WBC-22 Plasma parameters at outer strikepoint Te (eV) 20 5 1 ne (m-3) 2.5e19 1.05e20 5.6e20 Photon-emission excitation rate coefficients (m3 / s) C I (910 nm) 1.7e-15 1.5e-17 5.e-19 CD (431 nm) 5.6e-15 7e-15 1.5e-15 C2 (516 nm) 2e-14 4e-14 1.16e-14 C+ (514 nm) 5e-16 nil nil Photon efficiencies: Full hydrocarbon spectrum launched C I 4.4e-03 1.6e-3 1.7e-3 CD 5.1e-2 0.45 0.22 C2 1.1e-2 0.83 9.8 C II 4.2e-3

slide-12
SLIDE 12

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

12

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Detachment strongly suppresses signatures of chemical erosion at the OSP

  • HC brightness decreases to or

below detection limits (open

symbols) in detachment.

1020 1021 1 10 100

Te (eV) ne (m-3)

line-average density (1019 m-3) 2.5 3.5 4.5 5.5 1017 1018 1016 1017 1017 1018

Brightness (ph s-1 m-2 sr-1)

CD (431 nm) C2 (516 nm) BD (433 nm) SAPP #105079 #105196

detached

515 516 517

attached detached C2 Swan band at OSP

wavelength (nm)

slide-13
SLIDE 13

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

13

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Detachment strongly suppresses signatures of chemical erosion at the OSP

  • HC brightness decreases to or

below detection limits (open

symbols) in detachment.

  • BD behavior significant:

Must radiate in detached plasma (MFP ~1 mm) Verifies Te ~ 1 eV to sustain BD emission. Ultra-low Te cannot be cause of extinction of HC emission, since Eth ~ identical between BD & CD.

1020 1021 1 10 100

Te (eV) ne (m-3)

line-average density (1019 m-3) 2.5 3.5 4.5 5.5 1017 1018 1016 1017 1017 1018

Brightness (ph s-1 m-2 sr-1)

CD (431 nm) C2 (516 nm) BD (433 nm) SAPP #105079 #105196

detached

slide-14
SLIDE 14

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

14

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Detachment strongly suppresses signatures of chemical erosion at the OSP: Ychem ≤ 10-4

1020 1021 1 10 100

Te (eV) ne (m-3)

line-average density (1019 m-3) 2.5 3.5 4.5 5.5 1017 1018 1016 1017 1017 1018

Brightness (ph s-1 m-2 sr-1)

CD (431 nm) C2 (516 nm) BD (433 nm) SAPP #105079 #105196

detached

1022 1023 10-3 10-4 10-2

10-3 10-4 10-5 10-6

Γi (m-2 s-1)

Ychem (C/+)

based on CD based on C2

line-average density (1019 m-3)

2.5 3.5 4.5 5.5

from Langmuir probe

detached

slide-15
SLIDE 15

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

15

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

ISP behaves nearly identical to OSP:

No apparent difference between locations in Ychem

  • Implies Y chem ~0.3% in attached case.

No DTS for modeling.

  • No difference between net erosion

dominated OSP (stars) and sooty, redeposited ISP tiles.

1017 1018 1016 1017

Brightness at ISP (ph s-1 m-2 sr-1)

CD (431 nm) C2 (516 nm)

BD (433 nm) 1017 1018 1018 1019 D-γ (434 nm)

1021 1022

Te (eV)

Γi (m-2 s-1)

10 20

D-γ BD CD

Outer midplane

106 105

Brightness (counts)

line-average density (1019 m-3)

2.5 3.5 4.5 5.5 SAPP #105079 #105196 SAPP: V1

from Langmuir probes

detached ISP

vs. 〈ne〉

slide-16
SLIDE 16

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

16

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

ISP behaves nearly identical to OSP:

No apparent difference between locations in Ychem

  • Implies Y chem ~0.3% in attached case.

No DTS for modeling.

  • No difference between net erosion dominated

OSP (stars) and sooty, redeposited ISP tiles.

FURTHER OBSERVATIONS

  • BD presence: Te can support HC emission.
  • Lack of HC emission --> little chemical

erosion in detached ISP.

  • In stark contrast to the divertor, main-wall

chemical erosion increases with ~ constant yield.

1017 1018 1016 1017

Brightness at ISP (ph s-1 m-2 sr-1)

CD (431 nm) C2 (516 nm)

BD (433 nm) 1017 1018 1018 1019 D-γ (434 nm)

1021 1022

Te (eV)

Γi (m-2 s-1)

10 20

D-γ BD CD

Outer midplane

106 105

Brightness (counts)

line-average density (1019 m-3)

2.5 3.5 4.5 5.5 SAPP #105079 #105196 SAPP: V1

from Langmuir probes

detached ISP

vs. 〈ne〉

slide-17
SLIDE 17

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

17

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Divertor atomic carbon response to detachment varies greatly from HC

  • CI brightness is ~

proportional to ion flux.

  • Change in Zeeman splitting

indicates CI emitted from ionization front.

1 4

1 2 3 20 40

line-average density (1019 m-3)

2.5 3.5 4.5 5.5

solid symbols: OSP (V6)

  • pen symbols: ISP (V2)

SAPP #107403

CI Brightness (1018 ph s-1 m-2 sr-1)

Teff,CI (eV)

2

vCI (m/s)

102 103 104

∆ZCI emission (mm)

Thompson chemical erosion WBC

chemical erosion: WBC most probable velocity

from Zeeman splitting

models

Attached Detached

CI emission

ionizing plasma recombining plasma

divertor plate

slide-18
SLIDE 18

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

18

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Divertor atomic carbon response to detachment varies greatly from HC

  • CI brightness is ~

proportional to ion flux.

  • Change in Zeeman splitting

indicates CI emitted from ionization front.

  • Little or no change in Teff,

contradictory to chemical erosion.

  • Doppler shift remains

inconsistent with physical sputtering.

1 4

1 2 3 20 40

line-average density (1019 m-3)

2.5 3.5 4.5 5.5

solid symbols: OSP (V6)

  • pen symbols: ISP (V2)

SAPP #107403

CI Brightness (1018 ph s-1 m-2 sr-1)

Teff,CI (eV)

2

vCI (m/s)

102 103 104

∆ZCI emission (mm)

Thompson chemical erosion WBC

chemical erosion: WBC most probable velocity

from Zeeman splitting

models

slide-19
SLIDE 19

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

19

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Discussion on dependence of location for Ychem

  • Redeposited layers, e.g inner divertor do not have higher

chemical erosion.

No difference between ISP & OSP, regions yet dominated by net erosion &deposition respectively. At strikepoints, almost all C atoms/HC eroded are from a deposited film, since prompt redepositon on ~90%.

  • Results indicate against importance of boron in reduction.

Slighter higher Ychem after boronization Upper inner wall and main-wall, strong boron presence with no apparent reduction

  • Large energy / particle fluence in divertor remains as the

“cause” in the relative reduction.

slide-20
SLIDE 20

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

20

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

Discussion on the (near) extinction of HC emission in detachment.

  • Accuracy of yield < 10-4 unknown, but the general results of

WBC follow from simple examination

The HC dissociation chain is simplified by the lack of ionizing event, → the HC should produce C2 &/or CD BD shows C2 and CD should radiate efficiently.

  • The absence of C2 &/or CD argues strongly against:
  • 1. The importance of chemical erosion as a carbon source for

the plasma (does not produce C ions) and

  • 2. A large role for chemical erosion in determining net erosion

/ deposition (MFP of HC molecules < mm << MFP CI, CII).

slide-21
SLIDE 21

M A D I S O N U N I V E R S I T Y O F W I S C O N S I N

21

Carbon Workshop, Sept. 2003, Whyte

NATIONAL FUSION FACILITY

S A N D I E G O

D III –D

In summary we have verified and further quantified previous H-mode results of carbon erosion, with some new obervations

  • Chemical erosion is weak in the DIII-D lower divertor,

Ychem ~0.3 % with attached plasma.

  • Same shot comparison indicates that the divertor tiles have

less chemical erosion than main, inner wall.

Boron does not appear to be cause of reduction.

  • Detachment eliminates the spectroscopic signature of

chemical erosion, with a inferred yield through modeling < 10-4.