Skip to main content

Full text of "Intensity Biased PSP Measurement"

See other formats


^JUAA 



AIAA 2000-2526 

INTENSITY BIASED PSP MEASUREMENT 

C. S. Subramanian 

Florida Institute of Technology, Melbourne, FL 

T. R. Amer, D. M. Oglesby, C. G. Burkett 
NASA Langley Research Center, 
Hampton, VA 



21st AIAA Aerodynamic 
Measurement Technology 

and Ground Testing Conference 

19-22 June 2000 / Denver, CO 



For permission to copy or republish, contact the Anrerican Institute of Aeronautics and Astronautics 
1801 Alexander Bell Drive, Suite 500, Reston, VA 20191 



AIAA-2000-2526 



INTENSITY BIASED PSP MEASUREMENT 

Chelakara S. Subramanian*, Tahani R. Amer*, 

Donald M. Oglesby *, and Cecil G. Burkett, Jr.* 

NASA Langley Research Center, Hampton, VA 23681 



ABSTRACT 

The current pressure sensitive paint (PSP) technique 
assumes a linear relationship (Stem-Volmer 
Equation) between intensity ratio (I,/!) and pressure 
ratio (P/Po) over a wide range of pressures (vacuum 
to ambient or higher). Although this may be valid for 
some PSPs, in most PSPs the relationship is 
nonlinear, particularly at low pressures (<0.2 psia 
when the oxygen level is low). This non-linearity 
can be attributed to variations in the oxygen 
quenching (de-activation) rates (which otherwise is 
assumed constant) at these pressures. Other studies 
suggest that some paints also have non-linear 
calibrations at high pressures; because of 
heterogeneous (non-uniform) oxygen diffusion and 
quenching. Moreover, pressure sensitive paints 
require correction for the output intensity due to light 
intensity variation, paint coating variation, model 
dynamics, wind-off reference pressure variation, and 
temperature sensitivity. Therefore to minimize the 
measurement uncertainties due to these causes, an in- 
situ intensity correction method was developed. 

A non-oxygen quenched paint (which provides a 
constant intensity at all pressures, called non-pressure 
sensitive paint, NPSP) was used for the reference 
intensity (Inpsp) with respect to which all the PSP 
intensities (I) were measured. The results of this 
study show that in order to fully reap the benefits of 
this technique, a totally oxygen impermeable NPSP 
must be available. 

NOMENCLATURE 

Ai& B Stem-Vohner Coefficients 

/ Emission intensity at oxygen pressure P 

lo Emission intensity at zero oxygen partial 

pressure 

Inpsp Emission intensity of Non-PSP 

Ipsp Emissionintensity of PSP 

K Stem-Vohner constant 



Kf Quenching rate of fluorescence 

Kic Quenching rate of intemal conversion 

Kt, Quenching rate of oxygen 

T] Paint efficiency 

(j) Quantum Efficiency 

Pi Reference Pressure 

P2 Measured pressure 

INTRODUCTION 

PSP measurements provide a means for the recovery 
of global surface pressure distributions on 
aerodynamic test articles. A typical PSP (Figure 1) 
consists of a 25-40 ^.m thick reflective undercoat and 
a 25-40 \im thick coating of a luminophore dispersed 
in a binder layer. The binder is usually a polymeric 
material. The luminophore must be one for which its 
luminescence is quenched by oxygen. The principle 
of operation of pressure sensitive paints are well 
described in the literatures'"'. 

The intensity response of the paint, /, is related to 
incident intensity, io, paint efficiency, rj, and 
luminescence quantum efficiency, ^, by 



I=ion^ 



(1) 




undercoat 



Figure 1. PSP schematic Responce 



' Professor of Aerospace Engineering, Associate Fellow of AIAA 
at Florida Institute of Technology, Melbourne, FL 32901, 

* Aerospace Engineer 
'Analytical Chemist, 

* Technologist, 

Copyright ©2000 by the American Institute of Aeronautics And Astronautics, Inc. No copyright is asserted in the Unites States 
under Title 17, U.S. Code. The U.S. Government has royalty-free license to exercise all rights under the copyright claimed 
herein for Government Purposes. All other rights are reserved by the copyright owner. 

1 
American Institute of Aeronautics and Astronautics 



The paint efficiency is a function of the reflectivity of 
the primer coat, the concentration of the luminophore 
in the binder and the permeability/diffusivity of 
oxygen in the binder. The luminescence efficiency 
basically is a measure of luminophore performance, 
which is governed by its activation and de-activation 
rates (Figure 2). The activation is caused by the 
absorption of photons by luminophores received 
directly from the light source as well as from the 
reflective imdercoat layer, presuming the binder is 
inactive. 



Where Iq is the emission at zero oxygen level and I is 
the emission at any pressure P. K is equal to v , l{Kf 
+Ki^ which is assumed to be a constant. It is 
generally not practical to measure Iq in the wind 
timnel environment; since the tunnel would have to 
be pumped down to a vacuum. Instead of trying to 
achieve zero oxygen conditions the intensity of 
emission at "vwnd off', Ii, is used as the reference 
intensity, and the pressure at "wind off' is considered 
the reference pressure. Pi. In practice this is usually 
the local barometric pressure. In terms of the Stem- 
Vobner equation this takes the form of the ratio of 
the Stem-Volmer relation for two pressures. 



1 



JABLONSKl DIAGRAM 



■IC 



"ISC 



T, 



_So_ 



-Radiative 



- Non -Radiative 



Figure!. Photon activation and deactivation 
process 

The de-activation, on the other hand, occurs by the 
non-radiative processes; such as, internal conversion 
(IC), intersystem crossing (ISC, which are 
temperature dependent) and oxygen quenching 
(which is pressure dependent), and by radiative 
processes Wks fluorescence anA phosphorescence. In 
PSP applications, the emission intensity is correlated 
to the partial pressure of oxygen (proportional to 
oxygen concentration) by expressing the quantum 
efficiency in terms of de-activation rates as 



/ = 



hWf 



(2) 



Where KfiKicZnd k: , are the quenching rates of 
fluorescence, internal conversion and oxygen, 
respectively. It is assumed that under equilibrium 
conditions the denominator in Equation 2 is equal to 
the photon (activation) energy absorbed by the 
luminophore. 

Under the appropriate illimiination and constant 
quenching rates of fluorescence and internal 
conversion, the intensity of the luminescence 
emission from the paint is inversely proportional to 
the oxygen concentration, and, hence, the air pressure 
on the surface. The luminescence of PSP may be 
expressed in terms of the well-known Stem Volmer 
relation given by 



Io/r= 1+KP 



(3) 



wind off Iq_/I,= 1+KPi 
wind on Iq/I ^ I+KP2 

I,/l2=(l+KP2)/(l+KP0 



(4) 
(5) 
(6) 



I,/l2 ={1/(1+KP,)} + (KP2/(UKP0} (7) 

K and Pi are constants, thus (7) may be expressed as: 



I,/l2=A+BP2 

A = l/(l+KPi) 
B=K/(l+KPi) 



(8) 
(9) 



(10) 



Since luminescence intensity depends on illumination 
intensity, values for I are determined for each point 
on the wind tunnel model at each angle of attack. The 
values for A and B are then determined from a plot of 
I1/I2 vs. pressure, using pressure taps on the model for 
calibration. The accuracy of this type of calibration 
depends on maintaining constant and reproducible 
illimiination at every model position. Since the light 
intensity at the surface of the model changes with the 
angle of attack, the reference intensity (Ii ) at every 
model position must be measured. In order to 
correctly ratio these wind-off measurements to the 
wind-on measurements, spatial registration dots must 
be placed on the model. These enable the wind-off 
and wind-on images to be correctly aligned. The 
registration marks are usually round, black dots about 
6 mm in diameter. 

The objective of this study is to develop a paint and 
measurement system that would not require the "wind 
off' calibration and would correct for differences in 
illumination intensity over the model surface. Others 
have used dual limiinophore PSPs to correct for light 
intensity variations and also temperature variations^'*. 
However, mixing different luminophores in the same 



American Institute of Aeronautics and Astronautics 



paint matrix nearly always produces spectral 
interference between the different luminophores. If 
the registration dots could be prepared from paint 
containing a luminophore that is not quenched by 
oxygen, the emission from the dots can also serve as 
a light intensity reference. The luminophore in the 
registration dots should be one that is excited by the 
same illumination used to excite the pressure-sensing 
luminophores in the PSP. Theoretically, the optimum 
system would have the same luminophore in the 
registration dots as in the paint but contained in a 
binder that is oxygen impermeable. This way the 
pixel intensity at the dot can be used as the light 
reference mtensity. Having the same luminophore in 
the dot would eluninate the need for a filter wheel or 
filter shuttle on the camera in order to observe 
different wavelengths of light. The pixels 
representing the dot could be used for light 
referencing. However, practical binders have some 
oxygen permeability and even a small amount of 
quenching of the luminophore in the dot would cause 
serious errors. It is easier to find a luminophore, 
which is not quenched by oxygen than it is to find a 
totally impermeable polymer paint matrix. This 
would mean that the target dots would emit at a 
different wavelength of light than the PSP. Therefore, 
the reference luminophore should emit at a 
wavelength sufficiently different from that emitted by 
the pressure sensing luminophore to be resolved with 
different filters over the camera lens. Instead of 
taking the ratio of wind-off intensity to the wind-on 
intensity, the ratio of the intensity of emission from 
the nearest registration mark to the PSP emission 
intensity of interest would be used. Although this 
may not give a perfect correction for variations in 
light intensity, it should give a reasonably good 
correction. 

If the temperature sensitivity of the luminophore in 
the registration dots and the temperature sensitivity of 
the PSP are known, the limiinescence from the dots 
can be used to measure the temperature of the paint 
surface. This information could be used to make a 
correction for the effect of temperature on the PSP 
intensity. 

The presented approach avoids the low reference 
intensity (and hence camera resolution) problems 
normally encountered in atmospheric pressure 
referencing. Furthermore, if the intensity variation 
due to temperature of the NPSP is made the same as 
the PSP, then the proposed referencing method can 
be also used for temperature sensitivity correction of 
the PSP. Also, the NPSP can be used as target 
markers for model deformation determination in 
wind tuimel testing. 



Some tests were performed in the NASA Langley 
laboratory on a painted circular coupon in a test 
chamber to demonstrate this technique. The 
following section describes the experimental set up. 
Then, experiments with various PSP and NPSP 
combinations are explained. The results and 
discussions are presented in the following section, 
and fmally some conclusions are given. 



EXPERIMENTAL SET UP 

The tests were conducted at one of the NASA- 
Langley Research laboratory. Figure 3 shows the 
experimental set up, which consists of an adjustable- 
vacuum chamber to vary the pressure range, a 
pressure transducer to measure pressure reading, two 
lamps for excitation, two 12 bit CCD digital cameras 
or two 16bit CCD cameras, two T-type 
thermocouple thermometers to monitor the specimens 
and test chamber temperature, and a data acquisition 
system for the specified cameras. 

The test coupon size was 76 mm diameter. In the 
center of the coupon a 7 mm dia circle was painted 
with NPSP, and the remaining surface was painted 
widi regular PSP paint (See Figure (4)). Several 
coupons were made with different NPSPs and PSPs 
to test the best performing paints over a given range 
of pressures and temperatures. 



Praaturi Controll«f 



I. 



rsBi ^btH 

»T.m ^y^ »T-13» 




Figures. Schematic of Experiment Set-up 



American Institute of Aeronautics and Astronautics 




Figure 4. Picture and Schematic of the test 
Specimen 

EXPERIMENTAL PROCEDURE 

Several tests were performed on a painted circular 
coupon to demonstrate this technique and verify the 
theory. Six specimens of different PSPs and NPSPs, 
and camera combinations were tested. The pressure 
range was varied from 0.0-2.7 atmospheres, and the 
temperature range was varied from 15-35 °C. The 
following Table 1 shows the PSP and the NPSP paint 
and paint binder combinations tested. 



Spec 


PSP 
Lum. 


PSP 
Binder 


NPSP 
Lum. 


NPSP 
Binder 


1 


PtPFPP 


FEM-IBM* 


Ru-bypy 


Clear Coat 


2 


PtPFPP 


FEM-IBM 


PtPFPP 


Clear Coat 


3 


Ru-Bath 


RTV-118 


Ru-bath 


Clear Coat 


4 


PtPFPP 


FEM-IBM 


RPFPP 


PMMA 


5 


PtPFPP 


FEM-IBM 


PtPFPP 


Clear Coat 


6 


PtPFPP 


FEM-IBM 


Ru-bypy 


Clear Coat 



♦ 2,2,2-trifluoroethymethacrylate-co- 
isobutylmethacrylate 

Table 1. Paint and Binder combinations that were 
tested 



Figure 5 A and B show actual images of the test 
specimen at 0.035 and 14.7 psia, respectively. The 
PSP and NPSP in Figure 5A have their respectively 
maximum intensity levels at 0.035 psia. However, 
the PSP in Figure 5B is quenched and its intensity 
level is considerably reduced. 




Figures. A. Image of PSP and NPSP at 0.035 psia 
B. Image of PSP and NPSP at 14. 7 psia 



RESULTS AND DISCUSSIONS 

The results presented in the following Figures show 
how the PSP and NPSP intensities vary with respect 
to pressure, temperature, paint type and time. 

Figure 6 shows the pressure response curves of the 
PSP and NPSP for the specimen 6 at 0.01 psia 
pressure (lowest pressure) and at 26 °C. The PSP 
intensity is found to decrease by a factor of 10 when 
the pressure is increased from 0.01 to 20 psia, while 
there is hardly any change in the NPSP intensity. 
Thus, the NPSP provides a constant intensity to 
which the PSP intensity can be referenced. 



NPSP & PSP Nomallzed Intensities, Specimen 6 at 260 



f 1 

I 0.8 

I 0.6 

I 0.4 

i 0.2 



i- - T - - 




— ' 






\ 












\ 












-*-{\l\o)pso 
-^(Wo)npsp 


\ 










V 




























* * 





10 15 

Pressure (psia) 



20 



25 



Figure 6. Plots of the normalized PSP and NPSP 
intensities. 

Figure 7 is a plot of the ratio of the NPSP and the 
PSP normalized intensities, which can be used for 
calibrating a given PSP and NPSP combination. For 
the specimen 6 tests, one light source with a 
390+-40 nm filter was used to excite both NPSP and 
PSP. A 16-bit Photometries camera with a 650 +-10 
nm filter was used for the PSP imaging and a 16-bit 
Photometric camera with a 580+-9 nm filter was used 
for the NPSP imaging. There was a 6.25 mm (0.25") 
wide darkened ring around the NPSP. 



American Institute of Aeronautics and Astronautics 





Ratio of Normalized Intensities, Specimen 6 at 26C 


1- 

8 10 
1 8 

! I 




















^ 








^'' 






^^~ 






-♦-(l/io)npsp/(l/lo)psp 




^ 










^^ 










r 










( 


) 5 10 15 20 25 
Preuura (psia) 



Figure 7. Ratio of PSP and NPSP normalized 
intensities versus pressure for specimen 6 



Figures 8 shows the effect of temperature on the 
calibration curves for specimen 1 . It should be noted 
that the plots in Figure 8 were based on using 
intensity reference at the temperature of the run. The 
changes in the shapes of the calibration curves show 
that the PSP binders were temperature sensitive. 
Intensity reference (1 ref) also changed with 
temperature, which shows that the luminophore was 
also sensitive to temperature. Thus, the use of the 
PSP and the NPSP does very little to correct for the 
effect of the temperature on the PSP calibration 
curves. 



Effect of Temperature, Specimen 1 



25 
o 20 
^- 15 



•:^1Q- 



^ 5 



— - 




^. 


^ - 






Jt^ 


\<^ 




.i^ 




26C 

15C 




> 


^ 


^ 


-♦-Rabo 
-^RaBo 




^ 




RaBo35C 










..*^ 

w 


1 1 









10 



20 30 

Pressure (psia) 



40 



50 



Figure 8. The Effect of temperature on the ratio of 
normalized PSP and NPSP intensities versus 
pressure for specimen 1 

The intensity ratio variations show that the intensity 
ratio increased for increasing temperatures. For the 
specimen 1 tests, one light source with a 450 nm +-40 
filter is used to excite both NPSP and PSP. However, 
a 12-bit Photometries camera with a 650 +-10 nm 
filter is used for the PSP unaging and a 16-bit 
Photometric camera with a 580+-9 nm filter is used 
for the NPSP unaging. There was no darkened ring 
around the NPSP. Because of the spectral intensity 



separation, no cross-talk between PSP and NPSP 
intensities is evident. 



To better understand the effect of different binders 
and luminphore combinations, the calibration plots 
were developed for three different paint specimens, 1 , 
2 and 3 (see Table 1) as shown in Figure 9. 





Effect of Paint Specimen 






IntBiwIty Ratio 








.. 




















-•- Ratio Sperl 
-»- Ratio Spec.? 
Ratio Specs j 








u—.— H 




U^ 


t^^^_ 




















10 20 30 40 50 
Pressure (psia) 





Figure 9. Ratio of NPSP and PSP normalized 
intensities versus pressure for different paint 
specimens 

If each specimen can be analyzed separately, one can 
notice the effect of each paint combination. For 
example, specimen 1 which had Ru-Bypy as the 
luminophores for the NPSP was not quenched at all. 
This gave a good light reference(See Figure lOA and 
lOB.) 



WSP, PSP Nomallad ItttniWM vs. Pnasunt Speckmn 1 
M26C 



1.2 T— 1 j r 1 j 

II 0.6 t •-(MojumSpKl 

I- ?1 \L -*-(MoW»P Sp«ic 1 



((■■IS) 



Ratio of Nonnallzsd Intensities. Spec 1 









.. 






.^< 


^ 






_-^ 


^ 




^v 


r***^ 


' ^-•-RaboSfwcl r 


^ 











10 20 30 40 50 

PrvBMjr* Iptia) 



Figure lO.A. Plots of the normalized PSP and NPSP 
Intensities, sped. 
B. Ratio of PSP and NPSP normalized 
intensities versus pressure for sped. 



American Institute of Aeronautics and Astronautics 



specimens 2 and 3 use the same luminophore for PSP 
and NPSP, but with different binders. In this case, 
since the emission wavelength was the same for 
NPSP and PSP, significant spectral leakage of 
intensity occurred. The spectral leakage was 
minimized by placmg a darkened ring around the 
NPSP. Specimen 2 has PtPFPP as NPSP 
luminophore, which give a larger signal, but was 
quenched by oxygen, and reference intensity ratio is 
decreased by an increase in the pressure (see 
figurellA). Notice that the Figure IIB is curved 
downward due to that effect. 



NPSP, PSP Nonnallzed IntamiOw vs. Pressure, 
Specimen 3 at 26 C 



2 ^ '■— ■■^^^ii^i= r 



-(l/lo)ptp Spac 3 
-{Ulo)np«p Spac 3 



PrMsura (psia) 



Figure 12A . Plots of the normalized PSP and NPSP 
intensities, spec. 3. 



NPSP, PSP Normalized Intensities vs. Pressure, 
Specimen 2 at 26 C 



^ 12 

= 0.8 

■J 0.6 

I 0.4 

I 0.2 

o 

z 



■ Ml' 



- (Wo)psp Spec 2 

- (lyio)npsp Spec 2 



10 20 30 40 

Pressure (psia) 



50 



Figure 11 A. . Plots of the normalized PSP and NPSP 
intensities, spec. 2. 



Ratio of Normalized Intensities, Spec 2 



|i 15 

II- 



f — ^^ 



-Ratio Spec2 



20 30 

Prvuura (p*l«) 



Figure IIB. Ratio of PSP and NPSP normalized 
intensities versus pressure for spec. 2. 



Specimen 3 (See Figure 12A and B), the NPSP 
intensity is decreasing with increase pressure because 
of its binder was not totally oxygen impermeable. 



Ratio of Normalized Intensities, Spec3 


1 20 
* ' 

( 


































-♦- Ratio Sp«c3| 




^..^_- 




' 


^ 








10 20 30 40 50 
PrMsur* (p«ia) 



Figure 12B. Ratio of PSP and NPSP normalized 
intensities versus pressure for spec. 3 

Moreover, in contrast for the specimen 1 , the NPSP 
intensities for the specimens 2 and 3 were decreasing 
with increasing pressure because their binders were 
not totally oxygen impermeable. Figures lOB, 11 B 
and 12B show that the calibration sensitivity is more 
linear for specimens 1 as compared to specimen 2 
and 3. 

Figure 13 shows the result of three repeated 
calibration of specimen 3 at 26 ° C. Replicates 1 and 
2 were done without the darkened ring around NPSP, 
but replicate 3 was obtained with a 6.25 mm 
darkened ring. Each replicate took about 3 hours to 
complete. All 3 replicates were performed over a 
period of about 30 hours. For specimen 3, both PSP 
and NPSP had the same lummophores but different 
binders. The excitation source filter wavelength was 
450 +- 40 rmi and the emission filter wavelength was 
580 +-9 nm. The NPSP intensity was found to 
decrease by 20 percent for 0.01 - 40 psia pressure 
change. Some oxygen permeability of the NPSP 
binder and some spectral leakage of intensity are 
believed to be the cause of this. The isolation of 
NPSP from PSP by the darkened ring is found to 
minimize the spectral leakage of intensity. In Figiire 
13, where the calibration plots are presented as the 
ratio of normalized intensities, there is no noticeable 
change in the paint performance between replicates 1 
and 3. 



American Institute of Aeronautics and Astronautics 



a: 

c 
« 

c 





Effect of Replicates, Specimen 3 












J. 










^-"^^^-^^ 


p 








.y^ 










^a^ 


r 










ji^ 




—♦—Ratio ropl _ 
-•- Ratio rep2 _ 






r^ 










J 










r 










0- 


■ 











10 20 30 

Pressure (psia) 



Figure 13. Effect of replicates on specimen 3 
calibration at 26 degrees C 



CONCLUSIONS 

The following conclusions are drawn from the results 
of the tests performed on six specimens of different 
PSP and NPSP, and camera combinations in a 
pressure-vacuum chamber over a pressure range of 
0.0-2.7 atmospheres, and temperature range of 15-35 
"C. 

(a) The technique works well when the PSP 
and NPSP have distinctly different 
spectral emissivity. 

(b) The repeatability of the calibration 
relation is good and the temperature 
dependence of the calibration is small. 

(c) The NPSP could be used as registration 
points in wind tunnel testing. 

(d) When PSP and NPSP have the same 
luminescence probe molecules 

(in a different binder), spectral 
leakage/interference problems occur. 

(e) None of the tested NPSP binders for the 
single-luminophore binary paint were 
completely impermeable to oxygen. 

(f) This method does not provide a true 
globalize intensity conection, but only 
localized correction. 



REFERENCES 

l.B. C. Crites,(1993), Measurement Techniques — 
Pressure Sensitive Pamt Technique, Lecture Series 
1993-05, von Karman Institute for Fluid Dynamics. 

2. K. S. Schanze,, B. F. Carroll, S. Korotkevitch, 
M. J Morris., Temperature Dependence of 
Pressure Sensitive Paints, AIAA Journal (0001- 
1452), vol. 35, no. 2 

3. V. E. Mosharov, V. Radchenko, and A. Orlov, 
Binary Pressure Sensitive Paint: A Lot of 
Problems, 7* Annual PSP Workshop, Purdue 
University, West Lafayette, LA, Oct 1 1-13, 1999. 

4. D. M. Oglesby, B. D. Leighty, B. T. Upchurch, 
Pressure Sensitive Paint With An Internal 
Reference Luminophore, Proceeding of the 41^ 
Meeting of the International Instrumentation 
Symposium, Denver, CO, May 7-1 1, 1995. 

5. J. Harris and M. Gouterman, Referenced Pressure 
Sensitive Paint, Proceedings of the Seventh 

International Symposium on Flow Visualization, p. 
802, edited by J. Crowder, Seattle, WA. 

6. A. Bykov, S. Fonov, V. Mosharov, A. Orlov, V. 
Pesetsky, and V. Radchenko, (1997), Study Result 
for the Application of Two-component PSP 
Technology to Aerodynamic Experiment, AGARD 
Conference Proceedings CP-601, Advanced 
Aerodynamic Measurement Technology, 29-1 to 
29-8. 

7. Liu, T., Cambell, B. T., Bums, S. P. and 
Sullivan, J. P., 1997, Temperature- and 
Pressure-Sensitive Luminescent Paints in 
Aerodynamics, Applied Mechanics Reviews, Vol. 
50, No. 4, pp. 227-246 

8. T. R. Amer, C. Obara, W. Goodman, B. Sealey, 
C. Burkett, B. Leighty, T. Carmine," Pressure 
Sensitive paint measurement in NASA-Langley 
Wind Tunnels," Proceedings of the 45* 
International Instrumentation Symposium, May 
1999, pp.325-334. 



American Institute of Aeronautics and Astronautics