The future of MSE: Towards real-time measurements and current - - PowerPoint PPT Presentation

▶
the future of mse
SMART_READER_LITE
LIVE PREVIEW

The future of MSE: Towards real-time measurements and current - - PowerPoint PPT Presentation

Feb 09, 2023 163 likes •512 views

The future of MSE: Towards real-time measurements and current profile control in ITER and KSTAR M. F. M. de Bock a,b , D. U. B. Aussems a , R. Barnsley b , J. Chung c , G. Hommen a , J. A. van Hooft a , R. T. Huijgen a , J. Ko c , D. Johnson d ,

slide-1
SLIDE 1

The future of MSE:

Towards real-time measurements and current profile control in ITER and KSTAR

  • M. F. M. de Bocka,b, D. U. B. Aussemsa, R. Barnsleyb, J. Chungc,
  • G. Hommena, J. A. van Hoofta, R. T. Huijgena, J. Koc, D. Johnsond,
  • F. M. Levintone, M. Scheffera, B. Strattond, and M. Walshb

aEindhoven University of Technology, Eindhoven, The Netherlands bITER Organization, Saint-Paul-lez-Durance, France cNational Fusion Research Institute, Daejeon, Korea dITER-US, Princeton Plasma Physics Laboratory, Princeton, NJ, U.S.A. eNovaphotonics, Princeton, NJ, U.S.A.

slide-2
SLIDE 2

Motional Stark Effect

PAGE 1 3/3/2013

  • Current or q-profile
  • Instabilities, improved confinement, steady state …
  • Current profile control  real-time actuators, controller, sensor

ρ ρ T q

1 2 3

slide-3
SLIDE 3

Motional Stark Effect

PAGE 2 3/3/2013

  • MSE (Motional Stark Effect) diagnostic
  • Local, spectroscopic measurement of internal B-field (parts of it)

E = v x B v B + Es

g

E’ k m l

slide-4
SLIDE 4

Motional Stark Effect

PAGE 3 3/3/2013

  • MSE (Motional Stark Effect) diagnostic
  • Local, spectroscopic measurement of internal B-field (parts of it)
  • Line splitting ~ |E|

E = v x B B v + Es

g

E’ k m l

slide-5
SLIDE 5

Motional Stark Effect

PAGE 4 3/3/2013

  • MSE (Motional Stark Effect) diagnostic
  • Local, spectroscopic measurement of internal B-field (parts of it)
  • Line splitting ~ |E|
  • Line intensity π/σ ~ (1-cos2θ)/(1+cos2θ), with cosθ = k.E/|E|

E = v x B B v + Es

g

E’ k m l

slide-6
SLIDE 6

Motional Stark Effect

PAGE 5 3/3/2013

  • MSE (Motional Stark Effect) diagnostic
  • Local, spectroscopic measurement of internal B-field (parts of it)
  • Line splitting ~ |E|
  • Line intensity π/σ ~ (1-cos2θ)/(1+cos2θ), with cosθ = k.E/|E|
  • Line polarization angle tanγ=E.m/E.l

E = v x B B v + Es

g

E’ k m l ~ γ γ+π/2 ~

slide-7
SLIDE 7

Motional Stark Effect

PAGE 6 3/3/2013

  • MSE (Motional Stark Effect) diagnostic
  • Local, spectroscopic measurement of internal B-field (parts of it)
  • Conventional MSE

− Filter out π or σ line − Measure polarization angle: − Convert to q- and j-profiles

(EFIT, Ampère’s law …)

 

g B A B A B A B A B A B A

R Z R Z 5 4 3 2 1

) tan(     

E = v x B B v + Es

g

E’ k m l

slide-8
SLIDE 8

Future challenges

PAGE 7 3/3/2013

  • In the measurement…
  • Wall reflections
  • Polarization dependent mirror degradation
  • In the analysis…
  • Real-time polarization angles
  • Real-time q-profiles

[2] [1] [3]

[1] H. Yuh et al., Poster HTPD conference 2012 (not published) [2] A. Malaquias et al., EPS 2003 ECA Vol. 27A, O-3.4C [3] G. Hommen et al., Plasma Phys. Control. Fusion Vol. 55 (2013) 025007

slide-9
SLIDE 9

Wall reflections (ITER)

PAGE 8 3/3/2013

  • Background signal …
  • Line integrated bulk bremsstrahlung
  • Reflections of divertor bremsstrahlung + incandescence
  • unknown (coatings) partially polarized
  • spatial-temporally non-uniform

[1] H. Yuh et al., Poster HTPD conference 2012 (not published) A B C

slide-10
SLIDE 10

Wall reflections (ITER)

PAGE 9 3/3/2013

  • Background signal …
  • Line integrated bulk bremsstrahlung
  • Reflections of divertor bremsstrahlung + incandescence
  • Off-wavelength channels
  • in-situ calibration
  • cross-check in beam-off phases

A B C A B C [1] H. Yuh et al., Poster HTPD conference 2012 (not published)

slide-11
SLIDE 11

Mirror degradation (ITER)

PAGE 10 3/3/2013

  • First mirror (FM) exposed to plasma …
  • Imaging diagnostic, low signal  Large aperture
  • Deposition dominated location
  • Coating affects reflection and polarization!

[1] H. Yuh et al., Poster HTPD conference 2012 (not published)

aperture FM

slide-12
SLIDE 12

Mirror degradation (ITER)

PAGE 11 3/3/2013

  • First mirror (FM) exposed to plasma …
  • Coating affects reflection and polarization
  • Minimize effect by keeping incident angle close to normal

[1] H. Yuh et al., Poster HTPD conference 2012 (not published) [2] A. Malaquias et al., EPS 2003 ECA Vol. 27A, O-3.4C

slide-13
SLIDE 13

Mirror degradation (ITER)

PAGE 12 3/3/2013

  • Line Splitting calibration of polarization measurement
  • LS not affected by reflection and polarization changes
  • No line overlap in ITER
  • But: relation to B more complicated than polarization

[1] H. Yuh et al., Poster HTPD conference 2012 (not published)

slide-14
SLIDE 14

Mirror degradation (ITER)

PAGE 13 3/3/2013

  • Line Splitting calibration of polarization measurement
  • Cross-Calibration through equilibrium reconstruction
  • Simulations promising
  • but, very accurate fitting required
  • no test on yet current devices (KSTAR opportunity?)

[1] H. Yuh et al., Poster HTPD conference 2012 (not published)

slide-15
SLIDE 15

Real-time MSE

PAGE 14 3/3/2013

  • q-profile control needs real-time MSE (e.g. ITER and

KSTAR goals)

  • Real-time  1st time right
  • correct interpretation  know the spectrum
  • minimize perturbing effects  filter “best” part of spectrum
  • minimize assumptions  retrieve “all” data from signal itself
  • stable and fast analysis
  • Following slides: KSTAR as test case
  • MSE on KSTAR not real-time in 1st phase
  • but design is “real-time ready”
slide-16
SLIDE 16

Real-time MSE: spectrum simulation

PAGE 15 3/3/2013

  • Real-time  measurement 1st time right
  • Good understanding MSE spectrum

 good interpretation MSE measurement

[4] M. De Bock et al., Rev. Sci. Instrum. Vol.79 (2008) 10F524

E = v x B B v + Es

g

E’ k m l

slide-17
SLIDE 17
  • Real-time  measurement 1st time right
  • Good understanding MSE spectrum

 good interpretation MSE measurement

Real-time MSE: spectrum simulation

PAGE 16 3/3/2013

[4] M. De Bock et al., Rev. Sci. Instrum. Vol.79 (2008) 10F524

g

E’ m l E2,b vb B2 B1 B3 vc va E3,c E1,a

slide-18
SLIDE 18
  • Real-time  measurement 1st time right
  • Good understanding MSE spectrum

 good interpretation MSE measurement

Real-time MSE: spectrum simulation

PAGE 17 3/3/2013

[4] M. De Bock et al., Rev. Sci. Instrum. Vol.79 (2008) 10F524

g

E’ m l E2,b vb B2 B1 B3 vc va E3,c E1,a

slide-19
SLIDE 19
  • Real-time  measurement 1st time right
  • Good understanding MSE spectrum

 good interpretation MSE measurement

Real-time MSE: spectrum simulation

PAGE 18 3/3/2013

[4] M. De Bock et al., Rev. Sci. Instrum. Vol.79 (2008) 10F524 [9] J. Ko et al., KSTAR conference (2013) poster 25

slide-20
SLIDE 20
  • Real-time  measurement 1st time right
  • Good understanding MSE spectrum

 good interpretation MSE measurement

Real-time MSE: spectrum simulation

PAGE 19 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524

beam 2 beam 1 beam 3

slide-21
SLIDE 21
  • Real-time  measurement 1st time right
  • Good understanding MSE spectrum

 good interpretation MSE measurement

Real-time MSE: spectrum simulation

PAGE 20 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524 [9] J. Ko et al., KSTAR conference (2013) poster 25

slide-22
SLIDE 22
  • Real-time  measurement 1st time right
  • Good understanding MSE spectrum

 good interpretation MSE measurement

Real-time MSE: spectrum simulation

PAGE 21 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524

beam 2 beam 1 beam 3

slide-23
SLIDE 23
  • Optical filter width and shape
  • “smoothing” can reduce δγ, but too much increases it!
  • 2 cavity, 4A wide interference filter for KSTAR

Real-time MSE: Filter design

PAGE 22 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524 δγ [degrees]

slide-24
SLIDE 24
  • Optical filter central wavelength position
  • Fine tuning by tilting.
  • Can be done real-time to ensure optimal signal

Real-time MSE: Filter design

PAGE 23 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524

slide-25
SLIDE 25
  • PEM polarization measurement
  • polarization  intensity modulation
  • base frequencies (ω1 and ω2)
  • base retardances (R1 and R2)
  • Real-time MSE: Extracting all information

PAGE 24 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524 PEM1 harmonics PEM2 harmonics

tan(2g) = J2(R2) I2w1 J2(R

1) I2w2

E2,b vb B2 B1 B3 vc va E3,c E1,a

slide-26
SLIDE 26
  • PEM polarization measurement
  • ω and Iω  extracted with PLL
  • Retardance R ???

 1 set value on PEM controller …. but profile on PEM itself!  average over PEM aperture

Real-time MSE: Extracting all information

PAGE 25 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524

tan(2g) = J2(R2) I2w1 J2(R

1) I2w2

E2,b vb B2 B1 B3 vc va E3,c E1,a

x (cm) y (cm)

slide-27
SLIDE 27

tan(2g) = J2(R2) I2w1 J2(R

1) I2w2

  • PEM polarization measurement
  • Averaged retardance R from signal:

− measure up to 4th harmonic

Real-time MSE: Extracting all information

PAGE 26 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524

E2,b vb B2 B1 B3 vc va E3,c E1,a

I2w1 I4w1 = - J2(R

1)

J4(R

1)

slide-28
SLIDE 28
  • Simple analysis:
  • 1 Phase Locked Loop per channel
  • Requires enough signal

Real-time MSE: Synergy in analysis

PAGE 27 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524 PEM 1 PEM 2

slide-29
SLIDE 29
  • Simple analysis: 1 Phase Locked Loop per channel
  • Multi-channel analysis:
  • All channels go through same PEMs
  • Phase shift φ equal for all channels

 use synergy of multi-channel analysis  more signal, more accurate locking, also applies to R

Real-time MSE: Synergy in analysis

PAGE 28 3/3/2013

[5] M. De Bock et al., Rev. Sci. Instrum. Vol.83 (2012) 10D524 PEM 1 PEM 2

slide-30
SLIDE 30
  • Real-time constraint in real-time EFIT
  • fast conversion

− coarse grid − low order P’ & FF’ basis functions

  • especially good for boundary
  • internal profiles more difficult

Real-time q-profile

PAGE 29 3/3/2013

[6] D.A. Gates et al., Nucl. Fusion 46 (2006) 17–23 [7] L.L. Lao et al., Fus. Sci. Tech 48 (2005) 968–977

slide-31
SLIDE 31
  • Real-time constraint in real-time EFIT
  • Direct from MSE & shaping parameters
  • Boundary  ε, κ, δ

− magnetic reconstruction (rtEFIT) − optical reconstruction (OFIT)

  • MSE

 Rmagn. axis − where γ=0

  • Theory

 ε(ρ), κ(ρ), δ(ρ) − flux surface shape

Real-time q-profile

PAGE 30 3/3/2013

[8] R. Gianella et al., Rev. Sci. Instrum. 75 (2004) 4247 [3] G. Hommen et al., Plasma Phys. Control. Fusion Vol. 55 (2013) 025007

slide-32
SLIDE 32
  • Real-time constraint in real-time EFIT
  • Direct from MSE & shaping parameters
  • Boundary  ε, κ, δ
  • MSE

 Rmagn. axis

  • Theory

 ε(ρ), κ(ρ), δ(ρ)

  • KSTAR example:

− Optical reconstruction (OFIT) works with current camera images

Real-time q-profile

PAGE 31 3/3/2013

[3] G. Hommen et al., Plasma Phys. Control. Fusion Vol. 55 (2013) 025007

slide-33
SLIDE 33
  • Real-time constraint in real-time EFIT
  • Direct from MSE & shaping parameters
  • Boundary  ε, κ, δ
  • MSE

 Rmagn. axis

  • Theory

 ε(r), κ(r), δ(r)

  • MSE

− Assigns flux to surfaces − 2D flux map, q-profile, j-profile

  • computing time/frame: 1ms

Real-time q-profile

PAGE 32 3/3/2013

[3] G. Hommen et al., Plasma Phys. Control. Fusion Vol. 55 (2013) 025007

slide-34
SLIDE 34

Conclusions

PAGE 33 3/3/2013

  • MSE = powerful, local measurement of B-components
  • Use complete spectrum (not just 1 line)
  • Background reflections

− off-wavelength polarization measurement

  • Mirror degradation

− Line splitting measurment

  • Real-time requirements
  • real-time MSE needed for q-profile control
  • Know your spectrum
  • Use all signal information & multi-channel synergies

Thank you!