91-052

IONIZATION PROCESSES IN THE INTERELECTRODEREGION OF AN MPD THRUSTER

T.M. Randolph*, W. F. von Jaskowsky **. A.J. Kellyt and RG. JahnttElectric Propulsion Laboratory

Princeton UniversityPrinceton, NJ 08544

Abstract I. Introducton

A spectroscopic technique for The initial Propellant ionization phase

measuring ionization fractions in the is an important determining factor in MPD

interelectrode region of an MPD thruster thrusters efficiencies. High propellant

has been developed. A specially designed ionization fractions are needed to produce

MPD thruster has been constructed to allow the electrical conductivity necessary for

spectroscopic access to the interelectrode interesting electromagnetic thrust

region. Visible spectra from emitted components; however high propellant

plasma radiation have been recorded on ionization fractions also result in large

photographic film and analyzed by a frozen flow losses 1 -3. These frozen flow

scanner/computer system. Using relative losses can be attributed to slow

emission line intensities, electron recombination rates downstream in the

temperatures ion/neutral ratios and thruster; thus nearly all the ionization

electron densities have been obtained, energy of the initial propellant ionization

Spatial resolution along the thruster axis phase is dissipated in the propellant

of better than 0.2 mm has been achieved exhaust 1 . In order to understand the source

with radially integrated measurements. of this energy dissipation, an investigation

Data has been obtained with argon mass of the initial ionization phase within the

flow rates between 3.3 mg/s and 11 mg/s interelectrode region was conducted.

and operating currents between 110 A and Most theoretical studies of the initial

850 A in a 20 kW class steady state thruster. ionization process within the MPD thruster

The results of this study show peak have been unable to accurately predict

ionization fractions near the propellant experimentally observed ionization

inlet region and a strong scaling of fractions. Theoretical studies of the

interelectrode ionization fractions with ionization process typically underpredict2 / . ionization rates necessary to produce the

observed ionization levels of MPD

thrusters 4 - 6 . Recent investigations of

Graduate Student, Department of plasma instabilities have provided a

Mechanical and Aerospace possible explanation for such

Engineering. Member AIAA. discrepancies. Choueiri has hypothesizedthat superthermal electrons, produced by

Senior Research Engineer and Lecturer 1"anomalous" turbulent plasma heating,- SeniorResearch EngineerandIcturer m b rps r t enhanced

Emeritus, Department of Mechanical may be responsible for the enhanced

and Aerospace Engineering. Member ionization observed experimentally 7 .

AIAA. Experimental verifications of thishypothesis need to be directed toward the

Senior Research Engineer Department MPD thruster interelectrode plasma.of Mechanical and Aerospace The difficulty of employing variousEngineering. Member AIAA. diagnostic techniques in the MPD thruster

Engineering. Member AIAA. interelectrode region has made

SP nic investigations of the initial ionization

SProfessor, Department of Mechanical process difficult. Studying ionization

and Aerospace Engineering. Fellow processes with electrostatic probes has beenAIAA. limited for two primary reasons: the

inability of probes to measure neutral

1

91-052

number densities directly and the difficulty maintained below 2x10- 4 torr. with powerof obtaining spatial profiles of relevant being supplied by a 50 kW welding supply.parameters. The modified MPD thruster used to gain

The difficulty of gaining optical access optical access to the interelectrode regionto the interelectrode region has been the has a 2% thoriated tungsten cathode, aprimary limitation of spectroscopic graphite anode/body and a boron nitridemeasurements. Several spectroscopic insulator (fig. la.b). A 3 mm thick radialstudies have been made of the slot was machined from the interiorinterelectrode region using radial slots at surface of the insulator to the exteriorseveral axial positions 8 - 10. Russian surface of the graphite body. The 7 cm longspectroscopic experiments, performed with slot extends from the propellant inlet portgreater axial resolution, have indicated the to the thruster exit plane allowing opticalexistence of a narrow ionization "front"l 1- access to the entire interelectrode region A13 Observations of narrow ionization quartz window, secured by a ceramic"fronts" are indicative of "anomalous" holder, was used to prevent propellant flowplasma behavior and have provided the out the slot. Typical operating conditionsimpetus behind further spectroscopic for the thruster, with argon propellant.studies. involved currents between 110 A and 850 A

Two primary spectroscopic methods and mass flow rates between 3.3 mg/s andhave been employed for the determination 11 mg/s.of ionization fractions in MPD thrusters. One concern for obtaining accurateThe most common technique is the spectroscopic data was dischargeutilization of Stark broadening in the asymmetries. The thruster was checked foremission lines of a seeded species. asymmetries in its self induced magneticUnfortunately, the seeding levels necessary field by an existing probe systeml5to make this technique viable would Magnetic field measurements revealed nosignificantly upset the ionization levels in major radial asymmetries In the near fieldthe thruster. Another technique involves plume. Visual inspections of the anodethe measurement of absolute intensities in region during thruster firings were alsoemission lines. The measurement of made. These inspections revealed noabsolute intensities typically requires significant preferential current attachmentdifficult calibrations Inherently sensitive to the rad ' slot.to the optical transmission losses Another concern for obtaining reliableprevalent in this experimental spectroscopic data is the fouling of theconfiguration, interior quartz window surface. After high

To avoid the uncertainties prevalent in current thruster firings, in particular, thethese techniques, the method of relative optical transmission of the quartz windowline intensities has been employed. Using can be visibly reduced by ablated carbonthis method, electron temperatures and and tungsten deposited on the interiorIonization fractions can be determined by surface of the quartz window. Thisthe relative spectral intensity of different deposited material reduced the opticalspectral emission lines. Using a specially transmission significantly in thedesigned MPD thruster. It was possible to downstream portion of the Interelectrodeobtain such information from the emission region. The reduction in opticalradiation of interelectrode plasma species, transmission is remedied by periodic

window cleaning every few thruster firings.

II . FacilitiesI. Diagnostics

Spectroscopic measurements ofion/neutral ratios in the MPD thruster A. Optical Diagnostic Set Upinterelectode region were performed withinthe Princeton steady state thruster The optical system used to collect thefacility 3 , 14, 15. The vacuum facility radiation emitted from the MPD thrustercpnsists of a 1.5 m diameter, 6.4 m long Interelectrode region Is shown in figure 2.tank pumped by one 1.2 m diffusion pump Light emitted from the Interelectrodebacked by a roots blower and mechanical region is collected by an achromatic lenspump. Tank pressure is typically and reflected off a first surface mirror. The

2

91-0_ 5

dove prism is used to rotate the horizontal dispersion by the measured spatial line

image of the thruster slot to align with separation 1 6 . By comparing thevertical spectrometer entrance slit. The wavelength separation of observed spectral

spectrometer disperses the emitted lines to spectral references of likely atomicradiation along its output plane. The and molecular sources, it is possible toresulting spectra were recorded on Polaroid identify their respective transitions 1 7 -19

57 film by an manually triggered shutter. Although the resolution of theEach photographic spectrum contains both experimental instrumentation (± 0.2 A) isspatial and wavelength Information, the too limited for absolute line identification,abscissa corresponding to wavelength and the appearance of emission linethe ordinate corresponding to the thruster wavelengths and intensities closelyaxial position, corresponding to the spectral references is

Rigorous optical alignment procedures considered adequate for plasma speciesare performed for the lens/ identification.prism/spectrometer system. Initial optical The dominance of the argon propellantalignment is performed by a Helium-Neon species in the interelectrode region is quitelaser. A final alignment check is pronounced In the series of lineperformed by focusing the reflected light identification photographs (fig. 3). Argon IIfrom a mercury lamp through the thruster emission lines are prevalent through mostslot. Both checks were performed while the of the photographs, indicating that singlytank was evacuated thereby insuring the ionized argon is the dominantalignment of the optics under thruster interelectrode plasma species. Severaloperating conditions, strong argon I emission lines are also

observed, particularly in the 3900 A to 4400A spectral region (fig. 3). Only three argon

B. SpectralLine Identification III emission lines are Identified: thusrevealing that the interelectrode

Identification of the Important population of multiply ionized species isinterelectrode plasma species is low.accomplished by the identification of Several emission lines of impurityemission lines from the Interelectrode species are also Identified. Molecularplasma. The output of the spectrometer carbon, ablated from the interior anodecontains an approximately 450 A wide surface, is identified by the "Swan" bands atspectral region, requiring several 4685 A, 4698 A and 4715 A. Thorium II,photographs at different spectrometer from the impregr.ated cathode, is identifiedsettings to record a sufficiently large by several weak emission lines in the 4000spectral range. The 3400 A to 5800 A A to 4400 A spectral region. Iron I. thespectral region is recorded using 8 Polaroid source somewhat of a mystery, is identifiedphotographs. To produce the clearest by several emission line in the 3500 A topossible photographs, spectrometer slit 4000 A spectral region. Barium II.widths and photographic exposure times deposited In the interelectrode interior byare varied accordingly from 1 pim to 10 jrn earlier thruster firings using a Bariumand 1 s to 5 s, respectively. Argon impregnated cathode and ablated later, ispropellant, thruster operating conditions Identified by several strong emission linesare set at 10 mg/s mass flow rates and 450 A Including the 4309 A emission line (fig. 3).operating currents to obtain thesephotographs.

Spectral line identification is C. Photographic Data Reductionaccomplished by determining the spectralline wavelength separation and comparing After identifying emission lines in theit to existing spectroscopic tables. The MPD thruster interelectrode spectrum, thephotographic spatial separation between 3900 A to 4400 A spectral region is chosenthe spectral lines is measured with a for a more detailed investigation. Thecomparator. The spatial resolution is abundance of argon I and argon II emissionlimited by the Polaroid film grain size (20 lines, of similar intensities, made thisline pairs per mm) to ± 0.025 mm. Spectral spectral region Ideal for electronwavelength separations are determined by temperature and ion/neutral ratiomultiplying the spectrometer's linear determinations. Photographs of this

3

91-052

spectral region are taken during thruster A resulting plot of emission lineoperating conditions of 110 A to 850 A irradiance profiles is shown In figure 4.currents and 3.3 mg/s to 11 mg/s mass flow The irradiances of the three spectral linesrates. After development, these are displayed in arbitrary units versus thephotographs are individually scanned to thruster axial position. Figure 4 shows howobtain quantitative relative emission line the profiles abscissa corresponds with theintensity information. thruster axial geometry. The intensity

To obtain spectral irradiance drop at the 4.4 cm point in each line profileinformation from the spectral can be attributed to the fiberglass thread,photographs, a scanner and computer are used to secure the quartz window to theused. The electronic scanner digitized both thruster body. The dramatic characteristicspatial and photographic density intensity drops at both the inlet region andinformation from individual photographs thruster exit point can be attributed to thefor computer analysis. The spatial vignetting effect of the spectrometerresolution of a scan is limited to 300 dots mirrors. These optical effects significantlyper inch by the scanner pixel size. reduce confidence in data obtained in thesePhotographic density is reduced to grey thruster axis regions.level information by an 8 bit analog-digitalconverter and stored in a TIFF image file IV. Resultsformat. This process is repeated for eachphotographic spectrum. The spectroscopic measurements

Once stored as an image file, the data is resulting from the analysis of filmfurther reduced to obtain film irradiances. irradiances are radially integrated.The photographic reflectivity is converted Previous experimental work has shownto irradiance by a film characteristic curve significant radial variations inequation. Irradiance is corrected for film interelectrode plasma properties 8 - 10;wavelength sensitivity by a wavelength therefore measurements obtained in thisversus sensitivity equation. Both of these work are used only to show general trendsequations were obtained from curve fits to in the axial variations of interelectrodePolaroid factory calibration curves. Due to plasma properties.variations in film properties, little To obtain plasma parameters fromconfidence can be placed in this calibration spectral information obtained on film, thefor absolute irradiance values. This lack of interelectrode plasma radiation intensityconfidence is relatively unimportant for must be related to the film irradiance. Therelative intensity measurements which are film irradiance is proportional, by athe concern of this experiment, geometrical factor, to the plasma radiation

Once the film density to irradiance intensity. The primary concern of thisreduction is performed, it is possible to experiment is relative intensityobtain the integrated line irradiance measurements; therefore theprofiles of interesting atomic transitions, determination of the optical geometryTo obtain these integrated line irradiance factor is unnecessary.profiles for a given spectral emission line,the irradiance is summed over the spectrallinewidth while subtracting the A. Electron Temperaturesbackground irradiance2 0 .

Electron temperatures, needed for thedetermination of ion/neutral ratios, were

Llk= LXdX fLbd, obtained by the ratio of excited state

line line emission line intensities 2 0 .

Llk = emission line irradiance kTe = En - Eln

Lk = film irradiance In (lk Anm n Ilk

LXb = background irradiance \Xnm Alk g1 Inm

The background irradiance is determined En = excited state energyfrom an interpolated value of the pixels on Anm = transition probabilityboth sides of the spectral line. gn = excited state degeneracy

4

91-052

Inm = emission line intensity El = excited state energy w/r ground

The subscripts n and 1 refer to the upper Superscript + denotes the ion. Equilibrium

state of a given transition and the between the excited states and their

subscripts m and k refer to the lower state. respective ion or neutral ground state is

An optically thin transition is assumed, assumed. This assumption is marginally

This assumption is well justified, for non justified in MPD thrusters plasmas3. No

resonant transitions, by the comparatively assumption is made with respect to

low pressures typical of MPD thrusters3 . equilibrium between ions and neutrals.

Equilibrium is also assumed between the Ion/neutral ratios, made by the relative

excited states n and m. This equilibrium intensities of the 4014 A argon II and 4259

assumption is also well justified with the A argon I emission lines, are presented at

typical electron temperatures and densities the end of this paper. Shown in figure 7,

of MPD thrusters 2 large thruster current variations, at

Electron temperatures. made by the constant mass flow rates, dramatically

relative intensities of the 4014 A and 4104 effect the ionization levels of the

A argon II emission lines, are presented at interelectrode plasma. Shown in figure 8.

the end of this paper. Shown in figure 5, ion/neutral ratios, at constant j2/,k are

large thruster current variations at relatively unaffected by thruster current

constant mass flow rate s have a minimal ncreases. These two observations indicate

effect on electron temperatures. Note that a strong scaling of plasma ionization levels

the electron temperatures are also with the parameter j 2 . Note that the

relatively constant along the thruster axis profiles n figures 7 and 8 show a peak near

Somewhat similar observations can be the propellant inlet region; therefore

made in figure 6 where j 2 / is held initial propellant ionization appears to be

constant while the current is increased. In concentrated far upstream in the

general, only small electron temperature interelectrode region.variations are observed over a large range

of thruster operating conditions. These

observations coincide with previous C. Electron Densities

experimental work 2 1. In this reference,

Simpson determined that plasma energy Ater the determination of Electron

density variations dramatically effected temperatures and ion/neutral ratios, rough

ionization levels without significantly estimates can be made of electron

lines 2 0 2

effecting the electron temperature. density . a

B. Ion/Neutral Ratios 2NU (2me kTed i t

N Alk Inm j n t h e expansion region of the MPD

Using electron temperatures obtainedee e e e Uh ee e e

from excited state line ratios, ion/neutral

ratios can be determined from the relativetrintensities of ion and neutral emission = lectron

lines 2 0 . = ionization energy

Equilibrium between the ions and neutralsis also assumed in this expression. This

assumption may be particularly difficult to

N+ )m Alk g1 I+m U+ justify in the expansion region of the MPD

N x thruster because slow recombination ratesN Xlk Anm g+ Ilk U prevent ion/neutral equilibrium; therefore

(En -El electron densities obtained in this region

exp -e ) should be treated with caution.Electron density estimates of

N = number density approximately 1020 - 1021 m- 3 were made

U = electronic partition function with the electron temperatures and+ excited ion/neutral ratios presented in figures 5-8.

En = excited state energy w/r ion ground Good comparisons with electron densities,

5

91-052

measured in other experimental studies, port Indicate an initial propellantincrease the confidence in this technique ionization phase occurring over a veryfor electron densities determination 8 - 10 small length scale. Characteristic

Ionization phases occurring over lengthscales smaller than the classical ionization

D, Error Analysis mean free path would indicate the presenceof plasma turbulence on the ionization

Photographic measurements are process. A more detailed spectroscopicinherently full of uncertainties; thus a investigation of the propellant inlet isdiscussion of experimental errors is quite required to make this determination.important in interpreting the experimental The results of this work areresults. The primary source of error for preliminary in nature and should berelative intensity measurements, displayed treated with appropriate skepticism.in figures 5-8, is the bit resolution of the Future work on atomic modeling isanalog digital converter. Experimental necessary to reduce the reliance onconfidence is good near the thruster axis equilibrium assumptions. A more detailedmidpoint but becomes progressively worse spectroscopic investigation of the inlettoward either the inlet or exit plane. This region is to be performed after thelack of confidence results from low elimination of the spectrometer vignettingradiation signals in this region attributed effect. This additional work will enableto spectrometer vignetting. more accurate determination of

Errors involving uncertainties in ion/neutral ratios near the propellant inletvalues for transition probabilities were not port.Included in the analysis discussed above.Inaccuracies in the transition probabilitiesused result in a shifting of all the plasma Referencesparameter profiles away from their actualvalues. This results in errors in absolute 1. Bruckner. A. P.. "Spectroscopic Studiesmeasurements, not errors in comparing to of the Exhaust Plume of a Quasi-Steadydifferent profiles. Uncertainties in the MPD Accelerator", Ph.D. Thesis,4014 A, 4104 A and 4259 A emission line Department of Mechanical and Aerospacetransition probabilities are 10%, 26% and Engineering, Princeton University.6% respectively 2 2 - 2 4 These Princeton, New Jersey, 1972.comparatively low transition probabilityuncertainties should produce errors of no 2. Kilfoyle. D. B., "Spectroscopicmore than 15% in electron temperature Investigation of the Exit Plane of an MPDdeterminations; however the sensitivity of thruster", IEPC No. 88-027, 20thion/neutral ratios and electron densities to International Electric Propulsionelectron temperature will produce much Conference, Garmisch-Partenkirchen,greater errors in the determination of these West Germany. 1988.plasma properties.

3. Myers, R G. "Energy Deposition in LowPower Coaxial Plasma Thrusters". Ph.D.

V. Conclusion Thesis, Department of Mechanical andAerospace Engineering, Princeton

Ionization levels in the MPD thruster University, Princeton, New Jersey, 1989.interelectrode region have been determinedby spectroscopic investigations of a 4. Seals Jr., R K., Hassan, H. A., "Analysisspecially designed thruster. The prevalence of MPD Arcs with Nonequilibriumof argon I and argon II emission lines, from Ionization", AIAA No. 68-87. 6th Aerospacethe thruster interelectrode region, indicate Sciences Meeting, New York, 1968.that the interelectrode ionization balanceis between neutral and singly ionized 5. Shoji, T., Kimura. I., "Analytical Studyargon. Observations of ion/neutral ratios on the Influence of Nonequilibriumunder different thruster operating Ionization for Current Flow Pattern andconditions have indicated a strong scaling Flow Field of MPD Arcjets", AIAA No. 90-

with the parameter J 2 / i. Peak 2609, 21st International Electric

ion/neutral ratios near the propellant inlet

6

91-052

Propulsion Conference. Orlando. Florida. 14. Chamberlain. F. R.. Kelly. A. J..

1990. Jahn. R. G.. "Electropostitive Surface LayerMPD Thruster Cathodes", AIAA No. 89-

6. Sheppard. E., Martinez-Sanchez. M.. 2706. 25 Joint Propulsion Conference.

"Nonequilibrium Ionization in Plasma Monterey. California. 1989.

Accelerators". AIAA No. 90-2608. 21stInternational Electric Propulsion 15. Tilley. D. L.. Kelly, A. J., Jahn. R G..

Conference. Orlando. Florida. 1990. 'The Application of the Triple Probe toMPD Thruster Plumes", 21st International

7. Choueirl. E. Y.. "Electron-Ion Electric Propulsion Conference. Orlando.

Streaming Instabilities of an Florida. 1990.

Electromagnetically Accelerated Plasma".Ph.D. Thesis. Department of Mechanical 16. Randolph. T.. MAE Report No

and Aerospace Engineering. Princeton 1776.29, Princeton University. January -

University, Princeton, New Jersey. 1991. February 1990.

8. Clark K. E.. DiCapua. M. S., Jahn, R G., 17. Zaidel', A.N.. et. al., Tables of

von Jaskowsky, W. F., "Quasi-Steady Spectral Lines. IFI/Plenum. New York.

Magnetoplasmadynamic Arc 1970.

Characteristics". AIAA No. 70-1095. 8thElectric Propulsion Conference, Stanford, 18. Wiese. W. L., et. al., B. M., Atomic

California, 1970. Transition Probabilities v. II: Sodiumthrough Calcium, NSRDS-NBS 22.

9. Tahara, H.. Yasui. H., Kagaya, Y., Washington D. C.. 1969.

Yoshikawa. T.. "Experimental andTheoretical Researches on Arc Structure in 19. Pears, R. W. B., Gaydon, A. G.. The

a Self-Field Thruster". AIAA No. 87-1093, Identification of Molecular Spectra.

19th Electric Propulsion Conference, Chapman and Hall. London, 1965.

Colorado Springs. Colorado, 1987.20. Lochte-Holtgreven, W.. Plasma

10. Turchi, P. J., Davis, J. F., Roderick, Diagnostics. John Wiley and Sons Inc..

"MPD Arcjet Thrust Chamber Flow New York, 1968.

Studies". AIAA No. 90-2664. 21stInternational Electric PropulsionConference, Orlando, Florida, 1990. 21. Simpson. T. B.. "State of Ionization

and Composition of the Plasma of the MPD

11. Abramov, V. A.. et. al., Accelerator", Internal Report Electric

"Investigation of Electron Temperature and Propulsion Laboratory. Princeton

Plasma Radiation in a Quasi-Stationary University, Princeton, New Jersey, 1977.

High-Current Discharge Between CoaxialElectrodes", Proceedings of the 8th 22. Ludtke, T., Helbig, V., "Absolute

International Conference on Phenomena Transition Probabilities of the Argon II 4s-

in Ionized Gases, Vienna. Austria. 1968. 4p and 3d-4p Transition Array", Journal ofQuantitative Spectroscopy and Radiative

12. Kislov, A. Y., et. al., "Experimental Transfer. Vol. 44, No. 2, 1990.

Study of Current and PotentialDistributions Between Coaxial Electrodes 23. Rudko, R. I., Tang, C. L..

in a Quasi-Steady-State High Current Gas "Spectroscopic Studies of the Ar + Laser",Discharge", Proceedings of the 8th Journal of Applied Physics, Vol. 38. No. 12,International Conference on Phenomena 1967.in Ionized Gases, Vienna. Austria, 1968.

24. Wiese, W. L.. et. al., "Unified Set of13. Kislov, A. Y., et. al., "Distribution of Atomic Transition Probabilities forPotential in a Quasistationary Coaxial Neutral Argon", Physical Review A, Vol. 39.Plasma Injector". Soviet Physics, No. 5. 1989.Technical Physics. Vol. 13, pp. 736-738,1968.

7

91-052

J<2.5 cm>n .- 4.5 cm

Insulator

Cathode

Propellant Inlet

Anode/Body

Figure la: Thruster cutaway drawing

Anode/Body

WindowWno Cathode

Ceramic Holder

Fiberglass Thread

Figure Ib: Thruster front end drawing

8

91-052

Figure 2: Experimental layout

Photographic Shutter

. Spectrometer

Mirro Dove Prism

Entrance Slit

> Lens

Window

ThrusterPlume

Vacuum Tank

9

91-052

Figure 3 Interelec trode crljis.sionl spectrlir 39(X) A '1100 A

Insulalor

Cathode ne

rI J

4

10

.-. 941 "0

-3'~

a~i~~ruF A114104 '~

L

01 1 3 46Thruster Axial Position (Lm)

N'0

Insulator

Inlet - -_ _ _ _ _ _ _ _ _ _ _ _[ Cathode

Figure 4: Correlation of thruster geometry and graph abscisa.

1 1

91-052

1.4-

* *

1.0..

8 0.8-S[ J = 160A

J = 500 A

0.6-

0 1 2 3 4 5 6

Thruster Axial Position (cm)

Figure 5: Electron temperatures for argon propellant, mass flow rate = 5.5 mg/s.

1.4-

- . J.

1.2-

b8 0.8-

SJ = 330 A

* J = 590 A

0.6-

0 1 2 3 4 5 6

Thruster Axial Position (cm)

2 2Figure 6: Electron temperatures for argon propellant, J /mdot = 32 kA /(g/s).

12

91-052

1000-

4 . "" * J = 50160A

2. * . K . *

S *•-0 0 * . **

0 1 2 3 4 5 6

Thruster Axial Position (cm)

Figure 7: Ion/neutral ratios for argon propellant, mass flow rate = 5.5 mg/s.

4 * J = 590A

a . . *

2-

S 10081 4b 6 : . ,

4-

" "- . " ' .I . -• " . .

2 - ' . *- . ." l .. * /

,

10--I

0 1 2 3 4 5 6

Thruster Axial Position (cm)

2 2

Figure 8: Ion/neutral ratios for argon propellant, J /mdot = 32 kA /(g/s).

13