Polydiagnostics on the COST lamp

1
Polydiagnostics on the COST lamp
Joost van der Mullen Technische Universiteit
Eindhoven
Aim To calibrate various methods against each
others find the truth and nothing but the
truth For validating models
Madeira Model Inventory Workshop April 12-16
2005
2
The problems the challenges
Three methods to measure the temperature
Usually give three (or even more) answers
Even for so-called LTE plasma differences up
to 30
Difficult to answer the question is LTE
present? Or are the method (in)correct?
Impossible to validate the models
3
Final Goal
To find Easy/Global observables
To Characterise the plasma ne, Te etc
To determine the state of the lamp Light
technical properties Remaining Lifetime
Candidate Easy/Global observables (Filtered)
Emission Electronic behavior
4
Polydiagnostics on good defined plasma
Different people Various techniques
Limited amount of Lamps
5
The COST reference LampFamily
First generation Shortcomings burner wall too
hot ? limitation in life span power.
Second generation (in preparation) 1) Outer
envelope is filled with N_2 Additional
convection cooling   2) electrodes simplified (no
spirals but rods) Request on last COST meeting
plasma-electrode interactions.  
Invitation to work on 1ste generation requests
for the final design.   The followings types 1
ste generation are available a) Pure Hg
b) Hg with Na  c) Hg with Dy.
6
Provided from EINLighTRED
Eindhoven
Philips Aachen
INstitute for
Philips Licht
Philips NatLab
Lighting
Technology
Tue/N
ASML
Draka
Research and
EDucation
7
The plasmas in MH lamps
10 bar Hg 10 mbar add
Color non-uniformity
8
3D
Plasma orientation dependent
Gravitation
2-D
9
Main characteristics
Chemistry 10 bar Hg 20 mbar DyI3
Main feature Majority plasma properties Minor
ity species
10
General exploration phenomena
Demands High Efficiency Radiation Long life
span
11
Elemental whats in that name ??
Species H2O OH O H H etc.
H atomic concentration H atoms per volume
H elemental concentration all H atoms per
volume irrespective binding/state
H H 2H2O H
12
Radial segregation Diffusion
Wall T low Molecules Large slow
Centrum T high atoms small fast
Nelea va Nelem vm
13
Axial segregation
Large Nele at wall pushed down Small
Nele at centr pushed up
Net effect Emiting species Pushed-down
Or differently
Quick atoms can leave upstream easily Slow
molecules stay streaming downwards
14
Competition
Convection Diffusion Chemistry Radiation
15
Methods
Grand Modeling
Polydiagnostics
Radiation Transport
Thomson Sc
3D
16
Convergence
The Truth
17
The Final Goal
Can we by just looking to easy/global
observables Characterise the plasma ne Te,
etc Assess the status of the lamp
Efficiency Remaining lifetime
Examples Easy/Global observables
Emission/Filters V/I response
18
Settings
Model Diagnositcs validation for Several
settings
Gravitation Zero g Extreme g
Power Quantity Waveform
Vessel Geometry
19
Zero g methods limited
20
Normal conditionsmany possibilities
Guided by Atomic State Distribution Function
(ASDF)
21
The ASDF
                                       
 
22
Emission Spectroscopy

• Intensity as function of Wavelength
• Wavelength calibration big effort
• Calibrate Intensity of lines ALI
• Calculate Density of Dysprosium
• Atomic State Distribution Function (ASDF)
• Gives T gives various ns

23
Setup ES
dia
lamp
lens
ST-2000 CCD
ST-6 CCD
24
Spectral Impression grass fieldLine
identification not trivial
25
Radial profile Dy1
26
Abel inversion
?(r) emission as a function of radius r I(x)
measured lateral emission-line intensity
27
ALI
28
ASDF for central position
DyII
DyI
T5524 K
29
Future Plans

• Join forces with Plasimo
• Compare results with that of other techniques
• Still much work Spectrum identification Mea
  surements

ALI will be The reference frame For other
thechniques COST cooperation
Important Part of the collection Easy/Global
Observable
30
Absorption Spectroscopy
Dy groundstate density
Charlotte Groothuis
31
Linking absorption with density
32
Lateral
33
(No Transcript)
34
Lateral
35
Segregation parameter ? ?p /?z
36
X-ray absorption
Xiaoyan Zhu Evert Ridderhof
37
Procedure
(n T)any pos (n T)wall
Tg on any position
38
XRA on Helios lamp

• Exposure time 200s.

on off
258
464
788 852
1012
39
(No Transcript)
40
The Wall temperature as a function axial
position z
41
The Radial T profile as a F(z)
New
42
The Shape as a F(z)
43
The shape as a F(power)
44
The Radial T profile as a F(z)
T of 5000 K In Hg part Are low
45
X-ray Induced Fluorescence measurement of
segregation in MH lamps

• Tanya Nimalasuriya (TU/e)
• Evert Ridderhof (TU/e)
• John J. Curry (NIST)
• Craig J. Sansonetti (NIST)
• Sharvjit Shastri (APS)

46
Introduction
47
Advanced Photon Source
48
Experiment station
E.J. Ridderhof
49
XRF sketch
The x-ray beam is produced by the Sector 1
Insertion Device beam line at the Advanced Photon
Source at the Argonne National Laboratory
J.J.Curry NIST
50
XRF basic principle
An electron in the K shell is ejected from the
atom by an external primary excitation x-ray,
creating a vacancy.
An electron from the L or M shell "jumps in" to
fill the vacancy. In the process, it emits a
characteristic x-ray unique to this element and
in turn, produces a vacancy in the L or M shell.
51
XRF spectral lines
Principal fluorescence lines produced by K-shell
excitation in Dy. The excited levels correspond
to a singly ionized atom
X-ray induced fluorescence spectrum excited by 70
keV photons at x/R 0.56. x is displacement from
the arc axis in the direction of the detector and
R4.5 mm is the arc tube radius
J.J.Curry NIST
52
XRF advantages

• X-ray induced fluorescence
• - determines elemental densities of
  Dy,Hg
• - is effective anywhere in the burner
• No inversion technique is needed
• T profile with Hg densities

53
XRF-Spectra
1mm above bottom electrode x, z center
1mm above bottom electrode z center, x at wall
I
W
Ce
Dy
I
Dy
Ce
54
Dy density profile

6.7
21
36
50
)
-3

Density (cm
normalised radial position
55
Ratio elemental densities Dy/ Hg Relative
concentration
T. Nimalasuriya, J.J. Curry, C.J. Sansonetti,
E.J. Ridderhof
56
Diffusion versus (radial) convection
57
Temperature profile from Hg density
TXRF lt TXRA !!
XRF 140 W
XRA 142 W, X.Y. Zhu
58
XRF Conclusions

• Axial and radial segregation clearly observed

• T profile XRF shows similarities with XRA TXRF
  lower !!

• In future compare results with Absolute Line
  Intensity measurements and Laser Absorption
  Spectroscopy
• Polydiagnostics

59
Conclusions TS (versus XRA)

• TS for the first time applied on real lamp

• Indications that the LTE assumption is not
  valid
• Thermal Te - T gas ? 2000 K XRA compared
• Chemical Texc ? Te, b1 gt10.
• Plasma always ionizing even at I-zero-crossing!!
  

60
Zero g
61
Parabolic flights

• 20 seconds 1.8 G
• 25 seconds 0 G
• 20 seconds 1.8 G

62
 
63
(No Transcript)
64
Setup in zero g plane
65
Setup PFC
66
Parabolic Flights
67
(No Transcript)
68
Fisher model

• Competition between Diffusion and Convection.

69
Prediction forincreased convection

• Left hand side
• Decrease diffusion
• Increase convection
• More demixing

• Right hand side
• Decrease diffusion
• Increase convection
• Better mixing

70
The Convection/Diffusion competition
0g Diffusion solely 1gOptimal competition 2g
Convection dominant
71
Results for absorption
72
Sphere of Ullbricht integrated intensityJob
Beckers/Winfred Stoffels
 

• Highly reflective diffusive
• coating
• Integrates all light
• Homogenious light
• the whole sphere

73
ARGES burner, DyI3-salt, 5 mg Hg

1. Output increases cause of axial de-segregation
2. (right on the Fischer curve)
3. Output increases cause of disappearence of axial
   segregation (totally left on the Fischer
   curve) and new equilibrium.
4. Equilibrium comes back

74
Conclusions

• The total Light output varies with gravity.
• Difference of the light output can be explained
  by the theory of axial- lt-gt radial segregation
  of lamps at the right side Fischer curve.
• Lamps do not reach equilibrium at the end of
  a zero-g phase.
• The results inegrated emission agree with
  absorption

75
International Space Station
76
ISS data analysis

• Dy 642 and Hg 579
• lateral profile
• abel inversion
• T profile using absolute measurement of Hg
• density profile of Dy 642

77
Analysis Dy 642
78
Concluding Remarks
Polydiagnostics is an enormous field
Preliminary work on active and passive
spectr has been done
ALI the best platform for mutual calibration Line
identification not trivial A-values needed
Strong indications LTE not present under
high pressure conditions
There is need for much more COST projects