Copyright ©1997, American Institute of Aeronautics and Astronautics, Inc.

AIAA Meeting Papers on Disc, January 1997A9715783, AIAA Paper 97-0772

Analysks of a Rayleigh scattering measurement system in a hypersonic windtunnel

Charles TylerUSAF, Wright Lab., Wright-Patterson AFB, OH

Franklin E. EastepDayton Univ., OH

AIAA, Aerospace Sciences Meeting & Exhibit, 35th, Reno, NV, Jan. 6-9, 1997

Rayleigh scattering, a nonintrusive measurement technique for the measurement of density in a hypersonic wind tunnel, hasbeen under investigation at Wright Laboratory's Mach 6 wind tunnel. Several adverse effects, e.g., extraneous scatter offtunnel surfaces and condensation of flow constituents, hinder efforts to obtain accurate Rayleigh scattering measurements.Overcoming some of these difficulties, measurements have been achieved while the Mach 6 test section was pumped to avacuum, as well as for actual tunnel operation for various stagnation pressures at fixed stagnation temperatures. Stagnationpressures ranged from 0.69 to 6.9 MPa at fixed stagnation temperatures of 511, 556, and 611 K. Rayleigh scatter resultsshow signal levels much higher-than-expected for molecular scattering in the wind tunnel. Even with higher-than-expectedsignals, scatter measurements have been made in the flow field of an 8-deg half-angle blunt nose cone with a nose radius of1.5 cm. (Author)

Page 1

AIAA-97-0772-

ANALYSIS OF A RAYLEIGH SCATTERING MEASUREMENT SYSTEMIN A HYPERSONIC WEND TUNNEL

Charles Tyler*Flight Dynamics Directorate

Wright LaboratoryWright-Patterson AFB, OH

Franklin E.University of Dayton

Dayton, OH

ABSTRACT

Rayleigh scattering, a nonintrusivemeasurement technique for the measurement of densityin a hypersonic wind tunnel, has been underinvestigation at Wright Laboratory's Mach 6 windtunnel. Several adverse effects, e.g. extraneous scatteroff tunnel surfaces and condensation of flowconstituents, hinder efforts to obtain accurate Rayleighscattering measurements. Overcoming some of thesedifficulties, measurements have been achieved while theMach 6 test section was pumped to a vacuum, as wellas for actual tunnel operation for various stagnationpressures at fixed stagnation temperatures. Stagnationpressures ranged from 0.69MPa to 6.9MPa at fixedstagnation temperatures of 511, 556, and 61 IK.Rayleigh scatter results show signal levels much higher-than-expected for molecular scattering in the windtunnel. Even with higher-than-expected signals, scattermeasurements have been made in the flowfield of a 8-degree half-angle blunt nose cone with a nose radius of1.5cm.

INTRODUCTION

Rayleigh scattering is a nonintrusivemeasurement technique which has been usedsuccessfully over the last several years for simple

'Aerospace Engineer, Member AIAA^Professor, Department of Mechanical and AerospaceEngineering, Associate Fellow ALAA

This paper is declared a work of the U.S. Governmentand is not subject to copyright protection in the UnitedStates.

reacting flows1, combustors2, external flameexperiments3 and subsonic freejets4. Recently theRayleigh scattering technique has been extended into thesupersonic regime with limited quantitative results.Condensation and cluster formation have been found tohave an adverse effect on Rayleigh scattermeasurements5'6. Various efforts in eliminating suchadverse effects and other unwanted biases, e.g.extraneous scattering off tunnel surfaces, have beenmade.

Although the construction of baffles for theelimination of stray laser beams has been found byothers to reduce background scatter, such constructs donot eliminate laser glare off a model surface completely.Recently, a dual-line Rayleigh scatter system, using thegreen and yellow lines of a copper-vapor laser , hasbeen developed and tested by Otugen, et. al.7 Thisparticular technique identifies and eliminates surfacescatter background noise. Other recent techniques,which measure velocity, temperature and density insupersonic and other high-speed flows include FilteredRayleigh scattering8'9 and RELIEF Imaging10. FilteredRayleigh scattering is an alternate way of eliminatingbackground scattering. The above mentioned techniquessolve the problem of surface scatter background noise;however, they do not address the problems associatedwith flow condensation.

The condensation of air in a hypersonic windtunnel takes place at relatively low supersaturationratios. This is possibly condensation taking place ontonuclei of water and/or carbon dioxide, which exist in theair as minor components, and which condense wellbefore the saturation of nitrogen or oxygen is reached.The condensation of the water and/or carbon dioxidedoes not seriously affect the stream properties becauseof the small percentage in which they exist. However,a large number of nuclei are formed in the condensationof these minor components, which then act as nuclei for

1American Institute of Aeronautics and Astronautics

oxygen and nitrogen condensation at low degrees ofsupersaturation. Throughout the past, a substantialamount of research regarding condensation insupercooled hypersonic flowhas been performed.'1-12'13-14

Recently, Shirinzadeh ct.al.15 performed experiments inthe 15-inch Mach 6 High Temperature facility at NASALangley Research Center with the results that, in theabsence of condensation, it is possible to obtainquantitative measurements of density using Rayleighscattering techniques.

Experiments performed in the Mach 6 HighReynolds Number facility at Wright Laboratory were ata lower stagnation temperature of 556K and higherstagnation pressures of 0.69MPa and 1.61MPa. Densityprofiles, for select locations on a 8-degree half-angleblunt nose cone, obtained by Rayleigh scattermeasurements were compared to CFD results withpromising results. The offset in the Rayleigh scattermeasurements caused by condensation effects wascorrected.

EOUIPMENT/FACILITY DISCUSSION

The Rayleigh scattering measurement systemlocated at Wright Laboratory's Mach 6 High ReynoldsNumber facility is shown in figure 1. A standardNd:YAG pulsed laser, producing a frequency doubled532nm beam, pumps two oscillator-amplifier, tunabledye lasers. However, the output from one dye laser isblocked. The unblocked dye laser is tuned such that thewavelength of the exit beam is 613nm. This exit beamis then frequency doubled using a BBO (beta-batrium-borate) doubling crystal resulting in a beam with awavelength of 306.5nm. As a result of the doublingprocess and optical configuration the red beam, 613nm,and the UV beam, 306.5nm, are collinear.

The beam passes through a gas reference cellcontaining a gas at a known pressure and temperature.The gas in the reference cell is comparable to the gasbeing measured in the wind tunnel. For this experimentthe gas is air at standard pressure and temperature. Thereference cell provides a means of eliminating pulse-to-pulse laser power fluctuations. Two photomultipliertubes (PMT) are mounted within the reference cell. ThePMTs collect red and UV scattered light, respectively,and converts the light into signals, referred to as the"reference" signals. The reference signal is used tonormalize the actual signal, referred to as the "sample"signal, in the wind tunnel's test section from smallvariations in the relative laser pulse energy.

After exiting the reference cell the laser beamis directed into the test section through a fused-silicawindow. Two 90-degree turning prisms steer the beamsuch that it enters the window on one side of the testsection and strikes the wall on the other side.

A light collection system consisting of a 3()cmfocal length planoconvex lens, bandpass filter and twoPMTs is mounted on a three-dimensional traverseoutside the test section. The angle of observation isperpendicular to the direction of incident light. Thereference and sample signals are collected by a dataacquisition computer system which consists of gatedintegrators, an A/D converter and a 286 IBM-compatiblecomputer.

The Mach 6 High Reynolds Number facility atWright Laboratory is a blowdown tunnel which usesdried, compressed air. The air is heated to 500-61 IK bya heater bed of stainless steel balls prior to entering thestagnation chamber. The tunnel has an axisymmetric,31.36cm diameter nozzle contoured to produce anuniform flow which has a calibrated center Machnumber of 5.76.16 The tunnel was operated over a rangeof stagnation pressures, 0.69MPa-6.9MPa in incrementsof 0.69MPa, at fixed stagnation temperatures, 511, 556,611K. For stagnation pressures less than 4. IMPa the airexhausted from the tunnel is directed into a 2,831m3

vacuum sphere; otherwise the tunnel is exhausted toatmosphere.

RESULTS

The main assumption of the Rayleigh scattermeasurement system at Wright Laboratory is thatRayleigh coefficients are proportional to air density. Inother words, there is a linear relationship between theair density and the intensity of the scattered light. Withthis understanding, intensity measurements were takenin a no flow situation at atmospheric and very lowpressure conditions; pressures and temperatures wereobtained from static probes within the test section.Now, since intensities at two known density conditionswere obtained, a line can be drawn which shows therelation between density and intensity. By using thisrelation the surface scatter background noise can bedetermined and eliminated,

(slope) = M = (p 2 - P I ) (1)

American Institute of Aeronautics and Astronautics

surface scatter=I1-M*p1(2)

To better see the linear relation between density andscattered light the test section was slowly pumped downto a very low pressure. Rayleigh scatter measurementswere performed during the evacuation process, trackingthe density of the air in the test section to its lower limitof 0.019 kg/m3. These tracking measurements can beseen in figure 2.

Since the Rayleigh scatter measurement systemtracks the tunnel density readings quite well, the testsection was rapidly evacuated to a low pressure. Whenthe test section is quickly evacuated water vapor presentin the non-dried, air supplied from the atmospherecondenses and fills the test section with condensationparticles. Condensed water droplets cause bothphotomultiplicr tubes, red and UV wavelengths, in thelight collection system to reach saturated levels, as seenin figure 3. When the tunnel is started and operating athypersonic flow conditions, neither PMT reachsaturation. This observation contributes to the beliefthat although condensation of carbon dioxide andpossible clustering of oxygen and nitrogen occur, largeparticles are not present during tunnel operation.

Although large particles apparently are absentfrom the flow during tunnel operation, higher-than-expected intensity signals were acquired when the tunnelwas operating at the aforementioned hypersonicconditions. As shown in figure 4, the intensity of thesignals diverge from the expected measurements quitedramatically for the lower stagnation temperatures.Even at the lowest stagnation temperature, 5UK, andthe highest stagnation pressure, 6.9MPa, condensation ofthe principle components of air, nitrogen and oxygen,should not occur; however, carbon dioxide, which existsin air as a minor component, condenses well before thesaturation of nitrogen or oxygen, see figure 5. Thecondensation of carbon dioxide does not seriously affectthe stream properties because of the small percentage inwhich it exists. However, a large number of nuclei areformed in the condensation of carbon dioxide, whichthen may act as nuclei for oxygen and nitrogencondensation. Even if the condensed carbon dioxidedoes not cause the primary constituents of air tocondense, the carbon dioxide condensation particlesformed still foul Rayleigh scatter measurements.

Under the belief that the condensation particles

formed by the rapid expansion of the gas from thestagnation chamber are small and may possiblysublimate traveling through a shockwave, measurementswere made on a 8-degree half-angle blunt nose coneinstalled in the Mach 6 facility, shown in figure 6.Measurements were taken on a line perpendicular to thesurface of the cone at a location of 12.7cm back fromthe tip of the nose. Figure 7 shows the comparisonbetween Rayleigh scatter measurements andcomputational fluid dynamic results. Close to thesurface of the cone the measurements agree quite well,while further off the surface behind the shockwave andinto the freestream it is apparent that the Rayleighscatter measurements are higher-than-expected.Although the quantitative results disagree, it isreassuring to see that qualitatively the two techniques,Rayleigh scattering and computational fluid dynamics,appear similar.

Preliminary efforts have been made to correctfor the offset in the Rayleigh scatter measurements onthe blunt nose cone. First, the offset in the scattermeasurements made during the empty tunnel operationwas calculated. Accounting for this calculated offset inthe raw cone data, scatter measurements can becorrected, as depicted in figure 8. Although thecorrected Rayleigh scatter measurements approachcomputation fluid dynamic results the accuracy of thetechnique requires additional work.

CONCLUSIONS

Extraneous surface scatter background noiseand scatter off condensation particles create difficultiesin using Rayleigh scattering as a measurement techniquein hypersonic flows. Fortunately, the surface scatterbackground noise has been eliminated by takingscattered intensity readings at two known densityconditions, obtaining the linear relation between denstiyand scattered intensity and calculating the level ofextraneous background scatter. The reduction ofcondensation particles effects also have been addressedwith promising results. Further work will be performedtoward additional reduction and possible elimination ofcondensation effects hindering Rayleigh scatteringmeasurement efforts.

ACKNOWLEDGEMENTS

The author thanks all the people involved withthe Rayleigh scattering instrumentation developmenttest. The author appreciates the effort of work

American Institute of Aeronautics and Astronautics

generated by the Mach 6 facility crew Joe Scheuring,Tom Morris and Rick Alien, as well as the pump housecrew Mike Green and Dwight Fox. The author wouldalso like to thank the members of the ExperimentalDiagnostics section for all their patience, constantsupport and helpful suggestions.

REFERENCES

1. Bill, R.G., Namer, I., Talbot, L. and Robben, F,"Density Fluctuations of Flame Grid InducedTurbulence," Combust. Flame 44, 1982, pp. 277-285.

2. Namazian, M., Talbot, L., Robben, F. and Cheng,"Two-point Rayleigh Scattering Measurements in a V-shaped Turbulent Flame," 19th Symposium onCombustion, The Combustion Institute, 1982, pp. 487-493.

3. Seasholtz, R.G., Zupnac, F.J. and Schneider, S.J.,"Spectrally Resolved Rayleigh Scattering Diagnostic forHydrogen-Oxygen Rocket Plume Studies," AIAA 91-0462.

4. Escoda, C. and Long, MB., "Rayleigh ScatteringMeasurements of Gas Concentration Field in TurbulentJets." AIAA Journal. Vol 21, 1983, pp. 81-84.

5. Shirinzadeh, B., Hillard, M.E. and Exton, RJ.,"Condensation Effects on Rayleigh ScatteringMeasurements in a Supersonic Wind Tunnel," AIAAJournal. Vol 29, 1991, pp. 242-246.

6. Shirinzadeh, B., Hillard, M.E., Blair, A.B. andExton, R.J., "Study of Cluster Formation and its Effectson Rayleigh and Raman Scattering Measurements in aMach 6 Wind Tunnel," AIAA 91-1496.

7. Otugen, M.V., Seasholtz, RJ. and Annen, K.D.,"Development of a Rayleigh Scattering System fortemperature Measurements," 4th InternationalConference on Laser Anemometry Advancements &Apllications, Cleveland, OH, 1991.

8. Miles, R.B., Lempert, W.R and Forkey, J.,"Instantaneous Velocity Fields and BackgroundSuppression by Filtered Rayleigh Scattering," AIAA 91-0357.

9. Elliot, G. S., Samimy, M and Arnette, S. A.,"Filtered Rayleigh Scattering Based Measurements in

Compressible Mixing Layers," AIAA 92-3543.

10. Miles, R, Lempert, W, Forkey, J., Zhang, B. andZhou, D., "Filtered Rayleigh and RELIEF Imaging ofVelocity, Temperature, and Density in HypersonicFlows for the Study of Boundary Layers, ShockStructures, Mixing Phenomena, and the Acquisition ofIn-Flight Air Data," New Trends in Instrumentation forHypersonic Research, NATO Advanced ResearchWorkshop, 1992, pp. 5G.1-5G.8.

11 Durbin, E.J., "Optical Methods Involving LightScattering for Measuring Size and Concentration ofCondensation Particles in Supercooled HypersonicFlow," NACA Tech Note 2441, 1951.

12. Stever, H.G. and Rathburn, K.C., "Theoreticaland Experimental Investigation of Condensation of Airin Hypersonic Wind Tunnels," NACA Tech NOte 2559,1951.

13. Daum, F.L. and Gyarmathy, G., "Air andNitrogen Condensation in Hypersonic Nozzle Flow,"ARL 65-169, March 1967.

14. Emmons, H.W., Fundamentals of Gas Dynamics.High Speed Aerodyanmics and Jet Propulsion. Vol III,Princeton University Press, Princeton, New Jersey, 1958.

15. Shirinzadeh, B., Balla, R Jeffrey and Hillard,ME., "Quantitative Density Measurements in a Mach 6Flow Field Using the Rayleigh Scattering Technique,"International Congress on Instrumentation in AerospaceSimulation Facilities, IEEE 95-CH3482-7, July, 1995,pp. 13.1-13.7.

16. Fiore, A.W. and Law, C.H., "AerodynamicCalibration of the Aerospace Research LaboratoriesM=6 High Reynolds Number Facility," ARL TR 75-0028, Feb, 1975.

American Institute of Aeronautics and Astronautics

Fig. 1. Schematic of the Rayleigh scatteringmeasurement system.

TUNNEL DENSITY, kg/1113

Fig. 2. Rayleigh scatter measurement systemtracking of tunnel density.

O 400

O

200

.... 306.5 NANOMETERS

613,0 NANOMETERS

0 100 200 300 400 500 600 700 800 900 1000

DATA POINT NUMBER

Fig. 3. Capture of rapid test section evacuationprocess.

3S 10 IS 50 555TRGNOT10N PRESSURE, FtTH

Fig. 4. Rayleigh scatter measurements for variousempty tunnel conditions.

1SENTROPICEXPflNSION H20

N2

STflG CONDITIONSx - 17.6 BTM, 522.2» - 17.6 RT«, 555.6' - 17.6 flTM, 6U.1

Id'TEMPERHTURE IK I

Fig. 5. Saturation lines for water, carbon dioxide,oxygen and nitrogen.

American Institute of Aeronautics and Astronautics

o

S«s.tt

3.ft"'o5

S

O

^

t\

•LS

^

b v

1,

/-

\r J

•

ft' •* — —_. ...-"

t.

Q

>

.--

if

v •^

Vt

RESULTS0 . COMPUTATIONAL FLUID DYNAMICS0 - UNCORRECTED SAYLEIOH SCATTER .« - CORRECTED RAYLEIGH SCATTER

1 1 1 1 J-

DENSITY. KO/«3

Fig. 8. Corrected Rayleigh scatter data for4.83 MPa stagnation pressure condition.

Fig. 6. 8-dcgrec half-angle blunt nose coneinstalled in the Mach 6 wind tunnel.

1

i LJi0.5

0(

_

T"̂

?;^f±

t-"/

_

m\\

J>T

^

0

(

,7f

/

^

* -̂

-/

_

1

^-~

-

:0

u- -

:

4

~

D

—

[

IN

-

^= — H*r •*

~~

^"

-»-CFDFOR4.83MP«-»- RAYLEIQH TOR 4 ,83MP|-*-CFDFOR2.0IMP«-*- RAYLEIGH FOR2.01MPa

0.6 0.8

JllY.ka'mJ

_

Sj

_. ...

i.

Fig. 7. Comparison of Rayleigh scattermeasurements and CFD results.

American Institute of Aeronautics and Astronautics

Copyright ©1997, American Institute of Aeronautics and Astronautics, Inc.