Comparison of heat flux measurements by IR thermography ...
1
LCFS IR View Calorimeters Arc Length Coordinate Vertical Face Nose Shelf 10 09 Langmuir Probes 08 Inner Divertor GH Limiter 1. Abstract 4. Results 5. Conclusions 1. P. C. Stangeby, The Plasma Boundary of Magnetic Fusion Devices, IoP Publishing, 2000. 2. J. Marki et al., J. Nucl. Mater. 363–365 (2007) 382–388. Poster available at: mit.edu/brunner/www/ Thermocouple Profile Simulation 0 5 10 15 20 25 30 Time [s] 400 500 600 700 800 Temperature [K] Measured TC Simulated TC Ideal TC Shot: 1100310027 0 1 2 3 4 Energy Balance [MJ] Energy In Ohmic ICRF Energy Out Outer Divertor Inner Divertor Radiation E out/E in: 0.94 K AB GH Limiters molybdenum tile molybdenum slug Inconel sleeve Alcator C-Mod was recently upgraded with multiple diagnostics to measure the heat flux footprint on the outer divertor: • IR thermography system • Plasma conditions from Langmuir probes (LPs) • Embedded calorimeters to measure local energy deposition In addition to providing valuable consistency checks of the inferred heat flux profiles, these redundant measurements allow an assessment of the local sheath heat-flux transmission coefficient (γ). We report measurements of the sheath heat-flux transmission coefficient at multiple locations in the scrape-off-layer during an EDA H-mode discharge. The LP data are consistent with the presence of a heat flux “tail” measured by the IR system, extending into the far SOL. However, measured γ are found to deviate from theoretical values by up to a factor of 3, depending on position in the scrape-off layer, pointing to some as yet unresolved physics. Calorimeter-inferred energy deposition profiles are found to be within 20% of the IR-inferred values across the entire footprint. 2. Experiment • EDA: ELM-free regime of steady H-mode operation in Alcator C-Mod -B t = 5.4 T I p = 0.5 MA N l04 = 1.1*10 20 m -3 P ICRF = 3 MW • Three “good” Langmuir Probes (LP) - Lower probes showing signs of shadowing • Limiters and inner divertor tiles instrumented with thermocouples for energy balance calculations Comparison of heat flux measurements by IR thermography and probes in the Alcator C-Mod divertor D. Brunner, B. LaBombard, J. Payne, and J.L. Terry Plasma Science and Fusion Center, M.I.T, Cambridge, MA 02139, USA Supported by USDoE Coop. Agreement DE-FC02-99ER54512 Typical calorimeter assembly IR camera view of ramped tiles FLIR Titanium SC7000 -320x256 resolution -3-5 μm band 3. Analysis ρ=0 Sheath heat flux transmission coefficient Thermocouple interpretation • Calorimeter energy deposition measurements rely on being able to measure the thermal rise of the calorimeter body before energy is lost to the surrounding environment. • Our thermocouples are tipped with 3 mm of stainless steel, the response time of the thermocouple is not adequate to directly measure the thermal rise of the calorimeter. • Thus necessitating use of a numerical 1-D heat transfer algorithm with temperature dependent thermal properties to calculate the deposited energy: 1) Apply 1 second heat pulse to the calorimeter. 2) Pull temperature evolution at thermocouple tip and apply as boundary on 1-D thermocouple system. 3) Compare temperature evolution at various depths in the thermocouple simulation to measured profile. 4) Iterate until minimum error between simulated profile and measured is found. • To bound possible systematic errors in the 1-D model, the time interval over which the modeled and measured temperature signals are compared was varied. • The inferred deposited energies were found to vary by less than 10%. - Theory relating heat flux from a plasma to a material surface through LP measurable quantities (electron temperature and ion saturation current) 1 : (1) τ is the ratio of ion to electron temperature (T i /T e ) δ e is the secondary electron emission coefficient ι is the ratio of net current through the sheath to ion saturation current (J gnd /J sat ) - Often assumed: τ = 1, δ e = 0, and ι = 0 resulting in γ = 7.5 - γ has been measured 2 between 2 and 20 - Low values explained by LPs over reporting T e when there are fast electrons 3 - High values attributed to secondary electron emission, empirical formula 1 : (2) for secondary emission on a clean surface at normal incidence with monoenergetic electron beam. For tungsten: δ max = 1.4 and E max = 650 eV Calorimeters and IR measurements agree better than 20% - More variation above divertor nose because IR heat flux takes pixel line- outs at about mid-ramp, calorimeters above the nose are at either the top or bottom of the ramp (temperature variation between mid-ramp and top/bottom is usually ~20%) - Confirms TC analysis routine and allows us to use it on tile TCs: Now can examine energy balance -Thermocouples in inner divertor and limiters allow for energy balance calculations -Account for >85% of energy in lower single null discharges. -Currently don’t consider “double counted” radiation. Heat flux transmission coefficient • The new suite of diagnostics has greatly expanded C-Mod’s capabilities for divertor heat flux experiments. • These redundant systems provide important cross-checks for data validation. • Energy deposition profiles from calorimeter sensors are found to agree with IR-inferred heat depositions within 20%. • In comparing the IR camera and LP data, tests of the plasma-sheath heat flux transmission theory are performed, including the effects of a measured finite sheath current. It is found that the standard assumptions of T i = T e and δ e = 0 can not explain the measured values of γ at the divertor surface, being a factor of 3 off in some cases. • Secondary electron emission effects can not account for the discrepancy; this effect has the wrong dependence on T e . • Postulating a profile of T i = 50 eV pushes the theoretical values closer to measured values, although significant deviations still exist. There are no measurements of T i to support such a possibility at this time. • Motivated in part by these observations, significant improvements to C-Mod’s heat flux diagnostic set are planned for the next run campaign, including LPs located in direct view of the IR camera, fast-response thermocouples in the calorimeters, and surface thermocouples to calibrate the IR system at high temperatures. 0 2 4 6 8 10 12 ρ [mm] -2.0 -1.5 -1.0 -0.5 0.0 0.5 Floating Potential (V f ) Electron Temperature (T e ) 0.0 0.5 1.0 1.5 2.0 2.5 Plasma Density 0 10 20 30 40 Electron Temperature 0 5 10 15 Measured Heat Flux Transmission Coefficient (γ) -2 -1 0 Ground Current (J gnd ) Ion Saturation Current (J sat ) -3 LP08 LP09 LP10 LP Measurements During EDA H-Mode shot: 1100224008 [10 20 m -3 ] ι = [eV] • Compare IR and calorimeter inferred energy deposition profiles • Compare IR and Langmuir probe heat flux profiles with assumptions on secondary electron emission and the ratio of ion to electron temperature -10 -5 0 5 10 15 0 100 200 300 ρ [mm] LP08 LP09 LP10 IR & LP Parallel Heat Flux Shot: 1100303019 [MW/m 2 ] IR Camera t=1.2 1.4 s IR & Calorimeter Deposited Energy 5 10 15 20 Poloidal Arc Length [cm] 0 5 10 15 Energy [MJ/m 2 ] Calorimeters IR Camera Shot: 1100310027 High on Ramp Low on Ramp • IR camera show a long tail of about 15% of the peak heat flux extending far into the SOL. • Comparison of the IR-inferred heat flux profile with LPs (assuming γ = 7.5) confirms the tail to be real. 0 5 10 15 0 5 10 15 0 5 10 15 0 2 4 6 8 10 12 ρ [mm] T i = 50 eV with no secondary electrons shot: 1100224008 T i = T e with secondary electrons T i = T e with no secondary electrons 0 2 4 6 8 10 12 ρ [mm] 0 2 4 6 8 10 12 ρ [mm] • A sweep of the strike point allows exploration of sheath heat flux transport through a variety of plasma conditions. • γ is inversely correlated with electron temperature. • γ is correlated with the floating potential/ground current. • Comparing the measured γ with Eq. 1: - Equal ion and electron temperatures and no secondary electrons does not reproduce the profile. - Inclusion of secondary electrons does not improve the match. - Secondary electron emission has the wrong dependence on T e . - Including a hypothetical T i = 50 eV (Alcator C measured4 Ti/Te ≈ 2) improves the profile but does not resolve the disagreement. 3. P. C. Stangeby, Plas. Phys. Control Fus. 37 (1995). 4. A. S. Wan et al., Rev. Sci. Instrum. 57 (1986) 1542–1551.
Transcript of Comparison of heat flux measurements by IR thermography ...
LCFSIR View
CalorimetersArc LengthCoordinate
Vertical Face
Nose
Shelf
1009 Langmuir Probes08
InnerDivertor
GHLimiter
1. Abstract
4. Results
5. Conclusions
1. P. C. Stangeby, The Plasma Boundary of Magnetic Fusion Devices, IoP Publishing, 2000.2. J. Marki et al., J. Nucl. Mater. 363–365 (2007) 382–388.Poster available at: mit.edu/brunner/www/
Thermocouple Pro�le Simulation
0 5 10 15 20 25 30Time [s]
400
500
600
700
800 Temperature [K]
Measured TC
Simulated TC
Ideal TC
Shot: 1100310027
0
1
2
3
4
Energy Balance [MJ]
Energy In
Ohmic
ICRF
Energy Out
Outer Divertor
Inner Divertor
Radiation
E out/E in: 0.94
K
ABGHLimiters
molybdenum tile
molybdenum slug
Inconel sleeve
Alcator C-Mod was recently upgraded with multiple diagnostics to measure the heat flux footprint on the outer divertor: • IR thermography system • Plasma conditions from Langmuir probes (LPs) • Embedded calorimeters to measure local energy depositionIn addition to providing valuable consistency checks of the inferred heat flux profiles, these redundant measurements allow an assessment of the local sheath heat-flux transmission coefficient (γ). We report measurements of the sheath heat-flux transmission coefficient at multiple locations in the scrape-off-layer during an EDA H-mode discharge. The LP data are consistent with the presence of a heat flux “tail” measured by the IR system, extending into the far SOL. However, measured γ are found to deviate from theoretical values by up to a factor of 3, depending on position in the scrape-off layer, pointing to some as yet unresolved physics. Calorimeter-inferred energy deposition profiles are found to be within 20% of the IR-inferred values across the entire footprint.
2. Experiment
• EDA: ELM-free regime of steady H-mode operation in Alcator C-Mod -Bt = 5.4 T Ip = 0.5 MA Nl04 = 1.1*1020 m-3 PICRF = 3 MW• Three “good” Langmuir Probes (LP) - Lower probes showing signs of shadowing• Limiters and inner divertor tiles instrumented with thermocouples for energy balance calculations
Comparison of heat flux measurements by IR thermography and probes in the Alcator C-Mod divertorD. Brunner, B. LaBombard, J. Payne, and J.L. Terry
Plasma Science and Fusion Center, M.I.T, Cambridge, MA 02139, USA Supported by USDoE Coop. Agreement DE-FC02-99ER54512
Typical calorimeter assembly
IR camera view of ramped tilesFLIR Titanium SC7000 -320x256 resolution -3-5 μm band
3. Analysis
ρ=0
Sheath heat flux transmission coefficient
Thermocouple interpretation• Calorimeter energy deposition measurements rely on being able to measure the thermal rise of the calorimeter body before energy is lost to the surrounding environment. • Our thermocouples are tipped with 3 mm of stainless steel, the response time of the thermocouple is not adequate to directly measure the thermal rise of the calorimeter.• Thus necessitating use of a numerical 1-D heat transfer algorithm with temperature dependent thermal properties to calculate the deposited energy: 1) Apply 1 second heat pulse to the calorimeter. 2) Pull temperature evolution at thermocouple tip and apply as boundary on 1-D thermocouple system. 3) Compare temperature evolution at various depths in the thermocouple simulation to measured profile. 4) Iterate until minimum error between simulated profile and measured is found. • To bound possible systematic errors in the 1-D model, the time interval over which the modeled and measured temperature signals are compared was varied.• The inferred deposited energies were found to vary by less than 10%.
- Theory relating heat flux from a plasma to a material surface through LP measurable quantities (electron temperature and ion saturation current)1:
(1)
τ is the ratio of ion to electron temperature (Ti/Te) δe is the secondary electron emission coefficient ι is the ratio of net current through the sheath to ion saturation current (Jgnd/Jsat)- Often assumed: τ = 1, δe = 0, and ι = 0 resulting in γ = 7.5- γ has been measured2 between 2 and 20- Low values explained by LPs over reporting Te when there are fast electrons3
- High values attributed to secondary electron emission, empirical formula1:
(2)
for secondary emission on a clean surface at normal incidence with monoenergetic electron beam. For tungsten: δmax = 1.4 and Emax = 650 eV
Calorimeters and IR measurements agree better than 20% - More variation above divertor nose because IR heat flux takes pixel line- outs at about mid-ramp, calorimeters above the nose are at either the top or bottom of the ramp (temperature variation between mid-ramp and top/bottom is usually ~20%) - Confirms TC analysis routine and allows us to use it on tile TCs:
Now can examine energy balance -Thermocouples in inner divertor and limiters allow for energy balance calculations -Account for >85% of energy in lower single null discharges. -Currently don’t consider “double counted” radiation.
Heat flux transmission coefficient
• The new suite of diagnostics has greatly expanded C-Mod’s capabilities for divertor heat flux experiments.• These redundant systems provide important cross-checks for data validation.• Energy deposition profiles from calorimeter sensors are found to agree with IR-inferred heat depositions within 20%.• In comparing the IR camera and LP data, tests of the plasma-sheath heat flux transmission theory are performed, including the effects of a measured finite sheath current. It is found that the standard assumptions of Ti = Te and δe = 0 can not explain the measured values of γ at the divertor surface, being a factor of 3 off in some cases.• Secondary electron emission effects can not account for the discrepancy; this effect has the wrong dependence on Te.• Postulating a profile of Ti = 50 eV pushes the theoretical values closer to measured values, although significant deviations still exist. There are no measurements of Ti to support such a possibility at this time.• Motivated in part by these observations, significant improvements to C-Mod’s heat flux diagnostic set are planned for the next run campaign, including LPs located in direct view of the IR camera, fast-response thermocouples in the calorimeters, and surface thermocouples to calibrate the IR system at high temperatures.
0 2 4 6 8 10 12ρ [mm]
-2.0
-1.5
-1.0
-0.5
0.0
0.5 Floating Potential (Vf)Electron Temperature (Te)
0.0
0.5
1.0
1.5
2.0
2.5Plasma Density
0
10
20
30
40Electron Temperature
0
5
10
15
Measured Heat Flux Transmission Coe�cient (γ)
-2
-1
0
Ground Current (Jgnd)Ion Saturation Current (Jsat)
-3
LP08LP09
LP10
LP Measurements During EDA H-Mode
shot
: 110
0224
008
[1020 m-3]
ι =
[eV]
• Compare IR and calorimeter inferred energy deposition profiles• Compare IR and Langmuir probe heat flux profiles with assumptions on secondary electron emission and the ratio of ion to electron temperature
-10 -5 0 5 10 150
100
200
300
ρ [mm]
LP08LP09 LP10
IR & LP Parallel Heat Flux
Shot
: 110
0303
019
[MW/m2]
IR Camera
t=1.2 1.4 s
IR & Calorimeter Deposited Energy
5 10 15 20Poloidal Arc Length [cm]
0
5
10
15Energy [MJ/m2]
Calorimeters
IR Camera
Shot
: 11
0031
0027
High on Ramp
Low on Ramp
• IR camera show a long tail of about 15% of the peak heat flux extending far into the SOL. • Comparison of the IR-inferred heat flux profile with LPs (assuming γ = 7.5) confirms the tail to be real.
0
5
10
15
0
5
10
15
0
5
10
15
0 2 4 6 8 10 12ρ [mm]
Ti = 50 eV with no secondary electrons
shot
: 11
0022
4008
Ti = Te with secondary electrons
Ti = Te with no secondary electrons
0 2 4 6 8 10 12ρ [mm]
0 2 4 6 8 10 12ρ [mm]
• A sweep of the strike point allows exploration of sheath heat flux transport through a variety of plasma conditions.• γ is inversely correlated with electron temperature.• γ is correlated with the floating potential/ground current.• Comparing the measured γ with Eq. 1:
- Equal ion and electron temperatures and no secondary electrons does not reproduce the profile.
- Inclusion of secondary electrons does not improve the match. - Secondary electron emission has the wrong dependence on Te.
- Including a hypothetical Ti = 50 eV (Alcator C measured4 Ti/Te ≈ 2) improves the profile but does not resolve the disagreement.
3. P. C. Stangeby, Plas. Phys. Control Fus. 37 (1995).4. A. S. Wan et al., Rev. Sci. Instrum. 57 (1986) 1542–1551.
