STECHN I CAL REPORT ARB RL-TR-02336

-:C I LASER-EXCITED OPTO-ACOUSTI C

¥ PULSES IN A' FLAME

•- William R. Anderson• CL ~~John E. Allen, Jr. - •o

•_• ~David R. Crosley •

,o .- June 1981 .

17==1

USb ARM ARAETRSARHADDVLOMNOMN

BALLISTIC RESEARCH LABORATORY

4- .PULSDES POING A RUD FLAMEAN

Apr vedfrpbicrlease K.Adesrsbton ulmtd

10 n.;

Destroy this report when it is no longer needed.not return it to the originator.

3. .4

Secondary distribution of this report by originatingor sponsoring activity is prohibited.

Additional copies of this report may be obtainedfrom the National Technical Information Service,U.S. Department of Comerce, Springfield, Virginia22161.

7 I

The findings in this report are not to be construed asan official Department of th.e Army position, unlessso designated by other authorized docunnts.

I-

Ahe U" of trad *MS or [awactimpm nsc in tkt ort

'II dc~~~ not .mstituut indoi,.wt of my conwreialpoae.

M.-

I I . !N .- -. - -

UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE (When Data Entered)

REPORT DOCUMENTATION PAGE pRD INSTRUCTIONSBEFORE COMPLETI•G FORMd

1. REPORT NUMBER GCVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER

TECHNICAL REPORTARBRL-TR-_02336 V -

4. TITLE (and Suzlbtle) P ULS APE OF REPORT & PERIOD COVERED

LASER-EXCITED OPTO-ACOUSTIC PULSES IN A FLVIE- _ _BRTechnical __eport.t/ " iG. P- REPORT N•UMBER

l:-.-- 3•R.••-. .CONTRACT OR GRANT NUMBER(.)" William R.!,AndersonJohn E.jAllen, Jr.*David R.1 Crosleyf

-9. PERFORMING ORGAkIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASKUS Armv Armament Research and Development CAREA m ORK UIT NUMBERSUS Arm. Ballistic Research LaboratoryATTN: DRDAR-BLAberdeen Proving Ground, MD 21005 - 1L161102AAH431I. CONTROLLING OFFICE NAME AND ADDRESS 1227RE T"rDATE

US Army Armament Research and Development Command / /3UNE 1981US Ari, Ballistic Research Laboratory 3" - - F PUAGESATTN: DRDAR-BL 424. MONITORING AGEN'Ir NAME & tDORESS(Q1 dtf*.fl;rwnt frm C.ntrollhaj Office) 1: SECURITY CLASS. (of WO repot")

(

UNCLASSIFIED15s. DECLASSI FICATION/ DOWNGRADING-

SCHEDULE

1I. DISTRIBUTION STATEMENT (of 'da Repoot)

Approved for public release; distribution unlimited.17. DISTRIBUTION STATEMENT (of the abstract entered In Block 2D. If different from Report)

15. SUPPLEMENTARY NOTES*NAS-NRS Resident Research Associate; present address: Goddard Space Flight

Center, Greenbelt, 'MD 20770"**Present address: Mblecular Physics Lab, SRI international, Menlo Park,

CA 94025It. KEY WORDS (Coil inue an revmere side it necosareinvd lda iefy b- block ===ber,

toacoustic effectombustionluorescence

Alkali atomsQuenchin-

26 fWRA? (~ -seW N '~ ~ y ~(clt-jmlk)

? Pulsed pressure waves are produced by energy tran.1fer from electronicallyexcited Na or Li atoms seeded in an acetylene-air flame; the excitation isprovided by a resonantly tured, pulsed laser. Temperature increases ofapproximately one degree within the flame produce signals easily detected by amicrophone. A description of the phenomenon is given, and preliminary resultsj are presented for two possible applications: lqcalizee speed-of-sound'Weasurements and auenchina rat• dptp1-minmtinnq

•i(Xi

DD IM "mo w o 3 ED88 E UNClASSIFTED

/SICUMT-' CLASItCA1?OI OF ThIS PAGE (1110 Dat Sn~sreE)

It

TABLE OF CONTENTS

E •Page

LIST OF ILLUSTRATIONS ....... ................... .

I. INTRODUCTION ........... ....................... .

A. Laser Probes and Combustion Chemistry ............. ...

B. Cpto-acoustic pulses .. .. ....... . .......... ..........

II. EXPERIMENTAL DETAILS ....... ...................... 9

A. Apparatus ........ .......... ..... ........... 9

B. Qualitative Observations ......... ............... 15

C. Magnitude of the Pressure Wave .... .............. is

D. Form of the Pressure Wave ...... ................ .. 16

III. APPLICATIONS AS A PROBE ...... .................. ... 19

A. Concentration Measurements ......... .............. 19

B. Temperature Measurements: Speed of Sound ........... 23

IV. QUENCH PLATE ME_ASURE1%FEN7S .. .. .. .. .. .. .. . .. 26

- V. 4P LEVEL PLIMPING ......... ..................... .... 29

VI. CONCLUSIONS ............ ........................ 3.

37REFERENCES ................ ........................ .

39DISTRIBUTION LIST .......... ..................... .

S.•- 7"Sc /*

j =' .. .: .. . . _

Spec

-Rr -A"

-ll • .•,,=

LIST OF ILLUSTRATIONS

Figure Page

1. Experimental arrangement. The laser scanner, and aspectrometer often used for fluorescen e observations, arenot shown . ........... ..... .......................... I 0

2. Oscilloscope trace of an opto-acoustic pulse. The timescale is 20 psec/cm, and the trace is triggered as the laserfires. Data such as those shown in Figure 2 are obtainedfrom the difference between first two extrema, using thedual channel boxcar. The later oscillations with smaller

Z' amplitude are quite reproducible from pulse to pulse forfixed geometry and stable flame conditions ............ .

3. Opto-acoustic pulse signal as a function of laser• frequency . .. .. .. .. .. .. .. .. .. .. .. ... 13

4. Output of the ,,unitoring photodiode as the laser (with 6 cm-bandwidth) is scanned through the region of the D-lines.Zero corresponds to no laser intensity, so there is aboutfifteen percent absorption here ..... ............... ... 14

S. Dependence of pressure wave amplitude on flame-to-microphonedistance r ........... ......................... .... 17

6. Oscilloscope traces of opto-acoustic signal waveform, fox 4different geometrical arrangements. Each results from asingle laser pulse. Horizontal axis: 20 usec per majordivision ........... ..... .......................... IS

7. Microphone signal as a function of beam diameter (related tothe optical saturation studies) showing that microphones ofthree different diameters yield the same relative response . 20

8. Dependence of the microphone signal on laser power, at powerdensities much lower than saturation values .................

9. Dependence of the microphone signal on Nal concentration;presumably the Na concentration in the flame follows thesolute concentration ............... .................. .

10. Schematic illustration of the measurement of the speed of

sound. On three different laser pulses, the laser beam isfired into the flame at three different positions, and thearrival time of the pressure pulse at the microphone ismeasured ............. ....... ........................ 24

5iMME1JG PAGE

LIST OF ILLUSTR-ATIONS (Cont'd)

Figure Page

11. Schematic illustration of a system near optical saturation. EITwo level models are illustrated in (1) and (2); it is thesewhich are pertinent to the treatment in the text. The Almultilevel model (3) requires the incorporation of relaxationrate parameters ........... ......... ........... ..... 27

12. Plot of M-1 vs. I-1, in the form of Eq. (4), for low laserpower (large beam diameter) ............. ............. 30

13. Plot of M-1 vs. 1-1, in the form of Eq. (4), for high laserpower (focussed beam) ........... .................... 31

14. Inverse of the pressure-wave amplitude versus inverse ofthe spectral-power density. Error bars represent estimatesof the readability of the boxcar output. The line is anunweighted least-squares fit ...... ................ 32

|ZI: **.

t•I

U 5-u•. ,.

I. .-IRODUCTION

A. Laser Probes and Combustion Chemistry

The application of lasers to the study of combustion processesoffers a wealth of information of diverse kinds. Many of these newtechniques - in particular laser-excited fluorescence, ordinary Ramanscattering, and coherent anti-stokes Raman spectroscopy - are aimed atthe measurement of concentrations of individual molecular species atpartial pressures ranging from several atmospheres down to 10-10 torr. 1

The use of scanning lasers permits one to map out energy level populationdistributions within ground electronic states, with an ease and accuracy

previously possible only for excited states. From such information canbe obtained temperatures corresponding to different degrees of freedom.Measurement of laser absorption line-shapes within flames provides data

on energy transfer rates and translational temperatures. The recentsuccessful observation of two-photon excitation of Na in a 1-atm pressureflame2 suggests numerous possibilities for probing those species nowinaccessible by present-day cemmerciallv available lasers with one-photonexcitation.

The use of the full arsenal ot trhse laser techniques so fardemonstrated would prov.ide a nearly complete experimental characteri-zation of a process undergoing combustion. Of course, the use of lasersmeans, almost necessarily, a high degree of spatial resolution:focusing to 100 p beam diameter is generally easy. Pulsed lasers ofvarious types yield concomitant temporal resolution down to severalnsec. Conseouently spatial profiles, over time periods up-blurred by Aturbulence, can be obtained for these quantities of interest. It should

be noted t:at, in addition, laser probes are non-perturbative. T:.at is,they introduce no physical barriers into the gas flow or provide surfaceswhich could cause heterogeneous catalysis influencing the chemistry.Although th; presence of particulate matter can be a serious problem,extremes of temperature and pressure offer no hostility to the laserbeatE itself.

The field likely to benefit most from this detailed picture isthat of con-bustion chemistry. The information on transient, reactivespeci.s will provide the key to the mechanistic chemical kinetics hereto-fore usually only inferred fr om the composition of finalprdcs

'A. C. Eckbre-th P. A. Boncziik and J. F. Verd.:ý_eck, 'W-j of LaserRaran and Fluorescence Technicues for F.iracticaZl ComrimaZon agrostics",United Technologies Research Center Report R77- :2665-6, Eartford, CT,Febaar-an 1977.

2J. E. Allen, Jr., W. R. Anderson, D. R. Croslez and T. D. Fansler,"DEnergy TY7ansfer and Quenchirn Rates of Laser-Pun -ed etoniealUyExcited Alkalis in Fla,,cs" Seventeenth SýposizCi (international)on Combustion, Leeds, EkgZand, August 1978.

7!-

Data on state distributions will directly address the questions of

disequilibrium in combustion systems, widely recognized as existingbut seldom accounted for in mechanistic schemes. In fact, this combi-nation of experimental probes using laser-based techniques togetherwith tneoretical chemical kinetic modelling on modern large computershas been heralded3 as promising - finally - significant advancementin the enormously complex problem of the chemistry of combustion. Anunderstanding of the chemistry pertinent to ballistic systems willundoubtedly require such models tied in detail to copious laser-basedexperiments (as well as measurements with other techniques) under con-trolled laboratory conditions and proven by agreement with the neces-"sarily more limited data available on real systems.

Nearly all of the currently available laser probe methods rely onthe scattering or emission of light as the mode of detection. Whilethis can be a highly sensitive method under many conditions, there are

other situatiuns in which it is less suitable. For example, highexcited state quenching rates reduce the degree of emitted fluorescence,and Raman-scattered photons could bo absorbed by some interferingmolecular species. One recent alternative method is the optogalvaniceffect. 4 in which the yield of collisionally produced ions is enhancedby promotion, using laser excitation, to excited electronic states lyingnearer to the ionization continuum than does the ground state.

B. Opto-acoustic pulses

lie describe- here the observation and characterization of anothermeans of detecting the occurrence of selective absorption of tuned laser

radiation. This is the production of a pulsed opto-acoustic effect --pulsed sound waves -- within a flame operating at atmospheric pressure,following deposition of tne lase: energy into an electronic transition.

The flame is seeded with alkali atoms and a pulsed laser is tunedto the appropriate absorption line. While some of the electronicallyexcited atoms fluoresce, the vast majority lose their energy by colli-sion with other gases present in the flame. This energy ultimately(though rapidly) is converted into translational kinetic energy of theflame gases, producing a pulsed pressure wave which expands from theregion illuminated by the laser. Quantitative measurements are madeusing a microphone, and show that the technique is very sensitive and"accurate, and perturbs neither the gas dynamics nor chemical kineticsof the flame. Because the phenomenon itself depends on the rapid

'D. HartZey, A. L<-z and D. Hardesty, "Physics in Combustion Research"J,Physics Today, December 1975, 36-4-7.

R. B. Green, R. A. Keller, P. K. Schenck, J. C. Travis and C. G.Luther, "Opto-Galvanic Detection of Species in Flames", J. Amer.Chem. Soc. 98 17-8518 (1976).

8517- - -7

-; 4 - -

collisional conversion of electronic to translational energy (throughthe intermediary of molecular vibration), the sigmal is thus not reducedat high pressure as is the fluorescence. The results have provideddirectly useful information as well as projections concerning thesensitivity of the technique.

The opto-acoustic effect in general, viz., the conversion of opticalto acoustical energy through absorption and collisional quenching, haslong been known as a means of sensitively detecting the absorption ofradiation. What is newly demonstrated here is the nature of its occur-rence following pulsed excitation, and the existence of the effect underin s-tu conditions on a I-atm pressure fLame burner (in contrast to thetypical mode of operation asing specially constructed cells at lowpressure).

In addition to exploiting its detection capabilities, we have alsoutilized the effect to perform meazurements cf the speed of sound withinthe flame on a spatially resolvej basis. These data provide information

r on the translational temperature of the flame gases. This aspect mayprove especially useful for environments t.o hostile to admit thermocoupleprobes or processes too rapid to obtain usable thermocouple response.

The actual mode of formation of the pressure wave -- the fluiddynamics aspect -- is not investigated in any detail in these experi-ments. Observations have been made of the form of the pressure waveproduced; it appears to be dictated by density gradients in the flame andsurrounding atmosphere. Some modification of the experiment might per-mit the waveform to be used to extract bulk energy transfer rates, butwe have not pursued that end.

Most of the experiments to be described were carried out pumpingthe 3P state of Na. We have been dnable to detect opto-acoustic pulseformation when pumping the 4P level, aid we conclude that the net effec-tive conversion of electronic to translational energy, at least over thesalient time scale, is less efficient for this level. In conjunctionwith this experiment, we also consider some of -he results on energytransfer pathways among excited states of \a in the flame, obtainedusing multiphoton excitation of fluorescence.-iI

1i. EXPERIMEN-TAL DETAILS

A. Apparatus

The experimental arrangement is depicted in Figure 1. The flame isa fuel rich mixture (1.5 times stoichiometric' of C2H2 and air, and theslot burner is a standard type used for atomic absorption analysis. Na is

5IH .a Y. Paor (ed 97adm.H. 7. Pao (ed.), GOtoaccustic Sý?ectroscopy and Detection, AcadeicPress, New York, i977.

I3ý

cr..

Cf..

L zL

00

00

o r=

LUU

LU

C4'4

cl-

LUILae- 0

Er- 424C >.

0

co- - ---u------.-

introduced by aspiratlon of Na! sob:tion; :-,any of the experiments werecarried out at a Na concentrati:'n of the order of 1010 atoms cm- 3 ,as determined from both absorption and absclute fluorescence measurements.Additionally, a few experir-nts have been run using Li as the seed atom,or with Na in a Cl 4 I/ar flame. The experiments afe carried out atheights ranging from 0.F to 3 zm above the surface, that is, in the regionof the secondary reaction :one containing partially combusted gases.

The laser used is a Chrcnatix CMX--, a com.mercially available flash-lamp pumped tunable dye iaser having a pulse duration of 1 psec. Al-though the laser is capable of delivering 10 mJ/pulse in the region ofthe Na resonance lines, it was here operated typically at < 1 mJ/pulse.The bandwidth is nominally 3 cm- 1 and can be narrowed to 0.16 cm-4 byinsertion of an etalon. (In a separate experiment, not further describedhere, opto-acoustic pulses have also beer observed with a Na-laser-pumped-

*dye laser. The 5 nsec pulse lengths available with this instrument offeruseful versatility with the techniques).

The laser is focused to a few mm beam diameter and directed intotne flame at a spot 'x, 2 cm above the burner surface. When the laser 1Stuned to either of the two components of the 32p-32S transition in Na,the energy released in the quenching of the resonantly excited 3P stateproduces a pressure wave which propagates outward from the region illum-inated by the laser. This sound wave is detected by a condenser micro-phone located nu 4 cm from the flame. After amplification, 'he microphonesignal is fed to an oscilloscope and to a boxcar (gated) integrator,both of which are triggered by a Pulse from a photodiode detecting thelaser pulse itself. Figure 2 shows an oscilloscope photograph of asingle pressure wave.

In Figure 3 is exhibited a recorder tracing of the amplitude of thepressure wave signal vs. laser frequency. The laser is operote' here inits narroied mode and automazicali. ý:.nnel 'ver the wavelength reg-cnwhich includes both doublet comnonents. 'The -K';.es marked 'etalonreset are artifacts of the scanning sequ....e.. It is clear that the

pressure waves result fro- the reSa-aia electronic excitation and net

from some other mode of laser energy denosition within the flame gases,such as laser-induced breaK'.in. of the flane -ases.

The laser power was usually monitored by a laser calorimeter behandthe flauc. An alternate method •n~'i.eu spiatting cff a small portionof the beam before its entry into the flame and directing it to a spec-trometer outfitted with a photomultiplier. These detectors also servedat times for measuremcnt of absorption and -!uorescence. In addition,the optogalvanic signal, also in the form of pulses, was occasionallymonitored. All four phenomena - absorntion, fluorescence, optogalvanicpulses, and optoacoustic pulses - occur simultaneously when the laseris tuned to a resonance. Figure 4, shows a -measurement of the laserpower incident upon the photodiode as the laser is scanned across the

D-lines at large (6 cm-I) bandwidth. (This run was made at very high

'1

'-a

M.O MEMO

EE-E

Figure 2. oscilloscope trace of an optic-acoustic pulse. The time scaleis 20 ý.sec/cm, and the trace is triggered as the laser fires. Data such 4

as those shown in Figure 2 are obtained from the difference betweenfirst two extreina, using the dual chann'el boxcar. The l atLer oscilla-tions with smaller amplitude are quite reproducible from pulse to pulsefor fixed geometry and stable flame conditions.

12a

nnunmuum

04

'0i

10-

<uJ 4-2

Ln

0'

LUU

C4C

0.=

4 -r

)4P)

SC,

4-a.

LU2

0Si

r-4-

100

0 -4 0

14-

EC4c

Na density). From these runs, we can measure the amount of laser energyabsorbed.

Not shown in Figure I is a spectrometer, with associated optics and

photomultiplier, which was used to measure the fluorescence from the Na oparticularly in those experiments 2 involving ultraviolet and two-photonexcitation.

B. Qualitative Observations

When the laser is tuned to one of the resonance lines, the pulsedpressure wave is produced. At relatively high sodium density, so thata hundred microjoules or more of laser energy are absorbed, this waveis readily audible to an observer located within a few meters of theburner. In fact, the normal procedure for coarse manual tuning of thelaser is to locate the resonance lines by listening for the sound wave.

As measured by the microphone and displayed on the oscilloscopethe waveform shows a sharp rise as the sound wave reaches the detector,followed by a series of oscillations typically diminishing in magnitude.(See Figure 2). The two boxcar gates were set to correspond to the firstminimum and first maximum of the waveform; the difference between thesevalues was output to the recorder and constitutes the 'microphone signal'used below to denote the magnitude of the pressure wave.

C. Magnitude of the Pressure Wave

From the information directly available from the experiment, we canonly speculate on the mode of forma-tion of the pressure wave. Collisionalquenching of the excited state transfers the Na electronic energy intovibrational levels of the flame gases (as demonstrated by an experimenton electronic-to-vibrational-to-electronic energy transfer between Naand Li in the same flame 2 ). This energy is then rapidly converted totranslational energy, resulting in translational heating of the regionilluminated by t.'Ie laser b~eam. Because sound travels only about 1 Mduring the 1 usec heating time, and the laser bear is typically a littlelarger than this, we envision the energy deposition as resulting in aslightly hotter region where the laser beam has passed through the flame.A small shock wave may be initially formed; if so, it soon degrades intoa pressure wave, and propagates outward through the flame.

We describe a particular, though t)pical, experiment to measure theabsolute laser energy deposited and the resulting pressure wave amplitude.The ýa absorbs 0.22 WJ from a single laser pulse, as the beam traversesa 1 cm path through the flame, with a diameter of 15 mm. Over 99% ofthe excited atoms are quenched by collisions under our conditions, sothis is the amount of energy transferred into translation.

The flame gases are assumed diatomic here, with a density appropriateto 1 atm and the flame temperaturc (fron speed of sound measurements

1.3

__i I. I - -J I- ] i IjA--

described below) of 2300*K. The irradiated volume of 1.75 cm3 thus hasa heat capacity of 0.34 mJ/degree and is heated 0.64*K. This is a trulynegligible perturbation compared-with the temperature of the flame gases

*themselves. A pressure rise of 0.22 torr within the heated volume iscalculated using the ideal gas law.

At the microphone, the amplitude of the resulting pressure -avewili be lower, because of the outward expansion of the wave. In theseexperiments, it is the pressure amplitude (not energy density) which ismeasured. A fall-off inversely proportional to the distance T betweenflame and microphone is thus expected. Fig.re S shows a plot of themicrophone response vs. T- 1 , verifying this dependence.

We can then calculate the amplitude anticipated at the microphone.For the geometry used here, an amplitude of G.044 torr would result atthe microphone. The measured value, using a calibrated microphone,is 0.13 torr. In view of concerns about the homogeneity of the laserbeam over its spatial profile, some uncertainties in the measurementof the path length, and especially the lack cf knowledge concerning theactual -mode of formation of the pressure wave, we consider this to bereasonable agreement. confirming our simple physical picture.

D. Form of the Pressure Wave

The waveform of the sound wave is quite complex and we have examinedit extensively, although no quantitative information could be extracted.As noted above, it is quite reproducible from pulse to pulse for a fixedgeometry of laser/flam-e/microphone. Ilf, however, this geometry ischanged, considerable differences result. Figure 6 shows four oscillos-cope traces for fcur different geometrical arrangements. The scope istriggered on the left by the photodiode, and the time scale here is 20;isec/cm. (The amplitude referred to throughout this paper is measuredby setting the two boxcar gates to sample the first positive and negativeextrema of the waveforms, and taking the difference.) The characteri-ticsof the waveform, especially the smaller-amplitu-e oscillations followingthe first large wave, can be readily varied by introducing disturbances,particularly turbulence, into the flame. This implies that the form isdictated by reflections and/or interference from density gradients with-in the flame, though no analysis was attempted. Perhaps in a more care-fully controlled experiment probing those gradients, useful informationcould be obtained from the waveform.

It was originally hoped that the waveform might provide informationon the bulk energy transfer rates within the flame. Now, a sound wavehaving a frequency of the order of the inverse of some relaxation time(e.g., a vibrational energy transfer rate) within the system will undergoattenuation and phase shifts due to that relaxation process. Vibrationalrelaxation times at 1 atm are of the order of the laser pulse duration.The laser pulse thus is expected to produce a spectrum of sound frequencies,and it was hoped that a Fourier analysis of the opto-acoustic waveform

ii.

%* ~ ~ .~** -,

12 -

10

8

V A

6

4-

2

0 0.1 0.2 0.3r-1 ( cm-1 )

Figure 5. Dependence of pressure wave amplitude on flame-to-nicrophonedistance r.

17

4.. -. ,• i,.•ll I i iE LKIIm ~ II g''''

-•. I 1I1;., 1 I, I I

--- ! !- ---[ ---

J j "0- -' - __.._ __ _ _ 4J h

r -- ±

,-

I .,.. .... 30

C €4

ft -- u•,C

.0.

'o .

might yield information on these relaxation times within the flame.This approach was not successful, due to the dominance of the densitygradient effects and the bandwidth (140 kHz) of our fastest microphone.

"Air" mixtures of N2/02, Ar/02, and C02/02 were tried in this seriesof experiments, but no differences could be discerned. With a fastermicrophone it remains possible that this will be a useful applicationof this technique.

Even though the vibrational relaxation could not be directly1 _-observed in this manner, it is safe to conclude that the energy istransferred through the internal levels of the flame gases. N2, whichconstitutes some 60% of the total gas flow, has long been known to haveia very high quench rate for electronically excited alkalis, especiallyS= = • : a.

In a short series of experiments, three different microphones ofvarying diameter were used in order to ascertain whether the frequencyresponse (directly related to the microphone diameter) affected therelative signal size when other parameters were varied. Figure 7 showsa plot of microphone signal vs. beam diameter (the significance is relatedto the quenching/saturation studies discussed below). It can be seenthat the microphone diameter does not affect relative response for theseexperiments.

ili. APPLICATIONS AS A PROBE

A. Concentration Measurements

It has already been demonstrated that the formation of the opto-acoustic pulses introduces a negligible perturbation into the flame gases,insofar as either the chemical kinetics or gas dynamics is concerned.Here we consider the use of the technique as a sensitive probe of selec-tive absorption of laser radiation, i.e., as an alternative or complementto laser excited fluorescence.

Our simple picture of E-V-'\T transfer and local translationalheating suggests that the pressure wave amplitude should be proportionalto the laser power and to the concentration of the Na atoms. Figures8 and 9 show the dependence cf the microphone signal on these two para-meters. The power dependence is measured at a sufficiently low powerdensity so as to be in the linear absorption region, that is, we do notapproach optical saturation in Figure 8. In Figure 9, the microphonesignal is plotted against the concentration of the NaI solution aspirated

t into the flame. The nonlinear curve-of-growuth form of the plot reflectsthe effect of increasing optical thickness. The results shown here demon-strate that we may scale our observed signals to lower Na concentrationsand, with due regard to optical saturation, to higher laser power.

l

19-'=

z

Uj

Cr.

SI-I 4134 1/2"! 0 4136 1/4"

Z,4138 1/8"

S H M NORMALIZATION

MINIMUM

2 4 6

•jBEAM DIAM ETER

Figure 7. Microphone signal as a function of bean diameter (related tothe optical saturation studies) showing that microphones of threedifferent diameters yield the same relative response.

20

I InI -_J

o /~

T/

1.0

Lo

IIi0

UCL

~~0z0.

0 1 2 3 4 -

LASER POWER (RELATIVE)Figure 8. Dependence of the microphone signal on laser power, atpower densities much lower than saturation values.

21

°4>-- .. -, •

4T

-. II

3LU

z0

I'

U

z0

0

I ~1-

0 .05 .10 .15 .20No I CONCENTRATION (molesilitre)

Figure 9. Dependence of the microphone signal on KaI concentration;.presumably the Na concentration in the flame follows the soluteconcentration.

22

i iI-•

a- II

From these guidelines, we may project from our observed signal sizesand background noise a detectability limit for Na. We estimate that witha time constant of about 1 sec (corresponding to averaging over 30 laser

Z• :pulses), a signal-to-noise ratio of about 2:1 should be obtainable forI Na densities of the order of 107 atoms per cm3 in our 1 atm flame."t Although this is not as low a limit as one can obtain using laser-excitedi]" I fluorescence for this species, opto-acoustic pulses nonetheless do con-

stitute a sensitive detection technique. It should be emphasized that

we have undertaken no effort to shield the microphone from extraneoussources (pumps, etc.) of background noise; perhaps with two microphonesand differencing techniques the noise could be considerably reduced.

Although at higher total number densities, the quantum yield offluorescence is reduced due to the higher frequency of collisions, theopto-acoustic signal should remain essentially constant. Consequently,this may be the method of choice for some experiments carried out athigh pressure.

In addition, there may be some systems which exhibit selectiveabsorption of laser radiation but which do not readily fluoresce. Forexample, a triatomic or larger molecule may undergo rapid intersystemcrossing into a state of different multiplicity from the ground statebefore fluorescing. Collisions then would remove the energy of excita-tion before any long-lived phosphorescence would occuz. However, thecollisional transfer would still result in the formation of a pressurewave upon conversion into translational energy. Another molecular situ-ation might be a case in which the excited molecular state predissociates:pressure waves would be produced upon resonant tuning provided that theabsorption lines are still relatively sharp and not all of the absorbedphoton energy goes into breaking the chemical bond.

B. Temperature IMeasurements: Speed of Sound

The pulsed sound wave formation offers the ability to measure thespeed of sound, us, within the flame; us can then be related to the trans-lational temperature, providing an important means of characterizing theflame.

To measure the speed of sound in sirru, the laser beam is fired intothe flame at a particular point, as illustrated in Figure 10. The scope(or boxcar) is triggered at -he time of the laser pulse, so that thearrival time of the pressure -ave at the microphone can be accuratelymeasured. The laser beam is then moved within the flame (a 1 cm shiftis indicated in the figure). The difference in arrival times for thedifferent points, together with the distance between them, furnishes thespeed of sound within the flame. (Although the measurements carried outwere made by physically moving the beam, a better method would be tosplit the beam into two or three components and obtain the informationfrom one laser pulse, thus adding time resolution.)

II

amm.

A~ n,

Lu LLU

0

I S- -

E C

0 0V

- QJC

UUISzl

coo

-S-- 4.)W

Cý

cm an

TweL

The measurements were made at three different heights well intothe region of partially burned gases. The results were all the same,(9.7 ± 0.5) x 104 cm sec-1, within experimental error. The uncertaintycould be reduced by a more careful measurement of distances than wasdone for this demonstration experiment.

For a gas of average molecular weight R and average heat capacityratio �y, the speed of sound u is related to the translational temperatureby

Us RT/f] 1/2 (1)

From the stoichiometric composition of the initial flame gases, and

reasonable assumptions on the degree of combustion, average values werecalculated for M (molecular weight) and y (heat capacity ratio). Sincemost of the flame gas is made up of N2, this is not a critical pair ofnumbers. These values are then used to calculate the temperature usingEq (1). The result is 2280 ± 230'K, which may be compared to theadiabatic flame temperature for this flame 6 of 25450 K.

The translational temperature is a most important flame character-istic, from the standpoint of both the gas dynamics and the chemicalkinetics. The opto-acoustic pulse method offers the possibility of 4measuring the closely related speed of sound on a basis of high spatial

and temporal resolution. No perturbation such as a thermocouple need beintroduced into the flame, and extremely high temperatures (which arehostile to non-optical probes) appear to pose no fundamental difficulties.Such speed of sound measurements could well form the unique contributionof this method as a combustion diagnostic tool.

Of course, by carrying out an excitation scan over a series ofmolecular absorption lines or bands. using opto-acoustic pulses orfluorescence for detection, one can measure population distributions

corresponding to the temperatures of internal degrees of freedom as well.

In addition, the optoacoustic pulse method furnishes the abilityto generate pulsed sound waves locally within the flame, so as tostudy- their propagation through the flame gases, the interface with theambient environment, and the surrounding air. This may be of utilityin studies of noise pollution caused by combustion.

6A. G. Caydon and H. G. Wolf'hard, Flames: heir Stucture,, Radiation

and Temeratue 3rd Ed., Chapman and Hall, London, 19?0.

...........

*'1

IV. QUENCH RAIh MEASUREMENTS

Descriptive Equations:

f Under conditions of high laser intensity, one may create an appre-ciable population in the electronically excited state. This will be

the case when the rate, per atom, for absorption of laser light becomescomparable to the rate at which the atoms exit the upper state. Then,the number of excited atoms is no longer linear in the laser intensity.

With a focused laser beam, it is easy to obtain energy densities

approaching saturation of the optical transition in Na, even for thehigh quenching rates found in atmospheric-pressure flames. Daily 7 hassuggested operation ip this regime to obviate the need for quenchingrates to analv:e laser-excited fluorescence data, and recently Baronavskiand McDonald 8 have nsed fluorescence measurements and a series expansionof'the fluorescent signal equation to measure the quenching of C2 in aflame. Using the optoacoustic pulses, we have measured the totalquenching rate for Na in the acetylene-air flame.

We consider a two-level system (see Figure 11) and write a steady-

state equation for the excited state number density N e

ee

e.d -aB • I 2tO-BN0 - (BI +Q"A)N

The assumption of a steady state is valid since thu time scales involvedfor the absorption and quenching processes are significantly shorter thanthe laser pulse duration. Here, B and A are the Einstein coefficients

for absorption and spontaneous emission, respectively; tne term BINeaccounts for stimulated enission. I is the laser spectral power densitymeasured in (ergisec} per cm2 per unit frequency interval. Q is thetotal quench rate, per sec, for the upper level.

0Since the total Na number density is a constant N -i.e.Io+N +N=

e g 0

P-1 W. LDrziiy, "Saturat,-:on Ef~-ecz;z n Leser2- Mu-ced Fluorescence

S'vecttrc'scop., ADD!. Op; a6-6-571 (19?7).A. P. 3a-ornavsk` aria j. R. M. -D?!co'-rd, 'auentof C.Cnetain

in y x~gen-acetd ene Flm:An Alct fS~atza ton Spectros-eppj", J. Chem. Phys. 66 30-3301 (197?); A. P. Baronavaki and JT R.•- Do •, "Apo ication o f Satu-ation Spect-roscopv to the Measiemernt of

C2 , 3r-u Corcentratsionz in 0---acetylene Flaes", .4ppl. Opt. 1 1897-1901

07977).276

0)IL ciL ~-0

LLI0'a,

o4-'A

Li 0 a,

0 S-4.cno

go 0-

00

CL LLLJ~4 0 0.

(ft 4' C1- 4' L

(D 4a

CDC

L- +

cr0 <

U- u a n

.~ 4' 4 O

0,.- CVC

Eq. (2) may be solved for the steady-state value of Ne

S~BINo0V

(N ss. - Q+A+2BI (3)

Now the opto-acoustic pulse signal .M1 is proportional to (Ne)

e sss

where the constant represents the efficiency of conversion of electronicto translational energy, the microphone and amplifier characteristicsand geometrical considerations. Substituting into Eq. (2), one has forthe inverse of the microphone s-gnal t

c02 QA 1 (4)

Thus a plot of M (in arbitrary units) vs. I (in absolute units)should yield a straight line with a slope to intercept ratio of

Q+A-x QLIA-2B 2B/A

The relationship becween B and A is known from thermodynamic consider-ations:

)5B =- A (6)2hc

so that the experimental result serves to determine Q/A. But if A hasbeen separately measured, then an absolute value of Q may be extracted.

It may be parenthetically noted that these simple equations pertainto a two-level model only. Sodium itself has three levels participattingin this process, the 2 S1 / 2 , 2 Pl/2, and 2 P3 /2. Taking into account thetransfer between the upper doublet components, which requires a separate t

fluorescence scan of the D-iine emission, one may still obtain an anal-ysis of the data. This rcfinement for Na is discussed in Ref. 2. Fora yet more ccmplicated system such as a molecule with a series of rota-tional levels among which energy transfer may occur, the equations

28

7A° ;47'V

become considerably more complex and require much more independent in-formation on the energy transfer cross sections. Theoretical consider-ations of such a system are currently being pursued at the BallisticResearch Laboratory. 9

We nonetheless proceed with an analysis of the present data interms of a two-level model, in order to assess its utility. Anexamination of Eq. (4) shows that if the laser intensity is varied overa range far lower than saturation, the effective intercept will be zero.This is illustrated for an unfocussed laser beam in Ffgure 12, where thespectral power density is here given in erg cm- sec Hz- 1. Theexperiment was repeated after focuss ng the laser to a beam diameterone third the size, resulting in power densities an order of magnitudelarger. These results are plotted in Figure 13. The low power levelsof Figure 12 do not permit the distinction of a non-zero intercept,beyond experimental error bars, although the finite intercept is readilyapparent in Figure 13.

Figure 14, shows a plot of the data for one run, in the form of Eq.(4). The error bars represent estimates of the readability of the box-car signals; the line is an unweighted least squares fit to the data.In this instance we obtain Q/A = 410 ± 40. Using the value A = 6.3 x107 sec- 1 , this results in a Q = (2.6 - 0.3) x 1010 sec-i.

This is too large, by comparison with calculated values using Fpreviously determined cross sections for the quenching collisionsT0Part of the problem is due to the Gaussian nature of the laser beamspatial profile,il and our choice of intensity/range and Na density.Reference 2 contains more realistic results combining the opto-acousticpulse measurements with fluorescence results. Nonetheless, these resultsclearly demonstrate that the opto-acoustic pulse method forms a useful

detection technique.

V. 4P LEVEL PUMPING

SExcitation to the 4P level is providgd by frequency doubling thelaser to the resonancc wavelength of 3303A. Although the oscillator"strength of this transition is low, we can operate at high enough Nadensity to observe -zasureable excitation, as confirmed b) fluorescence

H . Gelb A V- -oa f

•. , A. J. •-ar ,n-nd D. R. Crosley, "Response o a 'olecularSyster, to Optical -ncetation Un. Saturatin Conditions", o be

P. Hooinwera and C. Th. J. AZkw.d, n "Q"u=hir-g of Fr-ited l-kalist Atoms and ReZated Effecta in Facnea. Part II. Measurm.ents and

Discussion", J. Quant. Sýect. Pad. TaTnsf. 6_j 847-874 (1966).

,'j. W. ,ailly, "Saturation of Fluorescence in Flmes Lith a Gaussian

Beam"• Avn.. Ozt. 17 -229 (1978).

)29

1 -4

_ nnmm n 9

VA

101

to.

o o0d~1

03.4C4 CL1

0 0~

~. 63

- - $ -., -;--V - - - TR7 - -- -

.~...

-LU

CV) 1-

Rta

.4V

CL

Ci

T &m

>LI-

10

C4 44

.1 4-'

30

C, Cl

s-i

-n CE 4) 4

In(s4!un~~~ ~% Cj~4gD .L

,l . S.6-

LO0

32 I

from the 4P level itself, as well as from other levels populated bycollision (see below). Using the amount of energy absorbed, we mayestimate the size of the expected opto-acoustic pulse signal in a mannersimilar to that outlined above for the 3P level. The anticipated valueis 18 mTorr at the microphone, but no signal is observed above the back-ground noise level of 0.2 mTorr.

We find the absence of an opto-acoustic signal from this higher-lying level quite surprising. One possibility is that there exists amarkedly different rate either for direct quenching of the 4P levelitself, or for the rate of subsequent V to T transfer in the flame gases.If the energy were to be trapped in internal modes, it might not formas sharply defined an initially heated region and thus might yield asound wave much reduced in amplitude. Another possibility is that the4P level is removed rapidly by chemical reaction, so that little of the

electronic energy is converted into translational energy. Although thislatter phenomenon has been seen to exist in flames for the 3P level,12it would have to be much faster than observed there in order to competewith collisional quenching in our system.

Further information about relative quench and energy transfer ratesis obtained from observations using two-photon excitation. 2 There, thelaser was tuned to the avelength corresponding to half the energy dif-ference between the 3S state and the 3D, SS or 4D state. Observationswere made of emission from all four high-lying states when the laseris tuned to excite each one of them, in turn. Ratios of the populationsof these levels were obtained from ratios of the fluorescence intensitiesand inserted into steady-state rate equations describing the collisionalenergy transfer. By invoking detailed balancing to relate upward anddownward transfer rates between pairs of states, a determination was madeof all ten transfer rates among the four levels as well as the totalquenching rates out of each state, normalized to the value of one of them.

It may be seen from the results of Reference 2 that the 4P leveldoes indeed possess a collisional loss rate (at least in this flame)which is considerably smaller than that of the other nearby levels. On

.the other hand, the absolute quench rate still remains high, faster thanone quenching collision per nsec. This value is obtained from ourmeasurements on the 3P level together with quenching cross sections forN2 obtained from low pressure experiments on both the 3Pl0 and 4P 1 3 levels.Consequently, there would appear to be no energy bottleneck attributableto at least the E to V aspect of the energy transfer.

IC. H. Miler IIi, K. Schofield and M. Steinberg, "Year SaturatiLonSLaser Induced Chemical Reactions of Ila (32 P1 ) in H 0.7. Fl~rzs

Chem. Phys. Letters 57 364-368 (1978).S1T. P. Gallaher, W. E. Cookeand S. A. Ede~stein, "Collisional Deacti- J

vation of the 5s and 4p States of ila by 112" Phs. Rev. A17 125-7131(1978).

33 -

7-71

If the quenching of the 4P level returned the Na predominantly tothe ground (3S) state, then about twice as much energy would be trans-ferred to the vibrational levels of the flame gases as occurs in a 3Pto 3S quenching collision. This would populate a different set ofvibrational levels in the flame gases, and conceivably one could havea different mode of V to T transfer and slower formation of the pressurewave. However, the results of Reference 2 argue against this possibilityas well. It is seen from those state-to-state transfer rates that halfof the transfer out of the 4P leaves the Na atom in the 5S, 4D or 3Dlevel (mostly the 3D). Much of the remainder is likely to terminateon the 3P, as fluorescence is observed originating from that level aswell. (In fact, the results of Reference 2 show altogether convincingevidence of step-wise relaxation in the flame, as opposed to a directreturn to the ground state. There is a smooth decrease of energytransfer probability with energy defect. This fact has importantimplications for treating, in a microscopic way, the occurrence of non-equilibrium populations, especially of reactive species, within hightemperature systems.) As a result, it appears that a similar range ofvibrational levels in the flame gases participate in the E+V+T transerout of the 4P level and the 3P level.

The small value of Q(4P) from Reference 2, which necessarily includesreactive collisions as well, and the size of k(4p-3d) suggest thatreaction of the 4P level does not occur rapidly. We thus do not under-stand why the 4P level excitation does not produce opto-acoustic pulses.Perhaps the answer lies in more detailed understanding of the dynamicalformation of the pressure wave.

VI. CONCLUSIONS

We have described a new type of opto-acoustic effect -- the produc-tion of pulsed sound waves in an atmospheric pressure flame followingdeposition of a laser pulse into electronic energy levels. The experi-ments on Na demonstrate that good signal to noise ratios can be obtainedat sub-ppb levels of the seeded atoms, using a co-..ercially availablelaser of modest power. (We add that opto-acoustic signals have alsobeen observed by pumping Na in the C2H2-air flame using a dye laserpumped by a N2 laser; a more quantitative study may reveal furtherdetails concerning the formation of the pressure wve, since the dyelaser pulse is in this case only several nsec long as opposed to 1 usecfor the flashlamp-pumped laser.)

The opto-acoustic pulse signal thus forms a sensitive detector ofselective absorption of laser radiation. There appears to bee no funda-mental reason why the technique should not be applicable to other species,including in particular molecular absorption systems. We remain mysti-fied, however, by the lack of opto-acoustic signal from the Na 4P exci-tation, which suggests that the situation is not entirely straightforwardand should be approached on a case-by-case basis. Nonetheless, we feel

34

ITIV

ý7:

that there is promise for the technique simply as a detection mode,particularly for species which absorb laser radiation but do notfluoresce readily (either due to high pressure or the rapid crossinginto a non-fluorescing state).

The form of the pressure pulse produced, if it yields to analysis,holds information on density gradients within the flame and perhaps onrelaxation phenomena within the flame gases. For the investigations ofquenching in Na under conditions approaching optical saturation, detect-

1~ ion via opto-acoustic pulses provides a measure of the excited statenumber density under conditions in which fluorescence detection ishampered by self-absorption problems. It may. thus be a useful complementto fluorescence in experiments designed to probe behavior near saturationover a wide range of Na density.

Finally, and perhaps most important, the method affords the abilityio produce locally within the flame a pulsed sound wave. This has pro-vided rapid and spatially resolved measurements of the speed of soundwithin the combusting system, and could form a useful tool to generatesound waves for an investigation of the noise produced by flames.

31

SJ

-I•

33N

S.. . . . . . . . . . . . .. . . .

REFERENCES

1. A. C. Eckbreth, P. A. Bonczyk and J. F. Verdieck, "Review of Laser

Raman and Fluorescence Techniques for Practical Combustion Diagnostics",United Technologies Research Center Report R77-952665-6, Hartford,CT, February 1977.

2, J. E. Alien, Jr., W. R. Anderson, D. R. Crosley and T. D. Fansler,"Energy Transfer and Quenching Rates of Laser-Pumped ElectronicallyExcited Alkalis in Flames", Seventeenth Symposium (International)on Combustion, Leeds, England, August 1978.

3. D. Hartley, M. Lapp and D. Hardesty, "Physics in Combustion Research",Physics Today, December 1973, 36-47.

4. R. B. Green, R. A- Keller, P. K. Schenck, J. C. Travis and C. G. 1:Luther, "Opto-Galvanic Detection of Species in Flames", J. Amer.Chem. Soc. 98, 8517-8518 (1976).

5. H. Y. Pao (ed.), Optoacoustic Spectroscopy and Detection, AcademicPress, New York, 1977.

6. A. G. Gaydon and H. G. Wolfhard, Flames: Their Structure, Radiationand Temperature, 3rd Ed., Chapman and Hall, London, 1970.

7. J. N. Daily, "Saturation Effects in Laser-Induced FluorescenceSpectroscopy", Appl. Gpt. 16, 568-571 (1977).

8. A. P. Baronavski and J. R. McDonald, "Measurement of C2 Concentra-tions in an Oxygen-acetylene Flame: An Application of SaturationSpectroscopy", J. Chem. Phys. 66, 3300-3301 (1977); A. P. Baronavskiand J. R. McDonald, "Application of Saturation Spectroscopy to theMeasurement of C2, 3 .1u Concentrations in Oxy-acetylene Flames",Appl. Opt. 16, 1897-1901 (1977).

9. A. H. Geib, A. J. Kotlar and D. R. Crosley, "Response of a MolecularySstem to Optical Excitation Under Saturation Conditions", to be

published.

10. P. N;oomayers and C. Th. J. Alkemade, "Quenching of Excited AlkalisAtoms and Related Effects in Flames. Part 11. Measurement andDiscussion", J. Quant. Spect. Rad. Transf. 6, 847-874 (1966).

11. J. W. Daily, "Saturation of Fluorescence in Flames with a GaussianBeam", Appl. Opt. 17, 225-229 f1978).

12. C. H. Muller III, K. Schofield and M. Steinberg, "Near SaturationLaser Induced Chemical Reactions of Na (3 2P3/2,1/2) in H2 /0 2 /N1Flames", Chem. Phys. Letters 57, 364-368 (1978).

37 -

Y3X=W AEBLWkTIjW

REFERENCES (continued)

13. T. F. Gallagher, W. E. Cooke and S. A. Edelstein, "Collisional

Deactivation of the 5s and 4p States of Na by N21, Phys. Rev. A17,

125-131 (1978).

I !3

---t

oIV

DISTRIBUTION LIST

No. of No. of

Copies Oranization Copies Organization

12 Commander 1 CommanderDefense Technical Infor US Army Armament MaterielATTN: DDC-DDA Readiness CommandCameron Station ATTN: DRSAR-LEP-L, Tech LibAlexandria, VA 22314 Rock Island, IL 61299

1 Director 1 DirectorDefense Advanced Research US Army ARRADCOM

Projects Agency Benet Weapons LaboratoryATIN.: LTC. C. Buck ATTN: DRDAR-LCB-TL1400 Wilson Boulevard Watervliet, NY 12189Arlington, VA 22209

1 Commander

2 Director US Army Watervliet ArsenalInstitute for Defense Analyses ATTIN: Code SARWI-RD, R. ThierryATTN: H. Wolfhard Watervliet, N"f 12189

R.T. Oliver400 Army-Navy Drive 1 CommanderArlington, VA 22202 US Army Aviation Research

and Development Command1 Commander ATTN: PRSAV-E

US Army Materiel Development P.O. Box 209and Readiness Command St. Louis, Mx 63166

ATTN: DRCD.MD-ST5001 Eisenhower Avenue 1 DirectorAlexandria, VA 22333 US Army Air Mobility Research

and Development Laboratcry2 Commander Ames Research Center

US Army Armament Research Moffett Field, CA 94035and Development Command

AATTN: DRDAR-TSS 1 CommanderDover, NJ. 07801 US Army Communications Research

and Development Command4 Commander Alr.: DRDCO-PPA-SA

US Army Armament Research Ft. Monmouth, NJ 07703and Development Command

ATTN: DRDAR-LCA, J. Lannon 1 CommanderDRDAR-LC, J.P. Picard US Army Electronics ResearchDRDAR-LCE, R.F. Walker and Development CommandDRDAR-SCA, L. Stiefel Technical Support Activity

-SCA L.IieDover, 1V 07801 Aq-fN: DELSD-L

Ft. Monmouth, NJ 07703

z

39

ii

DISTRIBUTION LIST

No. of No. ofCopies Organization Copies Organization

Commander 1 Office of Naval ResearchUS Army Missile Command ATTN: Code 473ATTN: DIRSMI-R 800 -N. Quincy Street

Redstone Arsenal, AL 35809 Arlington, VA 22217

1Commander CommanderUS Army Missile Command Naval Sea Systems CommandATTN: DRSMI-YDL ATMN: J.W. Murrin (SEA-62R2)Redstone Arsenal, AL 35809 National Center

Bldg 2, Room 6E081 Commander Washington, DC 20362

US Army Natick Researchand Development Command 1 Commander

ATTN: DRXRE, D. Sieling Naval Surface Weapons CenterNatick, MA 01762 ATTN: Library Br, DX-21

Dahlgren, VA 224481 Com•ander

US Army Tank Automotive Research 2 Commanderand Development Command Naval Surface Weapons Center

AMTN: DRDTA-UL ATTN: S.J. Jacobs/Code 240Warren, MI 48090 Code 730

Silver Springs, MD 209101 Commander

US Army White Sands Missile 1 CommanderRange Naval Underwater Systems Center

ATTN: STEWS-VT Energy Conversion DepartmentSWSM, NMI 88002 ATTN: R.S. Lasar/Code 5B331

1 adNewport, RI 028401 Commander

US Army Materials & Mechanics 2 ConrnanderResearch Center Naval Weapons Center -

A•N: DRXMR-ATL ATTN: R. DerrWatertown, MA 02172 C. Thelen

China Lake, CA 93555I Commander

US Army Research Office 1 Commander% ATTN: Tech Lib Naval Research Laboratory I

Research Triangle Park, NC ATTN: Code 6180*z7706 Washington, DC 20375

1 Director 3 SuperintendentUS Army TR.ADOC Systems Naval Post Graduate School:1 .nalysis Activity ATTN: Tech LibATiN: ATAR-SL, Tech Lib D. NetzerWhite Sands Missile Range, A. Fuhs'N 88002 Monterey, CA 93940

40

DISTRIBUTION LIST

No. of No. ofCopies Organization Copies Organization

Naval Ordnance Station AVCO Everett Research Lab Div

A.TN: S. Mitchell ATTN: D. SticklerTech Lib 2385 Revere Beach Parkway

Indian Head, MD 20640 Everett, MA 02149

3 AFOSR 1 Bell LaboratoriesATTN: L. Caveny ATTN: A. Tam

B.T. Wolfson Murray Hill, 1J 07974Bolling AFB, •. 20332

1 Calspan Corporation"2 AFRPL (DYSC) AATTN: E.B. Fisher

ATTN: D. George P.O. Box 400J.N. Levine Buffalo, NY 14211

Edwards AFB, CA 93523

4 o u o n1 Foster Miller Associates, Inc.4 National Bureau of Standards ATTN: A.J. Erickson

MATN: T. Kashiwagi 135 Second AvenueM. Linzer Waltham, MA 02154K. Smith

Washington, DC 20234 General Electric CompanyArmament Department

1 Lockheed Palo Alto Resesrch Lab ATan: DepartmenATTN: Tech Info Center- ATTN: M.J. Bulman

3521 Hanover Street Lakeside Avenue352 Haove SteetBurlington, XT 05402Palo Alto, CA 94304

Director General Electric CompanyFlight Propulsion DivisionNASA Goddard Space AT:Ta i

FlightCenterATTN: Tech LibATigh J. Aen Jr .Cincinnati, OH 45215ATTN: J.E. Allen, Jr.

Greenbelt, MD 20770 2 Hercules Incorporated

Alleghany Ballistit LabI Aerojet Solid Propulsion Co. AM-N: R. Miller

ATTN: P. Micheli Tech LibSacramento, CA 95813 .c iCumberland, MD 21501

2 ARO Incorporated IITRIATTN: N. Dougherty ATTN: M.J. Klein

W. Williams iO West 35th StreetArnold '3S, TN 37389 Chicago, IL 60615

Atlantic Research CroporationAMT: M.K. King

5390 Cherokee AvenueAlexandria, VA 22314

41

-. 1 m

DISTRIBUTION LIST

No. of No. ofCopies Organization Copies Organizetion

1 Olin Corporation 3 West Virginia UniversityBadger Army Ammunition Plant Department of ChemistryATTN: J. Ramnarace ATTN: G. HavrillaBaraboo, WI Morgantowp, WV 26506

3 Olin Corporation Aberdeen Proving GroundNew Haven PlantATTIN: R.L. Cook Dir, USAMISAA

D.W-. Fiefler ATTN: DRXSY-D275 Winchester Avenue DRXSY-M, H. Cohen

New Haven, CT 06504Cdr, 'JSATECONI

1 Paul Gough Associates, Inc. ATTN: DRSTE-TG-FATTN: P.S. GoughP.O. Box 1614 Dir. USACSL, Bldg E3516, EAPortsmouth, NH 03801 ATTN: DRDAR-CLB-PA

1 Phillips Petroleum CompanyResearch and Development DeptATTN: A.M. Schaffer124-RBSBartlesville, OK 74004

10 SRI InternationalATTNq: D. Crosley

333 Ravenswood AvenueMenlo Park, CA 94025

6 United Technolegies Research Corp..AMN: A.C. EckbrethEast Hartford, CT 06108

1 Georgia Institute of TechnologyATTN: B.T. ZinnAt]-'nta, GA 30332

8 University of ArkansasDepartment of PhysicsATTN: R. GuptaFayetteville, AK

5 Uni',ersity of Florida IDepartment of ChemistryATTN: J. WinefordnerGainsville, FL 32601 Ai

42