73 30 7 4 1

NASA T E C H N I C A L NASA TM X-68284M E M O R A N D U M

00NOIX

00

CASE F I L ECOPY

A 9700-HOUR DURABILITY TEST OF A

FIVE CENTIMETER DIAMETER IONTHRUSTER

by S. Nakanishi and R. C. FinkeLewis Research CenterCleveland, Ohio 44135

TECHNICAL PAPER proposed for presentation atPropulsion Specialist Conference sponsored by theAmerican Institute of Aeronautics and AstronauticsLake Tahoe, Nevada, October 29-November 2, 1973

https://ntrs.nasa.gov/search.jsp?R=19730022009 2020-03-22T22:20:57+00:00Z

A 9700-HOUR DURABILITY TEST OF A FIVE CENTIMETER DIAMETER ION THRUSTER

S. Nakanishi and R. C. FlnkeNational Aeronautics and Space Administration

Lewis Research CenterCleveland, Ohio

Abstract

A modified Hughes SIT-5 thruster has beenlife-tested at the Lewis Research Center. Thefinal 2700 hours of the test are described with acharted history of thruster operating parametersand off-normal events. Performance and operatingcharacteristics were nearly constant throughout thetest except for neutralizer heater power require-ments and accelerator drain current. A post-shutdown inspection revealed sputter erosion ofion chamber components and component flaking ofsputtered metal. Several flakes caused beamlet di-vergence and anomalous grid erosion, causing thetest to be terminated. All sputter erosion sourceshave been identified and promising sputter resist-ant components are currently being evaluated.

Introduction

Electron bombardment ion thrusters have poten-tial application for spacecraft attitude controland station-keeping.(1>2) Primary requirements forsuch thrusters are reliability and durability overmany thousands of operating hours.(3,4) Thrustvectoring capability is also desirable for manyapplications.

A 5-cm diameter thrust vectorable structurallyintegrated ion thruster (SIT-5) system was designedand developed by the Hughes Research Laboratoriesunder NASA Contracts NAS3-14129 and NAS3-14058.(5£)A modified SIT-5 thruster of a first generation de-sign has been tested at the Lewis Research Center.The purpose of the test was to determine the ulti-mate life of the thruster and to pinpoint problemareas not discernible in short term tests.

This paper presents the operational resultsspanning approximately the final 2700 hours of thedurability test. The initial 2000 hours of thetest with results of inspection and the ensuing5000 hours of continuous operation have been re-ported previously.(7>8) The final segment of the9715 hour test is described in a detailed opera-tional history of several thruster and neutralizerparameters. Performance profiles and thrustercharacteristics mapped at 6130 and 9030 hours willbe compared. Some of the findings of the post-shutdown inspection are included.

Apparatus and Procedure

A photograph of the modified SIT-5 thrusterinstallation is shown in figure 1. The specificmodifications described in detail elsewhere(7,8)included removal of the propellant reservoir, repo-sitioning of the neutralizer, and change of accel-erator optics. The electrostatic thrust vectorgrid was installed after the thruster had operated2027 hours with a translating-screen type vectorgrid. At that time, the thruster was disassembledfor photographic documentation. No refurbishingor repairs were made which would have compromisedthe legitimate continuation of the test and accu-

mulation of test hours on the thruster.

Figure 2 shows an interior view of the verti-cal vacuum tank before the test. The tank is1.37 m in diameter and 1.83 m tall. Mercury con-tained in the stainless steel pan was frozen byliquid nitrogen flowing through coils brazed toradial copper struts attached to the bottom of thepan. A cylindrical cooling coil brazed to verticalcopper strips formed the LN2 cryowall which ex-tended along the vertical wall of the tank.

The accelerator grid of the thruster was ap-proximately 75 cm above the frozen mercury targetwhen installed. To prevent backsputtering of con-densible conductive material upon the thruster, aset of nonmetallic (50% A1203, 50% Si02) baffleswere installed. No metallic surface other than themercury target intercepted the line of sight of thethruster ion beam. With the target frozen and thecryowall at liquid nitrogen temperature, the ulti-mate tank pressure reached the mid 10"? torr range.With the thruster in operation, the pressure roseto about 1x10-6 torr.

Details of the electrical system and test pro-cedures are described elsewhere.(7) An automaticdigital data acquisition system recorded thrusterdata at clocked intervals. A protective controlsystem designed to shut down the thruster in theevent of abnormal thruster or test facility condi-tions allowed unattended operation around theclock. The thruster was operated open-looped atnominally fixed conditions except for minor day-to-day adjustments. The ion chamber and neutralizerpropellant flow rates were held at about 34 mA and2.2 mA Hg+, respectively, throughout the test.Performance mapping data were obtained at approxi-mately 1000 hour intervals. A four-channel strip-chart recorder was used to chart the neutralizerparameters continuously.

Results and Discussion

History of Test

A history of the test from 7000 thruster hoursto the end at 9715 hours is shown in figure 3. Theelectrostatic vector grid was installed at 2027thruster hours so that the number of hours on thegrid is less than the thruster operating time bythat amount. The four parameters charted werethose most subject to change with time. As de-scribed previously,(8) other thruster parameterseither remained nearly constant throughout the testor were adjusted to fixed operating values.

The accelerator drain current shown in fig-ure 3(a) was about 0.09 mA from 2027 to 8000 thrus-ter hours. From 8000 to 9500 hours the drain cur-rent remained at about 0.10 mA then suddenly in-creased to more than 0.20 mA. Details of the final270 hours of the test when numerous grid shortsoccurred will be shown later.

The cathode keeper voltage (fig. 3(b)) was (

constant to the end of the test at a value whichprevailed from about 4000 hours.W Actually, thecathode keeper voltage first attained an operatinglevel of 12-13 volts as early as 400 hours into thetest. The voltage subsequently rose to about15 volts and gradually dropped to 12.5 volts by4000 hours. No heater power was applied to thecathode except during startups.

The neutralizer keeper voltage and floatingpotential plotted in figures 3(c) and (d) graduallyrose with time. The irregular variations in theseparameters were induced primarily by the changes inneutralizer heater power made during the test tostabilize neutralizer operation. Continuous stripchart records of neutralizer parameters to be pre-sented later will show the nature of some neutra-lizer transients.

A consistent trend associated with neutralizerheater power can be seen in figure 3(c). Increas-ing heater power decreased the keeper voltage andfloating potential. The decrease was followed bya gradual rise at approximately the same rate asthe overall average rise with time.C8) This riserate was about 1 volt per 1000 hours of operation,and it occurred even in the early hours of the testwhen heater power was 3 watts or less. The conclu-sion is that the mechanism causing the rise inkeeper voltage was not entirely dependent on heaterpower and neutralizer temperature.

Noted on the test history are four types ofrecurrent abnormal events. Neutralizer outage pre-cluded beam current operation because of highlynegative neutralizer floating potential, but theprotective control system shut off only the ionchamber discharge and the accelerating potentials.The cathode was permitted to continue operating inorder to facilitate restarts.

High neutralizer floating potential also oc-curred without a neutralizer outage. This abnormalcondition was rare in the first 7000 hours of test.It was especially numerous between 7100 and 7800hours but nonexistent thereafter. The cause ofthese occurrences is not known. An excerpt fromthe continuous strip chart record of neutralizerparameters will be presented later to show detailsof this abnormal condition.

Shorting between vector grid elements occurredfour times between 8000 and 9000 hours. Theseshorts occurred at a rate of 2 per 1000 hours ofoperation between 6000 and 7000 hours but were non-existent prior to that period.

Screen to accelerator shorts were less numer-ous although one had occurred as early as 5000hours into the test. In all instances, the shortswere cleared successfully either by application ofvectoring voltage or by a capacitor discharge. Acapacitor discharge usually quenched the neutra-lizer (in a manner similar to events early in thetest(7) before installation of inductive ballastsin the high voltage circuits).

Terminal history. A history of the final seg-ment of the durability test is shown in figure 4to an expanded time scale. An uneventful 110-hoursegment between 9580 and 9690 hours has beenomitted for brevity.

The drain current, neutralizer keeper current,and neutralizer floating potential only are shownbecause no long-term trends in ion chamber param-eters were discernible in this short time span.Thruster shutdowns occurred in close succession at9518, 9526, and 9534 hours because of screen-accelerator shorts. The accelerator drain current(fig. 4(a)) increased about 0.05 mA after each ofthe first two events. After the third screen-accelerator short was cleared, the drain-currentdropped 0.07 mA. It is believed that flakes ofsputtered metal shed from ion chamber surfaces weredefocusing the ions passing through partiallyblocked screen apertures. At worst, the first ofgrid element failures may have occurred at thispoint in the test. During the final 181 hours ofthe test, only one neutralizer-outage event oc-curred. The accelerator drain current continued tofluctuate with a gradual decline in the mean level.The test ended when a vector grid element cutthrough by a defocused beamlet dropped against theground screen mask to cause a permanent short.

The neutralizer keeper current and floatingpotential are shown in figures 4(b) and (c). Thethruster shutdown at 9518 hours was preceded bytwo hours of unsteady neutralizer operation. Thecontinuous strip chart record showed that at9516 hours, the neutralizer keeper current droppedto 0.2 A and the vaporizer temperature dropped17° C. The neutralizer keeper voltage and floatingpotential increased about 10 volts and containedfluctuations varying in amplitude from 1 to 5 volts.There was no apparent connection between the neu-tralizer disturbances and the thruster shutdown at9518 hours. Subsequent shutdowns caused by screen-accelerator shorts occurred during normal neutra-lizer operation. A neutralizer outage occurred at9555 hours. From that point to the end of testthe neutralizer operated at nearly constant condi-tions without interruption.

Neutralizer transients. A selected portion ofthe continuous strip chart recording of neutralizerparameters is shown in figure 5. Each divisionalong the paper travel corresponds to 3.75 minutes.The four channels from left to right are neutra-lizer floating potential * 0.1 V, neutralizerkeeper voltage x 0.1 V, neutralizer keeper currentx 10 A, and neutralizer vaporizer temperature(chromel-alumel thermocouple, M7 above 65.6° C).The operating time in hours is noted along the leftmargin.

The abnormal condition which occurred at about7666 hours was a series of high floating potentialspikes which did not persist long enough to acti-vate the protective control circuit. The relativetime positions of the four pens which are staggeredto permit cross-over along the chart width aremarked on each channel. At the chart speed used,however, the parameter leading the change cannot beresolved.

It is of interest to consider the power andthermal balance during this disturbance. From theinitial steady-state conditions, the neutralizerkeeper voltage decreased an average of 4 volts andthe keeper current increased 0.03 A for a net de-crease of 0.83 watt in neutralizer discharge power.The beam current and vaporizer heater power wereconstant throughout the transient. Neglecting theeffects of change in neutralizer floating poten-

tial, the 0.8 watt decrease in neutralizer dis-

charge power appears too small to justify a 13° Cdrop in vaporizer temperature. Similar transientsin which the floating potential varied only 5 voltsnevertheless showed the same amount of vaporizertemperature drop. Intentional variations in neu-tralizer heater power of the order of 3 watts withsmall variations in keeper voltage at constantkeeper current and constant vaporizer power changedthe vaporizer temperature only about 1° C. Itappears that the relatively large change in vapor-izer temperature observed in figure'5 was caused bymechanisms other than the changes in externallysupplied power. Also, sudden changes in bothkeeper voltage and floating potential occurred andthen corrected themselves spontaneously.

Performance Map

The durability test was run at a fixed operat-ing point and normally did not permit an examina-tion of thruster characteristics off the set point.Limited mapping of thruster characteristics wasperformed at approximately 1000 hour intervals toobserve any changes in these characteristics withoperating time. The results of performance mappingat 6130 and 9030 hours are presented herein. Anearlier mapping at 4250 hours is reported in ref-erence 8.

Cathode keeper current. The effects of vary-ing the cathode keeper current while maintaining aconstant beam current of 25 mA is shown in fig-ure 6. The levels and trends in the keeper voltage(fig. 6(a)) and the ion chamber discharge current(fig. 6(b)) at 6130 and 9030 hours were similar.The cathode keeper voltage reached a maximum at akeeper current between 0.35 and 0.4 A. A corre-sponding minimum in the discharge current occurredat a slightly lower value of keeper current.

The discharge voltage shown in figure 6(c) wasgenerally lower at the 9030 hour point. Becausedischarge voltage was inversely very sensitive topropellant flow rate, it is apparent that the flowwas higher at the later time. At both times, how-ever, the trends in discharge voltage as a functionof keeper current were similar. The maxima in thedischarge voltage corresponded to the minima in thedischarge current. Because the beam current washeld constant, it follows that the minimum dis-charge current was required where the dischargevoltage, and hence the ionization, was a maximum.

Discharge current. The effects of varyingdischarge current are shown in figure 7. Thecathode keeper current was maintained at 0.3 A.No cathode heater power was used. The beam cur-rent characteristic curves (fig. 7(a)) at 6130 and9030 hours were almost identical except for a morerapid fall-off in the later curve at discharge cur-rents below 0.4 A.

The accelerator drain current shown in fig-ure 7(b) decreased with increasing discharge cur-rent. This effect is attributable to less chargeexchange ions as beam current and propellant utili-zation efficiency increased. The higher drain cur-rent at 9030 hours was probably due to a slightlyhigher propellant flow rate.

The discharge voltage shown in figure 7(c) wasabout 1.5 volts lower at the 9030 hour point be-

cause of the higher propellant flow rate. At dis-

charge currents less than 0.4 A, the drop in dis-charge voltage below 37 volts and the loss inionization probably caused the rapid fall-off inbeam current (fig. 7(a)). In contrast, the dis-charge voltage at the 6130 hour point was 37 voltsor higher for discharge currents as low as 0.25 A.

Accelerating potentials. The variations indischarge parameters and accelerator drain currentwith changes in the net accelerating voltage (i.e.,changes in specific impulse) and the acceleratorpotential were relatively minor and consistent withtrends reported previously. As ion extraction im-proved with increasing total accelerating potentialdifference, the discharge current and voltage re-quired to maintain the design beam current de-creased slightly. Accelerator drain current re-mained nearly constant or increased slightly withincreases in either the net accelerating or accel-erator potentials.

Neutralizer keeper current. The gradual risewith time in keeper voltage and neutralizer floatingpotential for fixed point operation has been shownin the test history (fig. 3). The effects of vary-ing neutralizer keeper current are shown in fig-ure 8. The keeper voltage generally decreased askeeper current was increased (fig. 8(a)). Increas-ing the heater power also reduced the keeper volt-age , but the effects were dependent on the numberof operating hours. At 6130 hours, the neutralizercould be operated with no heater power. At 9030hours, this was no longer possible. At 7100 hours,10 watts of heater power reduced the keeper voltagebelow the 6130 hour values. With 10 watts ofheater power at 9030 hours, the keeper voltage wasstill higher than at 6130 hours with no heaterpower.

It is interesting to note that at 9030 hoursthe neutralizer could be operated at a given levelof keeper voltage if the 5 watts reduction inheater power were compensated by a comparable in-crease in keeper discharge power. This balance in-dicates that some minimal temperature is requiredfor neutralizer operation at a given keeper voltagebut that the required temperature varies with oper-ating time.

Although the basic design was identical, thethruster cathode showed essentially no change withtime (fig. 3), and it was operable throughout thetest with no heater power. The one major differ-ence between the neutralizer and the cathode wasthe level of mercury flow rate. The cathode oper-ated in the region of 33-35 mA equivalent; the neu-tralizer, in the region of 1.7 to 2.5 mA.

The corresponding variations in neutralizerfloating potential are shown in figure 8(b). Thefloating potential increased rapidly as keepercurrent was reduced below a critical "knee" value.The "knee" value of keeper current was also de-pendent on heater power level and operating time.Higher heater power moved the knee to lower valuesof keeper current. For most of the test, thekeeper current was held at 0.45 A which was wellabove the critical value for all conditions. Ingeneral, the combination of heater power level andoperating time which affected keeper voltage hadsimilar effects upon the floating potential.

Performance Profile

Performance profiles and thruster operatingparameters at 6130, 9030, and 9700 hours are givenin table I. The relatively constant thruster oper-ation indicated by the test history is reflectedin the power values. Beam power, discharge power,accelerator drain, and total cathode power werenearly constant throughout the test.

The major change in total input power withtime was due to the increase in neutralizer heaterpower required to permit unattended stable opera-tion of the neutralizer. Because of thermal feed-back to the vaporizer, higher heater power was par-tially offset by a lower vaporizer power require-ment to maintain a constant temperature. Total in-put power thus increased about 9 watts by the endof test to cause a 5 percent drop in power andoverall efficiency. Specific impulse variedslightly because of differences in total propellantutilization efficiency. Power to thrust ratio in-creased by about 20 watts per mlb.

Post-Shutdown Inspection

Electrostatic vector grid. Figure 9 is aphotograph of the thruster in vertical positiontaken immediately after retraction from the vacuumtest facility. The broken vector grid elementwhich caused the decision to terminate the test isin visible contact with the opening in the groundscreen mask. The ground screen and adjacent out-side structures were remarkably clean and free ofany loose metallic deposits.

A photograph of the vector grid taken from thescreen or ion chamber side is shown in figure 10.Care was taken to remove the grid assembly in avertical position to minimize jarring and disturb-ing the flakes. Flakes can be seen lying in theperipheral regions of the screen grid and the mag-netic pole piece.

A comparison of the flake locations and theeroded vector grid elements indicated that anoma-lous erosion was caused by flake-diverted beamions. A more gradual type of erosion was caused byhighly divergent ions from each beamlet hole andparticularly from the 12 outermost holes.

Figure 11 is a photograph taken from the ac-celerator side of the vector grid. The broken gridelement has been removed. Five grid elements wereeroded through, but only one element (shown re-moved) dropped completely out of position. Theothers were either restrained from falling by thecharge-exchange protective shields or held by theorthogonal elements lying cross-wise above them.

Numerous darkened areas outside the hole pat-tern can be seen on the face of the screen grid.These were arc sites formed where capacitor dis-charges were used to clear a short between thescreen and accelerator. Sputtered metal, probablyfrom the neutralizer, was deposited in those re-gions not shielded by the ground mask.

Neutralizer. A close-up photograph of theneutralizer taken after the test is shown in fig-ure 12. The double layer of 0.0127 mm thick tanta-lum foil used as an erosion rate detector marksthe approximate boundary of the intercepted ionbeam. The boundary lay on a 45 degree line drawn

from the trailing edge of the outermost beam holegrid element. The tantalum foil eroded at a rateof only 13 microns per 500 to 600 hours of unde-flected beam operation. The maximum depth of ero-sion on the neutralizer keeper cap is estimated tobe 250 to 300 microns at the corner which projectedmost deeply into the ion beam.

A photograph of the neutralizer hollow cathodeis shown in figure 13. Erosion occurred at theaperture and the chamfered surface of the tip. Anenlarged drawing of the tip disk sectioned along adiameter is shown in figure 14. The dotted outlineshows the aperture size and the chamfered surfacewhen new. During 9715 hours of operation the aper-ture eroded from 0.254 to 0.549 mm at the neck.

Preliminary data taken by A. Weigand of LeRChas indicated that neutralizer tip erosion can bea strong function of neutralizer flow rate. Simu-lation of total depletion of the insert's emissivecoating by operating cathodes without any insertor emissive coating has shown that erosion of thetip will take place at the flow rates used duringthis endurance test. Weigand"s data has indicatedthat for 5 cm neutralizer flow rates at or above6.9 mA the tip erosion rate is reduced by over2 orders of magnitude even if the insert has beentotally depleted. Thus neutralizer operation atslightly higher flow rates, with a small penaltyon propellant utilization efficiency, appears tobe an attractive operating mode to prevent exces-sive tip erosion.

Neutralizer insert. The neutralizer insertwas examined for possible correlation between neu-tralizer operation and insert condition. Fig-ure 15 is a photograph of the insert after it wasunrolled and stretched out. Parts of the down-stream edge (nearest the neutralizer tip) of theinsert appeared to have reached high enough temper-atures to cause embrittlement and erosion. Theseregions were either sintered layer-to-layer or sobrittle that crumbling occurred. The upstream por-tion of the insert remained intact and springy. Asmall rod was required to keep the insert from re-coiling.

Both the upstream and downstream regions ofthe insert were examined with an optical microscopeand an electron scanning microscope with elementalidentification by an energy dispersive analyzer.Barium was detected generally in the upstream re-gion, but the downstream edges were practicallydevoid of barium except in small isolated or shel-tered regions. The conclusion is that an ampleamount of the barium emissive mixture was left inthe insert, but it was not available or present inlocalized areas where required for good neutralizeroperation.

Cathode pole assembly and thruster back plate.A photograph of the cathode pole assembly andthruster back plate which were exposed to the ionchamber plasma is shown in figure 16. The extentof erosion can be seen by comparison with figure 17which is a photograph of the same component after2027 hours of operation. The magnification washigher so that the initial stages of erosion couldbe seen around the screw threads and the hexagonalcorners of the nut.

After 9715 hours, all but 0.009 gram of thenut (original estimated weight, 0.099 gm) was

eroded. Only two threads of the 0-80 stainlesssteel screw remained above the baffle. The baffleitself was only slightly eroded at the downstreamperipheral edge.

Slight erosion was found on the thruster backplate. Four radial traverses were made with a sur-face analyzer to determine the depth of surfaceerosion. Although not axially symmetrical themaximum erosion was about 8 microns at a radius ofabout 2 cm.

J. Power and E. Wintucky of LeRC have proposedthe use of pyrolytic graphite coatings over cathodepotential surfaces to reduce the sputtering ofthese components. The low sputter yield of carbonshould greatly reduce the material loss rates fromcathode potential surfaces. Tests of pyrolyticgraphite baffles and thruster back plates are cur-rently being performed to evaluate these sputterresistant components.

Anode. A photograph of the ion chamber in-terior looking from the cathode end is shown infigure 18. The anode showed no sign of erosion.Peeling of sputtered metal which formed the flakeswas evident around most of the upstream end of theanode. It is believed that the main source ofsputtered metal was the baffle assembly and thethruster back plate. The flakes which peeled offwere typically 0.021 mm thick and magnetic.

The exact cause of peeling is not known.Thermal cycling because of thruster shutdowns, themismatch of expansion coefficients, exposure toatmospheric moisture at 2027 hr, and the mechani-cal stability limit of sputtered material probablyall contributed to the loosening of an alreadyweakly bonded layer.

It has been proposed that an opaque finelywoven screen anode would retard the peeling off ofsputter deposited material on the anode. Concep-tually, the sputter deposited films would havethickness variations around each wire of the wovenanode. At points where two wires cross or contacteach other the film thickness should be near zero.As a result of the woven anode structure, allsputtered film patches are surrounded by near zerothickness films. Thus, even if a sputter depositedfilm cracks or peels in the region of thickest dep-osition, the flakes should remain attached at theplaces where the film is thin and strongly adher-ent. If the screen anode is finely woven the dis-tances between maximum and minimum film thicknessesare small, thus reducing the probability of flakestotally detaching from their thinner surroundingregions. Tests of the screen anode are being madeto determine its effectiveness in the preventionof loose flakes. The combination of reduction insputter rate of the cathode potential surfaces andthe peel resistant anode could prevent formation ofloose flakes.

Concluding Remarks

A five centimeter diameter mercury bombard-ment ion thruster has been life tested for9715 hours. The test was interrupted once at2027 hours to replace the translatable screenthrust vector grid with an electrostatic type.

During the final 2715 hours of the test, thecathode and ion chamber discharge characteristics

were essentially the same as those observedearlier in the test. Comparison of thruster per-formance profiles taken during the lifetest showedthat overall efficiency dropped 4% and power tothrust ratio increased 19 w/mlb. The principalcause of these changes was the increase in neutra-lizer heater power required for stable operation.Proposed solutions to avoid neutralizer perform-ance changes are presently being life tested.

Accelerator drain current increased signifi-cantly only after sputtered metal flakes shed fromanode surfaces fell upon the screen grid of theion accelerator system. Beam diversion from par-tially blocked screen holes increased direct ionimpingement and eventually eroded thrust vectoringgrid elements. Electrical shorting of the gridincreased in number and frequency as the test pro-gressed. All shorts were successfully clearedeither by steady application of power or by acapacitor discharge.

Post-test inspection showed varying degreesof erosion of the grid elements, neutralizer tip,cathode baffle assembly, and thruster back plate.Examination of the neutralizer insert showed de-pletion of the barium emissive mixture at thedownstream edge nearest the neutralizer tip. Sub-stantial amounts of barium were found on all up-stream areas of the insert.

Except for the electrostatic type vector grid,no other thruster component indicated imminentfailure at 9715 hours.

Table 1 Performance profile

Test hoursNet accelerating potential, VAccelerator potential, VBeam current, mACathode flow rate, mA equivNeutralizer flow rate, mA equtoBeam power, wDischarge power, wComponent powerAccelerator drain, wCathodeHeater, wKeeper discharge, wVaporizer, wTotal cathode power, w

NeutralizerHeater, wKeeper discharge, wVaporizer, wTotal neutralizer power, w

Total input power, wPower efficiency, %Utilization efficiency, %Overall efficiency, %Thrust, mlbSpecific impulse, secP/T ratio, w/mlb

61301300-90025

32.12.232.515.9

0.18

90301300-90025

33.02.232.516.6

97001300-90025

33.42.432.516.3

0.2 0.28

04.375.539.90

2.408.901.1612.4670.945.973.033.50.412630173

03.765.168.92

9.9010.900.3221.1279.341.071.128.10.412560193

03.785.769.54

9.909.980.4220.3078.941.269.828.80.412515192

REFERENCES

1. Owens, W. L., Jr., "Optimization of Ion Propul-sion for North-South Stationkeeping of Com-munications Satellites," Paper 72-1150, Nov.1972, AIAA/SAE, New Orleans, La.

2. Hudson, W. R., and Banks, B. A., "An 8 cm-Electron Bombardment Thruster for AuxiliaryPropulsion," Proposed Paper, Oct.-Nov. 1973,AIAA, Lake Tahoe, Nev.

3. Isley, W. C., and Duck, K. I., "Propulsion Re-quirements for Communications Satellites,"Paper 72-514, Apr. 1972, AIAA, New York, N.Y.

4. Free, B., and Huson, G., "Selected ComparisonsAmong Propulsion Systems for CommunicationsSatellites," Paper 72-517, Apr. 1972, AIAA,New York, N.Y.

5. Hyman, J., Jr., "Design and Development of a SmallStructurally Integrated Ion Thruater System,"NASA CR-120821, Oct. 1971, Hughes ResearchLabs., Malibu, Calif.

6. King, H. J., Collett, C. R., and Schnelker, D. E.,"Thrust Vectoring Systems. Part I: 5-cm Sys-tems," NASA CR-72877, 1970, Hughes ResearchLabs., Malibu, Calif.

7. Nakanishi, S., and Finke, R. C., "A 2000-HourDurability Test of a 5-Centimeter-Diameter Mer-cury Bombardment Ion Thruster," TM X-68155,1972, NASA, Cleveland, Ohio.

8. Nakanishi, S., "Durability Tests of a 5-Centimeter-Diameter Ion Thruster System," Paper72-1151, Nov. 1972, AIAA/SAE, New Orleans, La.

C-72-305

Figure 1. - Test installation of modified SIT-5 thruster.

W

C-71-1845

Figure 2. - Tank interior before test.

LEGEND OF EVENTS

A. NEUTRALIZER OUTAGEB. HIGH NEUTRALIZER POTENTIAL

z£ -2°r*5! C^ <

Q ^ S nl0

s « a 2or° s <

NEUTRALI

o: [~ 10LLJ »M o; g 3

- 10

-40,-

fei> -20-

Q

LEGEND OF EVENTS

A. NEUTRALIZER OUTAGEB. HIGH NEUTRALIZER FLOATING POTEN1C. VECTOR GRID SHORTD. SCREEN-ACCEL SHORT

J

_^__^___^___ \i i i

(a) ACCELERATOR DRAIN CURRENT.

i i i

(b) CATHODE KEEPER VOLTAGE.7ER HEATER POWER, W

7 7 clOi,7^10

—r!0 r.lQ^rlMfr.Si. 10j •> A .11?

1f J ]i5p

i i

(c) NEUTRALIZER KE

D Dp _ ^ Df\ 0 r1 1 B \|rDr I L

i

|m L_y "V1 II

EPER VOLTAGE.

i cc cc

t^s ^-v "f

1

6.4 7.0 8.0 9.0THRUSTER OPERATING TIME, HR

(d) NEUTRALIZER FLOATING POTENTIAL.

Figure 3. - History of test.

r,AL . '40

1 -30OL

ol -20

i .10u

0

, .5

.4

2£~ '3

P '.1

0

^ -4Q

1 1 -30s

_i > -2010.0X103 z

1 -10

£ o95

-

— •

1

V.

D

1 i

SCREEN-ACCEL SHORT

1 1 |_yi| 1 1V

(a) ACCELERATOR DRAIN CURRENT.

1

H!

D

i

b)N

D

— ̂

\

1 I

EUTRALIZE

D i

i i

1 1 1,1 1 1

R KEEPER CURRENT

1 1 1 A 1 1 1

00 10 20 30 40 50 60 70 9580" 9690 9700 10 20THRUSTER OPERATING TIME, HR

(c) NEUTRALIZER FLOATING POTENTIAL.

Figure 4. -Terminal history of test.

£-7624

TIME

7667 HR -

7666 HR

CHANNEL IDENTIFICATION

VG NEUTRALIZER FLOATING POTENTIALVNK NEUTRALIZER KEEPER VOLTAGEJNK NEUTRALIZER KEEPER CURRENTTNV NEUTRALIZER VAPORIZER TEMP.

RELATIVE"PEN"LOCATIONS

vGxO.1V

I I

VNKxO. IV

i

NKxlOA

'NVMVFROM65.6°-REFERENCE

1 3 8_)10

Figure 5. - Strip-chart record of neutralizer parameters.

IffiLLJLUi<i

Oct:

o Qi

o

o:

xo

o

14

12

10

TEST HOURS

° 6130o 9030

< .6

0 0 . 4

40

38

36

(a) CATHODE KEEPER VOLTAGE.

(b) DISCHARGE CURRENT.

I34.1 .2 .3 .4

CATHODE KEEPER CURRENT,

(O DISCHARGE VOLTAGE.

.5 .6

Figure 6. - Effects of varying cathode keeper current at25 mA beam current and no heater power.

ct:

=>o

28 r-

24

20

16

.12

O

Qi

Ot/1

TEST HOURS

a 6130Q 9030

(a) BEAM CURRENT.

-.10LU2Di3O

^ .08<co:

.06(b) ACCELERATOR DRAIN CURRENT.

42,-

40

38

36

34.2 .3 .4 .5

DISCHARGE CURRENT, A

(c) DISCHARGE VOLTAGE.

.6

Figure 7. - Effects of varying ion chamber dis-charge current. Cathode keeper current,0.3 A.

TEST HEATERHOURS POWER,

W

32 ,—

IW (a) NEUTRALIZER KEEPER VOLTAGE.

-50,-

-40

g -30o

-20

-10 I.1 .2 .3 .4 .5

NEUTRALIZER KEEPER CURRENT, A

(b) NEUTRALIZER FLOATING POTENTIAL.

Figure 8. - Effects of varying neutralizer keepercurrent.

Figure 9. - SIT-5 thruster after test.

C-72-4280

Figure 10. - Electrostatic vector grid after test showing metal flakes onscreen electrode.

E-7624

C-72-4286

Figure 11. - Electrostatic vector grid after test.

C-72-4262

Figure 12. - Neutralizer after test.

C-72-4276

Figure 13. - Neutralizer hollow cathode after 9715 hours.

NOTES: DOTTED LINE IS OUTLINE WHEN NEWALL DIMENSIONS IN MILLIMETERS

Figure 14. - Cross-section of neutralizer tip disk.

vOt*

W

V DOWNSTREAM EDGE

liiimmf

INSIDE END OF /INSERT COIL -I

C-73-991

Figure 15. - Unrolled neutralizer insert after 9715 hours.

CM0 1

JlHIIIIIll

C-72-4274

Figure 16. - Cathode pole assembly and thruster back plate after 9715hours.

-72-228

Figure 17. - Cathode pole assembly after 2027 hours.

C-72-4287

Figure 18. - Anode after 9715 hours.

NASA-Lewis