r12.7 report on solar blind and active pyrometry system...

19
1 Grant Agreement No. 228296 SFERA Solar Facilities for the European Research Area SEVENTH FRAMEWORK PROGRAMME Capacities Specific Programme Research Infrastructures Integrating Activity - Combination of Collaborative Project and Coordination and Support Action R12.7 Report on solar blind and active pyrometry system Due date of deliverable: Month 26 Actual submission date: Month 52 Organisation name of lead contractor for this deliverable: CNRS

Upload: lamdieu

Post on 14-Mar-2018

215 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

1  

 

 

 

Grant Agreement No. 228296

 

 SFERA  

Solar Facilities for the European Research Area

 

SEVENTH FRAMEWORK PROGRAMME

Capacities Specific Programme

Research Infrastructures

Integrating Activity - Combination of Collaborative

Project and Coordination and Support Action

 

 

R12.7  Report  on  solar  blind  and  active  pyrometry  system  

Due  date  of  deliverable:  Month  26  

Actual  submission  date:  Month  52  

 

Organisation  name  of  lead  contractor  for  this  deliverable:  CNRS  

Page 2: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

2  

 

 

Contents  

1.   Introduction  .....................................................................................................................................  3  

2.   Optical  temperature  measurements  ...............................................................................................  3  

Description  of  the  method  ..................................................................................................................  3  

Instruments  for  optical  temperature  measurements  .........................................................................  3  

Techniques  to  calculate  the  body  temperature  ..................................................................................  5  

Specifications  for  equipment  ..............................................................................................................  6  

Calibration  issues  ................................................................................................................................  6  

Main  error  sources  ..............................................................................................................................  7  

Literature  ............................................................................................................................................  8  

3.   A  CSP  specific  problem:  Solar  blind  pyrometry  ...............................................................................  9  

Presentation  of  the  problem  ...............................................................................................................  9  

First  solution:  solar  blind  pyrometry  —  atmospheric  or  glass  absorption  ..........................................  9  

Literature  ..........................................................................................................................................  12  

4.   The  emissivity  problem,  a  possible  solution:  bi-­‐Color  pyro-­‐reflectometry  ....................................  13  

Presentation  of  the  bi-­‐color  pyroreflectometry  ................................................................................  13  

Method  .............................................................................................................................................  13  

Literature  ..........................................................................................................................................  15  

5.   An  alternative  to  solar  blind  measurements:  active  pyrometry  ....................................................  16  

Pulsed  thermography  ........................................................................................................................  17  

Phase  lock-­‐in  active  laser  pyrometry  ................................................................................................  17  

Evolution:  temperature  measurements  in  solar  simulators  and  solar  furnaces  ...............................  17  

Literature  ..........................................................................................................................................  19  

 

   

Page 3: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

3  

 

1. Introduction  This  deliverable  presents  possible  methods  in  order  to  measure  the  temperature  of  materials  under  concentrated  solar  irradiation  such  at  the  focus  of  solar  facilities.  An  initial    

This  first  chapter  is  derived  from  the  chapter  included  in  SFERA  deliverable  R12.4  Guidelines  for  Testing  of  CSP  components,  but  with  additional  precisions.  Reading  other  chapters  from  R12.4  such  as  temperature  measurements  with  probes  should  be  of  interest  to  the  reader.  

2. Optical  temperature  measurements  

Description  of  the  method  The  temperature  of  materials  can  be  determined  by  measuring  the  amount  of  radiative  heat  emitted,  typically  in  the  infrared  spectrum.  

The  Planck  Law  allows  to  calculate  the  power  hemispherically  radiated  by  a  local  point  of  a  black  body  at  a  given  temperature  for  any  wavelength.    

𝐿! =!"!!

!

!!∙ !

!"#!!!!"# !!

      (1)  

Lλ  Angular  spectral  luminance  (W·m-­‐2·sr-­‐1·m-­‐1)  λ  Considered  light  wavelength  T  Surface  temperature  of  the  body  cλ  Speed  of  light  in  the  considered  medium  (m/s)  h  =  6,626  17×10-­‐34  J.s  Planck  constant  k  =  1,380  66×10-­‐23  J/K  Boltzman  constant  

Conversely,  by  measuring  the  light  power  over  the  complete  hemisphere  for  a  given  wavelength,  one  can  therefore  determine  the  temperature  of  the  black  body.  

However,  actual  materials  are  not  black  bodies:  part  of  this  theoretically  calculated  power  is  not  emitted.  The  emissivity  coefficient  compensates  for  this  reduced  emission.  The  emissivity  coefficient  depends  on  the  considered  wavelength,  the  temperature  of  the  equipment,  its  composition,  its  surface  aspect  (polished,  grinded…).  Therefore,  in  order  to  determine  the  temperature  with  optical  instruments,  one  must:  

o Measure  the  luminance  emitted  from  the  sample  

o Determine  the  emissivity  of  the  sample  for  the  considered  conditions:  measuring  wavelength,  body  temperature,  surface  aspect,  view  factor  or  etendue.  

Instruments  for  optical  temperature  measurements  Two  families  of  instruments  can  be  referenced:  

o Spot  instruments  typically  called  pyrometers.  

Page 4: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

4  

 

o Mapping  instruments,  usually  called  infrared  cameras.  

 Typical  sensitivity  of  different  detectors  technology  depending  on  their  temperature  and  the  incoming  

wavelength,  from  [Legrand  2002]  

The  technologies  used  for  both  instruments  are  similar,  except  for  their  organization  in  the  device  (single  detector  for  a  pyrometer,  matrix  or  scanner  for  cameras):  

o Thermal  detectors:  incoming  radiation  is  used  to  heat  the  detector  which  then  provide  a  signal.  The  wavelength  sensibility  depends  on  the  coating  used  on  the  detector.  Different  technologies  have  been  investigated  for  optical  temperature  measurements:  

o Bolometers  are  RTD  sensors:  electric  resistance  variation  due  to  the  temperature  of  the  sensor.  Such  sensors  are  used  in  pyrometers  or  in  cameras  as  described  in  [WOO  1993].  

o Pyroelectric  sensors:  the  sensor  delivers  an  electrical  charge  depending  on  the  received  electromagnetic  radiation.  The  sensor  composition  is  a  ferroelectric  material.  These  sensors,  despite  having  lower  sensitivity,  are  interesting  as  they  are  relatively  fast  and  have  wide  wavelength  sensitivity,  even  without  cooling  of  the  sensor.  

Page 5: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

5  

 

However,  the  sensor  is  sensitive  only  to  variation  of  the  incoming  electromagnetic  flux,  so  absolute  measurements  are  not  common  with  this  technology.  

o Thermopile  sensor:  the  heated  element  is  a  thermocouple  or  a  serie  of  thermocouples.  Therefore,  the  incoming  radiation  leads  to  the  measure  of  a  voltage  which  depends  on  this  incoming  heat  and  a  reference  temperature  (cold  junction).  

o Quantic  detectors:  the  detector  delivers  a  signal  proportional  to  the  number  of  received  photons.  Different  signal  can  be  delivered:  

o Photovoltaic  sensors:  the  incoming  photons  beam  leads  to  a  flow  of  electrical  current.  This  is  the  technology  used  for  everyday  cameras:  the  technology  is  typically  cheap,  extremely  fast,  but  usually  for  short  wavelengths  (visible  or  near  infrared)  thus  high  temperature.  

o Photoconductivity:  as  the  detector  sensitive  material  absorbs  incoming  photons,  its  conductivity  changes.  This  is  the  principle  used  in  PbS,  PbSe  or  InSb  detectors.  

Every  technology  has  different  characteristic  of  cost,  speed,  wavelength  sensibility,  signal  to  noise  ratio,  and  R&D  efforts  are  ongoing  to  improve  all  these  aspects.  

Techniques  to  calculate  the  body  temperature  The  signal  from  the  instruments,  whatever  their  technology,  are  used  in  different  ways  in  order  to  calculate  the  temperature  of  the  considered  body.  Beyond  the  problem  to  get  a  trustable  signal  related  to  the  luminance  of  the  body  in  the  chosen  wavelength  band,  the  problem  of  the  lack  of  real  knowledge  of  the  emissivity  property  of  the  body  has  lead  to  the  development  of  different  techniques:  

o Pyrometers  are  used  to  determine  the  average  temperature  of  the  sensed  surface  and  infrared  cameras  to  determine  local  temperature  maps.  The  user  has  to  provide  the  suitable  emissivity  from  reference  books  or  dedicated  measurements.  

o Bi-­‐  or  tri-­‐color  pyrometers  can  be  used  without  knowing  the  emissivity.  These  devices  are  pyrometers  operating  at  2  or  3  wavelengths,  and  under  the  assumption  that  the  emissivity  of  the  material  is  the  same  at  all  these  wavelengths,  the  redundant  measurements  allow  to  determine  the  temperature.  However,  this  assumption  is  not  always  true  despite  the  usual  precautions  (nearby  wavelengths).  The  tri-­‐color  pyrometers  are  used  to  at  least  check  this  assumption,  where  bi-­‐color  pyrometers  would  always  provide  a  result  without  hinting  if  it  is  realistic  or  not.  

o Bi-­‐color  pyro-­‐reflectometer  are  bi-­‐color  pyrometers  that  additionally  measure  the  directional  reflectivity  of  the  material  at  both  wavelengths.  If  the  material  is  opaque  and  the  bidirectional  reflectance  distributions  are  proportional  between  both  wavelengths  (whatever  the  incoming  and  outcoming  radiation  direction,  the  ratio  is  constant  between  the  reflectivity  taken  at  each  wavelengths),  the  system  of  equation  can  be  solved  to  determine  the  temperature  without  knowing  the  hemispherical  emissivity.  These  assumptions,  despite  

Page 6: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

6  

 

being  weaker  than  for  bi-­‐  or  tri-­‐color  pyrometry,  are  still  not  always  validated,  for  example  in  case  of  coated  materials.  

Development  is  currently  under  way  engineering  cameras  based  on  bi-­‐color  pyrometry  and  on  pyro-­‐reflectometer  principles  to  determine  temperature  maps  with  fewer  assumptions  on  the  emissivity  of  the  materials  for  more  robust  instruments  than  the  current  common  single  wavelength  or  bandwidth  infrared  cameras.  

Specifications  for  equipment  To  choose  a  pyrometer:  

-­‐ Define  your  temperature  range.  Define  both  the  minimal  and  maximal  expected  temperature.  Few  instruments  are  able  to  cover  both  very  high  temperature  (>  2000K)  and  near  ambient  temperatures  (<  500K).  

-­‐ Define  your  geometry.  Define  both  the  desired  probed  spot  size  and  the  distance  from  which  you  will  install  the  instrument.  Be  aware  that  low  cost  pyrometers  (less  than  a  few  thousand  euros/dollars)  typically  have  a  low  quality  alignment  of  the  provided  aiming  lasers  and  the  actual  position  of  the  probed  spot.  You  can  check  this  by  moving  a  diaphragm  in  the  beam  while  probing  a  large  hot  surface  such  as  a  radiative  plate.  The  source  to  instrument  beam  characteristic  is  the  etendue.  

-­‐ Define  your  wavelength(s).  Depending  on  both  your  material  properties,  the  expected  temperature  and  your  environment  condition,  near  or  medium  infrared  are  best  suited.  Generally,  the  lower  the  temperature,  the  longer  the  wavelength  due  to  the  Planck  law:  the  body  will  radiate  more  energy  hence  a  higher  signal.  The  Wien’s  displacement  law  can  be  used  as  a  hint,  as  it  gives  the  peak  wavelength  of  the  thermal  emission  of  a  perfect  black  body  depending  on  its  temperature:  

𝜆!"#   ∙  𝑇 = 2898  µμ𝑚.𝐾  

𝜆!"#    peak  wavelength  of  the  black  body  thermal  emission  in  µm    

T  temperature  of  the  black  body  in  K  

If  you  plan  front  side  measurements,  that  is  from  the  side  irradiated  by  the  concentrated  solar  energy,  choose  solar  blind  wavelength  without  reflected  light.  Typical  solar  blind  wavelengths  are  2.7  µm  (about  100  nm  wide,  due  to  water  in  atmosphere)  or  4.3  µm  (similarly  narrow  band,  due  to  CO2  in  atmosphere)  and  5  to  7  µm  (due  to  water  and  at  the  start  to  mirrors’  glass  absorption).  

Calibration  issues  While  calibrating  your  pyrometer  with  a  black  body  or  checking  in  it  with  a  radiative  plate:  

-­‐ Always  include  the  windows  and  filters  that  will  be  in  your  experimental  setup.  

Page 7: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

7  

 

-­‐ Reproduce  as  best  as  possible  the  geometry  from  your  experiment  such  as  distances,  angles  between  the  probed  body  and  your  instrument.  

-­‐ Atmospheric  conditions:  especially  for  solar  blind  pyrometry,  check  accordingly  the  ground  H2O  or  CO2  levels  during  calibration  to  have  them  similar  to  those  prevailing  during  your  experiment.  

Main  error  sources  -­‐ Emissivity  

The  emissivity  is  the  ratio  of  light  emitted  by  a  body  compared  to  the  ideal  black  body.  Ideally,  the  emissivity  of  your  material  should  be  measured.  Tables  from  bibliography  only  give  a  possible  range  of  the  parameter,  surface  properties  can  make  it  change  up  to  50%,  hence  the  determined  temperature  by  similar  ratio.  Emissivity  depends  on:  

o Chemical  composition  of  the  material.  

o Physical  state,  including  the  crystal  structure.  

o Surface  roughness:  both  low  and  high  frequency  shape  compared  to  the  pyrometer  wavelength  have  an  effect  such  as  diffraction  in  the  grooves.  

o Temperature  of  the  body.  

o Coatings.  Low  thickness  coating  (a  few  microns)  can  be  transparent  or  semi-­‐transparent  at  your  wavelength.  You  may  measure  emitted  light  from  the  coating,  the  body,  or  both;  and  there  can  be  a  thermal  gradient  between  the  body’s  surface  and  the  coating’s  surface,  leading  to  a  complex  mix  of  signals  measured  by  the  pyrometer.  

-­‐ Solar-­‐blindness  If  your  pyrometer  is  not  completely  solar  blind  (e.g.  using  an  industrial  pyrometer  for  the  glass  industry  at  4.7  –  5  µm),  residual  reflected  concentrated  light  will  be  added  to  the  light  emitted  by  your  material.  If  the  flux  concentration  and  the  material  surface  properties  do  not  change,  this  can  be  dealt  with  as  a  systematic  error  and  corrected,  otherwise  the  temperature  will  be  overestimated.    

-­‐ Dirt  As  any  optical  measurements,  any  dirt  on  the  optics  will  change  the  results.  As  it  is  an  infrared  measure,  there  can  be  dirt  not  visible  by  the  human  eyes,  such  as  residues  from  unsuitable  cleaning  liquids  (soap…).  Alcohol  based  liquid  are  advised,  preferably  iso-­‐propanol,  but  check  the  suitability  with  your  windows  materials  (some  infrared  windows  could  be  dissolved)  and  your  working  wavelengths  (no  residues).  

-­‐ Noise  Parasite  reflection  can  impede  the  optical  measurement,  notably  for  low  temperature  bodies.  Electric  noise  should  be  reduced  as  always  with  proper  grounding  and  shielding.  

Page 8: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

8  

 

-­‐ Multi  color  pyrometry  Depending  on  the  distance  between  the  several  wavelengths  used  by  the  pyrometer,  all  the  above  cited  errors  may  be  similar  if  the  wavelength  are  near  (no  more  than  a  few  hundred  nanometers)  or  the  errors  may  be  different  if  the  wavelength  are  far  from  each  other  (several  microns).  In  the  latter  case,  the  errors  should  be  corrected  or  at  least  check  individually  for  each  wavelength  used.  

Literature  D.  Hernandez,  G.  Olalde,  A.  Beck  and  E.  Milcent  (1995):  Bicolor  pyroreflectometer  using  an  optical  fiber  probe  ;  Rev.  Sci.  Instrum.,  66,  5548.  

P.B.  Coates  (1977):  Wavelength  Specification  In  Optical  and  Photoelectric  Py-­‐  rometry;  Metrologia,  13,  1.  

P.B.  Coates  (1981):  Multi-­‐Wavelength  Pyrometry;  Metrologia,  17,  103.  

M.  Schubnell,  H.R.  Tschudi  and  Ch.  Muller;  Temperature  measurement  under  concentrated  radiation;  Sol.  Energy,  58,  69,  (1996)    

SFERA  Deliverable  R12.4:  Guidelines  for  Testing  of  CSP  components  

Etendue:  http://en.wikipedia.org/wiki/Etendue  

Woo  R.A.,  Foss  R.A.  (1993):  Micromachined  bolometer  array  achieves  low  cost  imaging,  Laser  Focus  world,  vol  6,  pp101-­‐106.  

Legrand  A.C.  (2002):  Thermographie  multispectrale  haute  et  basse  temperature,  Application  au  contrôle  non  destructif,  PhD  thesis,  Univ.  Bourgogne.  

   

Page 9: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

9  

 

 

3. A  CSP  specific  problem:  Solar  blind  pyrometry  

Presentation  of  the  problem  If  the  equipment  for  which  we  want  to  know  the  temperature  is  irradiated  by  concentrated  solar  energy,  the  aforementioned  instrument  will  measure  both  the  self-­‐emitted  light  from  the  equipment  and  reflected  concentrated  solar  light.  This  is  true  even  for  apparently  non-­‐reflective  materials  or  cavities  (except  if  completely  designed  as  a  black  body).  If  the  intensity  of  the  self-­‐emitted  light  is  relatively  low  even  the  poorly  reflected  intense  incoming  concentrating  light  can  dominate  it.  

 

The  bright  spot  at  the  focus  of  the  concentrating  solar  facility  comes  both  from  reflection  of  the  incoming  strong  light  and  from  the  thermal  emission  of  the  heated  material:  different  methods  must  be  used  in  order  to  sort  the  signals,  such  as  choosing  wavelengths  with  only  on  of  the  signal  (solar  blind  pyrometry)  or  adding  a  controlled  heating  modulation  (active  pyrometry).  

First  solution:  solar  blind  pyrometry  —  atmospheric  or  glass  absorption  

For  such  conditions,  the  wavelength  for  the  measurements  must  be  chosen  for  which  there  is  no  or  little  concentrating  solar  light.  This  is  the  case  either  in  Earth’s  or  Sun’s  atmosphere  absorption  bands,  or  on  the  concentrating  optical  system  absorption  bands  as  illustrated  in  the  figure  below.  

Page 10: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

10  

 

Such  measurement  are  called  solar  blind  as  they  don’t  see  solar  radiation  per  design.  Refer  to  [Hernandez  2004]  to  the  investigation  to  multiple  wavelengths  suitable  of  high  concentration  solar  furnaces  where  the  problem  is  exacerbated.  

 Solar-­‐blind  temperature  measurement  developments  for  tower  applications  [Ballestrin  2011]:  Solar  spectrum  based  on  a  MODTRAN  simulation,  solar  reflected  spectrum,  two  band-­‐pass  filters,  black-­‐  body  radiance  at  several  temperatures,  mirror  reflectance,  and  quartz  transmittance.  

Different  wavelengths  are  usually  considered  for  solar  blind  measurements  thanks  to  atmospheric  absorption  bands:  

o From  about  2.5  to  2.8  µm  (typically  called  the  2.7  µm  band):  atmospheric  absorption  from  CO2  and  H2O.  

o From  about  4.2  to  4.4  µm  (typically  called  the  4.3  µm  band):  atmospheric  absorption  due  to  CO2.  

o From  5.5  to  7.3  µm:  atmospheric  absorption  due  to  H2O.  

o Beyond  14.2  µm:  atmospheric  absorption  from  CO2  and  H2O.  

open with direct solar irradiance of 1000 W m-2, peak irradiance at the focus [12] is 3034 kW m-2, total power is 69 kW and the 90 % power focal diameter is 26.2 cm.

3. Solar-Blind IR camera prototype

A new IR camera based on an InSb detector has been designed. This detector works in the 1.5-5µm spectral range and a software-controlled filter wheel enables the centered band-pass filters (Table 1) to be used on these two wavelengths bands (Fig. 2) creating a solar-blind IR camera.

Filter properties Filter 1 Filter 2

CW, Central wavelength (nm) 3320 4720

HW, Full width at half maximum (nm) 60 90

Table 1. IR camera band-pass filters.

Figure 2 shows the solar spectrum based on a MODTRAN code simulation [13], the solar reflected spectrum, two band-pass filters, black-body radiance at several temperatures, mirror reflectance, and quartz transmittance. Once the solar radiation is reflected by the heliostat mirror and concentrator, some solar radiation from the sample is reflected onto the IR camera (Fig. 1). This IR camera has two special band-pass filters, which avoid or minimize the reflected solar radiation from the sample. The low reflectivity of the mirrors over 3000 nm [14] allows wavelength bands to be defined where the solar radiation is almost negligible (Fig. 2).

Fig. 2. Solar spectrum based on a MODTRAN simulation, solar reflected spectrum, two band-pass filters, black-body radiance at several temperatures, mirror reflectance, and quartz transmittance.

Page 11: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

11  

 

 

Atmospheric  absorption  bands  in  the  infrared  spectrum  

 

These  absorptions  wavelengths  depends  obviously  on  the  content  level  of  the  considered  gas  in  atmosphere  and  the  length  of  the  light  path  through  this  atmosphere:  for  example,  to  measure  the  temperature  of  receivers  installed  at  the  top  of  solar  towers,  the  instrument  is  usually  installed  hundred  of  meters  away.  For  such  applications,  the  H2O  bands  are  not  advised  as  the  amount  of  water  in  the  atmosphere  changes  a  lot.  Corrections  are  possible  as  it  is  relatively  easy  to  measure  the  ground  humidity,  but  it  leads  to  greater  uncertainties  of  the  results.  The  CO2  bands  are  better  for  such  applications,  as  the  CO2  level  is  pretty  stable  in  open  space  conditions.  

When  considering  the  light  path  from  the  Sun  to  the  solar  receiver,  an  other  component  can  absorb  part  of  the  solar  energy  radiated  by  the  sun:  the  mirrors.  Indeed,  even  if  the  typical  total  thickness  of  the  glass  on  the  light  path  is  just  a  few  millimeters,  this  adds  at  least  two  possible  windows  for  solar  blind  measurements:  

o UV  measurements  

o 4.7  to  8  µm  measurements,  with  confirmation  depending  on  the  actual  glass  used.    

Page 12: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

12  

 

 Typical  glass  optical  properties,  from  the  users  manual  of  the  Heitronics  KT15  pyrometer  

When  using  such  absorption  bands,  the  change  of  the  optical  components  should  be  monitored  carefully.  For  example,  when  replacing  broken  mirrors  with  new  ones,  their  thickness  and  their  chemical  composition  may  be  different,  leading  to  a  different  absorption  of  the  solar  radiation  and  errors  on  the  temperature  measurements.  

Literature  D.  Hernandez,  G.  Olalde,  J.M.  Gineste,  C.  Gueymard  (2004):  Analysis  and  experimental  results  of  solar  blind  measurements  in  solar  furnaces;  Journal  of  Solar  Energy  Engineering,  Vol.  126,  pp  645-­‐653.  

J.  Ballestrína,  A.  Marzoa,  J.  Rodrígueza,  I.  Cañadasa,  F.  J.  Barberob  (2011):  Two  wavelength  bands  for  IR  thermometry,  SolarPaces  2011.  

 

   

Page 13: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

13  

 

4. The  emissivity  problem,  a  possible  solution:  bi-­‐Color  pyro-­‐reflectometry  

Presentation  of  the  bi-­‐color  pyroreflectometry  The  pyroreflectometry  is  a  punctual  technic  which  allow  to  measure  simultaneously  the  true  temperature  and  the  emissivity  of  an  opaque  surface.  The  method  has  been  developed  at  the  CNRS-­‐PROMES  at  two  wavelengths.  The  principle  of  the  pyroreflectometry  is  to  get  the  true  surface  temperature  of  metallic  target  (with  ε  ≠  1)  by  measuring  both  the  bidirectional  reflectivity,  ρ’’,  and  the  temperature  radiance,  TR,  at  two  wavelengths.  If  these  two  wavelengths  are  chosen  in  solar  blind  windows,  this  method  is  destined  to  exhibit  the  best  measurement  performance  for  the  surface  temperature  of  CSP  components.  The  development  of  such  solar  blind  bi-­‐color  pyro-­‐reflectometer  is  currently  conducted  at  CNRS-­‐PROMES.  

With  the  CNRS-­‐PROMES  current  non  solar  blind  apparatus,  the  bidirectional  reflectivity,  ρ’’  is  measured  from  pulsed  laser  diodes  working  at  λ1  =  1.3  μm  and  λ2  =  1.55  μm.  The  radiance  temperature,  TR,  is  deduced  directly  from  the  total  flux  emitted  by  the  target  and  collected  by  the  photodiode  by  using  the  Planck’s  law  and  a  calibration  of  the  apparatus  with  a  blackbody.  

The  bicolor  pyroreflectometry  is  not  based  on  standard  hypothesis  such  as  Lambertian  sources  or  grey  body.  Indeed  the  method  introduces  the  diffusion  factor  η:  the  ratio  of  the  hemispherical  reflectivity  ρ’∩  to  the  bidirectional  reflectivity  ρ’’.  π  corresponding  to  illumination  direction  of  the  target  with  a  laser  and  the  observation  direction  of  the  pyroreflectometer.  

 The  main  hypothesis  is  to  consider  that  the  diffusion  factor  is  independent  with  λ1  and  λ2.  This  key  parameter  is  a  thermo-­‐optical  property  directly  linked  to  the  surface  state:    

o When  η  converges  to  1,  the  surface  is  Lambertian:  ρ’’  is  constant  for  any  incident  and  viewing  direction.  

o When  η  converges  to  0,  the  surface  is  specular:  ρ’’  is  strongly  dependent  on  the  incident  and  viewing  direction.  

The  invariance  of  η  with  wavelengths  of  measurement  has  been  experimentally  validated  at  the  solar  furnace  with  thermocouple  for  temperature  validation  and  multi-­‐directional  reflectometer  on  several  metallic  surfaces  [Hernandez  2009]  (notably  with  the  DISCO  instrument).  

Method  Thanks  to  this  approach,  the  advantage  is  to  express  the  emissivity  as  a  function  of  the  diffusion  factor:  

ε  =  1-­‐  ρ’∩  =1-­‐π.η.ρ’’     (2)  

The  measured  radiance  can  be  written  has  following  for  each  wavelength:  

R(T,λ)  =  (1−π.η(T).ρʹ′ʹ′(T,λ  )).R(T,λ)     (3)  

Page 14: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

14  

 

The  Wien  approximation  of  the  Planck  law  is  used  in  our  case  since  this  approximation  is  true  as  soon  as  λ•T  <<  14400  μm.K.  The  T  as  well  as  η  can  be  then  deduced  by  solving  a  system  of  two  equations,  respectively  for  λ1  and  λ2  with  two  unknown  parameters  (T  and  η):  

1/T  =1/TRλ1  +  λ1  /C2  ln(ελ1)  =1/TRλ1  +  λ1  /C2  ln(1−π.η(T).ρʹ′ʹ′(T,λ1))     (4)  

1/T  =1/TRλ2  +  λ2  /C2  ln(ελ2)=1/TRλ2  +  λ2  /C2  ln(1−π.η(T).ρʹ′(T,λ2))     (5)  

��with  C2  =  (h.c)/kB    =  1.4388.10-­‐2  mK    where  h  is  the  second  Planck  constant,  c  is  the  speed  of  light  and  kB  is  the  Boltzman  constant.  

The  solution  converges  to  T*=  T  and  η.  

In  order  to  study  the  sensitivity  of  the  method  on  bidirectional  reflectivity  measurement,  the  error  calculation  has  been  evaluated  from  (4)  and  gives  the  relation:  

∆!∗

!∗=  

!∗∙∆!!,!

!!,!! +  !∙!∙!

!! ∙!∙ 𝜂 ∙ ∆𝜌" + 𝜌" ∙ ∆𝜂   (6)  

In  the  analysis  presented  in  [Delchambre  2011],  it  has  been  concluded  that  the  technic  is  sensitive  to  the  reflectivity  measurement  and  the  accuracy  on  T  increases  when  the  measurement  of  the  bidirectional  reflectivity,  ρ’’,  is  optimum.  In  [Delchambre  2012],  the  parasitic  effects  of  the  radiation  emitted  from  a  plasma  are  studied:  as  this  situation  has  similitudes  to  the  use  of  non  solar  blind  windows,  the  reader  should  refer  to  this  work  for  further  information,  including  pyro-­‐reflectometry  camera  and  tri-­‐color  pyro-­‐reflectometry.  

The  choice  of  the  working  wavelength  range  is  also  of  a  main  interest.  Indeed,  on  the  contrary  to  the  standard  bicolor  thermography  and/or  active  pyrometry  where  ∆λ  =  (λ2  –  λ1)  has  to  be  maximised  to  reduce  the  uncertainty  on  the  surface  temperature  [Loarer  2011],  the  error  on  temperature  with  the  pyroreflectometry  method  depends  on  the  choice  of  λ1  and  λ2  (see  equ.  6).  

It  is  worth  noting  that  the  choice  of  wavelength  was  initially  “due  to  their  proximity  (for  diffusion  factor  invariance),  to  the  silica  optical  fibers  and  the  availability  of  industrial  components  of  low  cost  in  the  spectral  range”  [Hernandez  2009].  However  the  choice  of  wavelength  range  can  be  modified  and  the  choice  of  wavelengths  (in  order  to  keep  the  method  applicable)  has  to  take  into  account,  in  priority,  the  bi-­‐directional  reflectivity  of  the  material.  

Indeed  the  equation  system  can  be  solved  only  when  the  bidirectional  reflectivity  is  different  at  both  wavelengths  in  particular  when  ρ’’  λ1  <  ρ’’  λ2  (λ1  <  λ2)  where  mathematically  there  is  only  one  solution  (which  is  observed  on  most  of  the  metallic  surfaces)  [Hernandez  2005].  

When  ρ’’ʎ1  >  ρ’’ʎ2,  two  calculated  solutions  can  be  obtained.  Thus,  it  is  necessary  to  use  a  third  wavelength  to  determine  the  convergence  to  a  single  point.  

When  ρ’’λ1  =  ρ’’λ2,  the  surface  is  a  grey  body  and  the  surface  temperature  can  be  calculated  with  the  color  temperature,  TC,  deduced  from  the  standard  bi-­‐color  pyrometry  method:  

TC  =  (1/λ1  –  1/λ2)  /  (1/λ1T1  -­‐  1/λ2T2)  

Page 15: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

15  

 

Literature  The  content  of  this  chapter  is  greatly  derived  from  the  following  paper  with  minor  modifications:  

E.  Delchambre,  MH.  Aumeunier,  T.  Loarer,  D.  Hernandez,  Y.  Corre,  E.  Gauthier  (2012):  Performance  of  the  Pyroreflectometry  in  Magnetic  Fusion  Devices  for  Plasma  Wall  Interaction  monitoring,  11th  International  Conference  on  Quantitative  InfraRed  Thermography,  11-­‐14  June  2012,  Naples  Italy.  

The  following  report  is  a  great  and  complete  introduction  and  description  of  the  bi-­‐color  pyro-­‐reflectometry  and  the  practical  instrument  developed:  

D.  Hernandez  (2010):  Bases  de  la  pyroréflectométrie  bicolore  et  presentation  du  pyref.  

D.  Hernandez  et  al.,  Rev.  Sci.  Instr.  80,  094903  (2009)  

E.  Delchambre  et  al.,  Phys.  Scr.  T145  (2011)  014078  (4pp)  

T.  Loarer,  Contrib.  Plasma  Phys.  51,  No.  2-­‐3,  201–206  (2011)  

D.  Hernandez  et  al.,  Measurements  42  (2009)  836-­‐843  

D.  Hernandez  et  al.,  Rev.  Of  Sci.  Instr.  76,  024904  (2005)  

   

Page 16: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

16  

 

5. An  alternative  to  solar  blind  measurements:  active  pyrometry  Active  pyrometry  regroups  3  different  methods,  all  based  on  studying  the  impact  of  a  known  heat  solicitation  in  addition  to  the  steady  state  of  the  body:  

o Pulsed  thermography:  amplitude  based  (PT)  

o Pulsed  thermography:  phase  based  (PPT)  

o Lock-­‐in  thermography:  phase  based  (PLT)  

Application  for  temperature  measurement  for  CSP  applications  is  still  under  research:  it  should  be  noted  that  a  main  application  of  these  methods  is  to  detect  the  presence  of  defects  in  materials,  by  assuming  good  knowledge  of  their  thermal  properties:  non  destructive  testing  (NDE).  This  is  valid  for  fields  where  the  material  abuse  history  is  known  with  good  confidence.  

 

Pulsed  thermography:  log  plot  of  the  relative  temperature  increase  vs  time  decay  as  typically  used  for  non  destructive  testing,  from  [Balageas  2010].  

The  envisioned  application  here  for  CSP  applications  is  mostly  its  counterpart:  the  materials  are  assumed  perfect  or  nearly  perfect  (anisotrop,  no  defects),  but  with  unknown  or  high  uncertainty  on  their  thermal  properties,  as  it  is  the  case  for  materials  abused  in  operating  facilities  with  little  consistent  knowledge  of  the  extents  of  stress  (thermal  gradients  due  to  cloud,  environmental  pollution…).  The  usual  approach  can  also  be  used  in  order  to  determine  the  real  optical  and  thermal  properties  of  the  material  and  then  use  these  properties  with  pyrometers.  

Page 17: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

17  

 

Pulsed  thermography  A  pulsed  laser  is  used  to  induce  surface  heating  in  the  materials:  1  –  50  ms  long  single  shot  or  with  low  rate  repetition.  Fast  pyrometers  or  imaging  cameras  are  used  to  observe  the  induce  thermal  radiation  emission  from  the  component.  3D  modeling  is  used  to  fit  the  observed  behavior  of  the  component  to  a  theoretical  but  accurate  thermo-­‐optical  model.  [Shepard  2007][Rashed  2006][Li  2010].  

Phase  lock-­‐in  active  laser  pyrometry  A  pulsed  laser  (1  Hz  –  1  kHz  sinusoidal  or  square  modulation)  is  used  to  induce  surface  heating  in  the  materials.  Fast  pyrometers  are  used  to  observe  the  induce  thermal  radiation  emission  from  the  component.  Lock-­‐in  amplifier  are  used  to  measure  the  amplitude  and  the  phase  of  the  observed  thermal  waves  which  brings  knowledge  about  the  material  properties  and  state:  

o The  depth  penetration  of  the  thermal  waves  depends  on  their  frequency.  The  faster  the  modulation,  the  more  in  surface  the  wave  stays.  

o The  amplitude  of  the  thermal  waves  depends  on  the  surface  emissivity,  intensity  of  the  heating  laser  modulation,  and  thermal  properties  of  the  heated  material.  

 Principle  of  the  lock-­‐in  method  and  thermal  wave  description  as  presented  in  [Melyukov  2010]  

 

Evolution:  temperature  measurements  in  solar  simulators  and  solar  furnaces  

Choosing  solar  blind  windows  may  prove  to  be  difficult  or  impractical  in  the  field.  One  alternative  is  to  modulate  the  incoming  concentrated  energy  in  order  to  identify  the  reflected  component.  

This  method  shares  some  similitudes  to  the  phase  lock-­‐in  active  pyrometry,  but  the  modulated  energy  does  not  heat  the  surface  sample  because  it  is  chosen  much  faster  than  the  surface  thermal  response  of  the  body.  

The  modulation  is  not  an  additional  laser  but  the  modulation  of  the  incoming  concentrated  solar  beam:  either  modulation  of  the  lamps  electrical  supply  for  a  simulator  or  the  use  of  a  fast  shadowing  

PHASE LOCK-IN LASER ACTIVE PYROMETRY FOR SURFACE LAYER CHARACTERISATION OF TOKAMAKS WALLS

by D. Melyukov*, C. Sortais*, A. Semerok*, P.Y. Thro*, X. Courtois**, D. Farcage*

*CEA Saclay. DEN/DANS/DPC/SCP/LILM **CEA Cadarache, DSM/IRFM/SIPP Abstract

A lock-in method of surface layer characterization of tokamak walls is presented for determining the thermo-physical properties of a layer (like adhesion, thickness and others) deposited on a substrate, by comparing the experimental results of the phase shift between an excitation signal from a modulated laser and the thermal response of the layer, with the predicted phase shift obtained by a theoretical 3D model. A fast method of phase shift calculations for 3D analytical model was developed. The model validation is done on the stainless steel sheets of different thicknesses and the main experiments were performed on an ITER-like sample. 1. Introduction

In modern fusion reactors, the erosion of the plasma facing surfaces results in dust deposition on the tokamak “cold” surfaces, in form of layers which could trap tritium. Metallic surface coatings are also used on some components. To provide efficient operation of tokamaks, it is essential to characterize such layers. In-situ fast surface characterization without the reactor component disassembly is required. The lock-in pyrometry appears as a very suitable method: together with a laser (1 Hz - 1 kHz repetition rate frequency), applied for surface heating it can be used to characterize the micrometric layer properties, for example its adhesion to the substrate, its thickness and others.

The lock-in method is based on the propagation of thermal waves in the material and its interaction with discontinuities or non homogeneities. The depth penetration of thermal waves depends on their frequency. This makes it possible to study thermo-physical properties of materials. Any thermal wave has amplitude and phase. In case of contactless temperature measurements by pyrometry method, the amplitude strictly depends on the surface emissivity, heat flux intensity and other optical parameters that are unknown. But the phase of thermal wave is independent of these parameters. This advantage is often used in the thermal characterization of the materials.

Local laser heating of a layer causes a 3D thermal conduction regime and thus a 1D model becomes not valid. On the other hand the 3D theoretical models calculation takes very long time especially in the case of numerical solution. A fast method of phase shift calculations for the analytic theoretical 3D model of the laser heating of the layer was developed [1,2]. Comparison of both theoretical and experimental results makes it possible to determine the main layer parameters. The lock-in pyrometry system was developed and tested on a WLOH�VDPSOHV������ȝP�:-layer on CFC substrate and graphite tile with deposited layer of 10-���ȝP�WKLFNQHVV�RI�different adhesion).

Figure 1. Principe of the lock-in method and thermal wave description.

Laser t

Substrate

Layer

Reflected thermal wave

Passed thermal wave

Page 18: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

18  

 

modulator  for  solar  facilities.  A  fast  pyrometer  is  used  to  observe  the  luminance  from  the  sample,  which  is  an  addition  of:  

o The  thermal  emission  of  the  sample;  

o The  reflectance  of  the  incoming  concentrated  solar  or  lamps  beam;  

o Part  of  the  time  reflectance  from  the  modulated  additional  energy  from  the  concentrated  beam.  

 

From  [Alxneit  2011]:  Setup  of  the  experiment  with  a  solar  simulator.  (1)  sample,  (2)  vacuum  furnace,  (3)  lens,  (4)  optical  fiber,  (5)  mechanical  chopper,  (6)  diaphragm,  (7)  narrow  bandwidth  transmission  filter,  (8)  PbS  detector,  (9)  Hg/Xe  short  arc  lamp,  (10)  function  generator.  

A  lock-­‐in  amplifier  at  the  frequency  of  the  modulation  allows  to  discriminate  the  thermal  emission  from  the  sample  from  the  reflected  light,  allowing  the  calculation  of  the  temperature  of  the  material  by  assuming  knowing  its  emissivity.  

As  this  method  doesn’t  rely  on  absorption  bands,  it  can  be  applied  to  determine  the  temperature  of  irradiated  surfaces  even  with  solar  simulators  which  lamps  have  thermal  continuous  spectrum.  

This  method  will  be  developed  in  SFERA  II,  WP12,  with  planned  tests  with  a  solar  simulator  and  a  solar  furnace.  

The  complete  principle  is  presented  in:  

Alxneit  I.  (2011):  Measuring  temperatures  in  a  high  concentration  solar  simulator  –  Demonstration  of  the  principle,  Solar  Energy  85  (2011)  516–522  

Page 19: R12.7 Report on solar blind and active pyrometry system …sfera.sollab.eu/downloads/JRA/WP12/Deliverable_R12.7... ·  · 2014-01-22Project and Coordination and Support Action !!

 

 

 

19  

 

Literature  D.  Melyukov,  C.  Sortais,  A.  Semerok,  P.Y.  Thro,  X.  Courtois,  D.  Farcage  (2010):  Phase  lock-­‐in  laser  active  pyrometry  for  surface  layer  characterisation  of  tokamaks  walls,  ������������������������������������������������������������������  �����������������������������������������������������������������10th  International  Conference  on  Quantitative  InfraRed  Thermography,  Quebec.  

S.  M.  Shepard,  J.  R.  Lhota,  and  T.  Ahmed  (2007),  Nondestructive  Testing  and  Evaluation  22,  113  

A.  Rashed,  D.  P.  Almond,  D.  A.  S.  Rees,  S.  E.  Burrows  and  S.  Dixon  (2006),  Rev  Quantitative  NDE  26ed.  D.  O.  Thompson  and  D.  E.  Chimenti,  500-­‐506.  

T.  Li,  D.  P.  Almond,  D.  A.  S.  Rees,  B.  Weekes  (2010):  Crack  imaging  by  pulsed  laser  spot  thermography,  Journal  of  Physics:  Conference  Series  214  (2010)  012072,  doi:10.1088/1742-­‐6596/214/1/012072  

Alxneit  I.  (2011):  Measuring  temperatures  in  a  high  concentration  solar  simulator  –  Demonstration  of  the  principle,  Solar  Energy  85  (2011)  516–522