NASA Technical Memorandum 106307

AIAA-93-2392

///'t _

/ _3/7'2---

Low Power Arcjet System Spacecraft Impacts

Eric J. Pencil and Charles J. Sarmiento

Lewis Research Center

Cleveland, Ohio

and

D.A. Lichtin, J.W. Palchefsky, and A.L. Bogorad

Martin Marietta Astro Space

Princeton, New Jersey

Prepared for the29th Joint Propulsion Conference and Exhibit

cosponsored by the AIAA, SAE, ASME, and ASEE

Monterey, Califomia, June 28-30, 1993

(NASA-TM-106307)SYSTEM SPACECRAFT

33 19

L_W PqW_R ARCJET

IMPACTS (NASA)

G3/20

N94-17856

Uncl as

01931?2

https://ntrs.nasa.gov/search.jsp?R=19940013383 2018-07-15T00:23:06+00:00Z

LOW POWER ARC JET SYSTEM SPACECRAFT IMPACTS

Eric J. Pencil* and Charles J. Sarmiento**NASA Lewis Research Center

Cleveland, OH 44135

D. A. Lichtint, J. W. Palchefskytt, and A. L. BogoradtttMartin Marietta Astro Space

Princeton, NJ 08543

Abstract

Application of electrothermal arcjets on communicationssatellites requires assessment of integration concerns identifiedby the user community. Perceived risks include plumecontamination of spacecraft materials, induced arcing orelectrostatic discharges between differentially charged spacecraftsurfaces, and conducted and radiated electromagnetic interference(EMI) for both steady state and transient conditions. A Space Actagreement between Martin Marietta Astro Space, the RocketResearch Company, and NASA's Lewis Research Center wasestablished to experimentally examine these issues. Spacecraftmaterials were exposed to an arcjet plume for 40 hours,representing 40 weeks of actual spacecraft life, and contaminationwas characterized by changes in surface properties. With theexception of the change in emittance of one sample, allmeasurable changes in surface properties resulted in acceptableend of life characteristics. Charged spacecraft samples werebenignly and consistently reduced to ground potential duringexposure to the powered arcjet plume, suggesting that the arcjetcould act as a charge control device on spacecraft. Steady stateEMI signatures obtained using two different power processingunits were similar to emissions measured in a previous test.

Emissions measured in UHF, S, C, Ku and Ka bands obtained a nullresult which verified previous work in the UHF, S, and C bands.Characteristics of conducted and radiated transient emissionsappear within standard spacecraft susceptibility criteria.

Nomenclature

A Area of solar cell array, 128era2

AF Antenna factor, dB/mBB Broadband emission level at 1

meter, dBI_V/m/MHzBNF Bandwidth normalization

factor, dBI:L Cable loss, dB

Cs Simulated solar constant,

137.2 mW/cm2FF Fill factor, %

Isc Short circuit current, mA

I(Z.) Solar spectral irradiance,mW/(cm2 _tm )

NB Narrowband emission level at

1 meter, dB/gV/m

Pinput Input power from solarsimulator, mW

* Aerospace Engineer, On-Board Propulsion Branch, Member AIAA.** Electrical Engineer, On-Board Propulsion Branch.t Principal Hardware Design Engineer, Member AIAA.tt Principal Development Engineer.ttt Senior Development Engineer.Copyright © 1993 by the Martin Marietta Corporation. Published by the AmericanInstitute of Aeronautics and Astronautics, Inc. with permission.

Pmax Maximum solar cell power,mW

R Resistance, ohmsRB Spectrum analyzer resolution

bandwidth (3 dB Gaussian),MHz

Voc Open circuit voltage, mV

a AbsorptanceEmittance

11 Solar cell efficiency, %_, Wavelength, ;lmp Reflectancep(_) Reflectance spectral

response4_ Spectrum analyzer displayed

voltage, dBgV or gV

|ntroductiQn

Electric propulsion users areconcerned with several integrationissues including conducted andradiated heat fluxes, plumemomentum and contaminationimpacts, conducted and radiatedelectromagnetic interference,transmission interference, and

spacecraft charging. In addition tocomponent development of a 1 kWclass hydrazine arcjet system,l,2,3efforts have focused on resolvingintegration issues associated withnorth-south stationkeepingapplications on geosynchronouscommunications satellites.

In the recent past, researchwas directed toward examiningcharacteristics of the arcjet plume.Langmuir probe surveys of thearc jet plume characterized electronnumber densities and temperaturesfor various nozzle geometries andoperating conditions.4,5. 6 It wasfound that the 1 kW class arcjetplume is weakly (less than 1%)ionized.4 These profiles were thenused in a source flow model7 for

estimates of the impact on signalstransmitted through the far-fieldplume regions simulating realisticpropagation paths. The plumeeffect on antenna performance wasminimal for the 1 kW class arcjet.7,s

An experimental study ofspacecraft compatibility of anoperational arcjet system wasperformed by TRW under contractto NASA. A flight-type arcjetsystem was mounted on aFLTSATCOM qualification modelsatellite in a large vacuumc h am b e r. 9 Measurement of theradiated and conducted

electromagnetic emissions revealedthat radiated emissions from the

arcjet and its power processor werewithin acceptable limits above 500MHz. A low frequency broadbandsignature exceeded the MIL-STD-461C limit below 40 MHz. Since

communications satellites typicallytransmit in higher frequencyranges, communications would notbe affected by these emissions.Satellite telemetry was monitoredduring arc jet ignition and nosignificant changes in signals werenoticed. An array of calorimeterslocated at a distance of 1.8 m and 2.3m from the thruster exit measured amaximum heat flux of 0.18 suns. No

visible degradation or massdeposition was observed on witnessplates placed at various locations inthe plume.

A 1.8 kW hydrazine arcjetsystem has been baselined fornorth-south stationkeepingapplication on the Martin MariettaSeries 7000 communicationssatellites. In order to address

residual user integration concerns,an integration test was performedunder a Space Act agreementbetween Martin Marietta Astro

Space, formerly General Electric'sAstro Space Division; the RocketResearch Company (RRC); andNASA's Lewis Research Center

(LeRC). Tests were performed in a

large vacuum facility at LeRC usinga flight-type arcjet system.Spacecraft material samples andsome test instrumentation were

provided by Martin Marietta andone of two power processing units(PPU's) was provided by RRC, the

2

arcjet system contractor, andPacific Electro Dynamics (PED), thePPU subcontractor. In this test,

plume contamination on spacecraftmaterial samples, electrostaticdischarge phenomena, andelectromagnetic compatibility wereinvestigated 10.

Potential contamination of

spacecraft surfaces wasinvestigated by positioningspacecraft material samples relativeto the arcjet thruster simulating asatellite configuration. Duration ofthe exposure was 40 hours whichrepresented approximately 40weeks of satellite lifetime or 6% of

the total thruster operation life.Contamination impacts werequantified via changes in thesurface properties of the spacecraftmaterials. Surface properties ofabsorptance, emittance, andresistance were measured for the

material samples. Effects on asilicon solar array were quantified

by measuring the initial and finalcurrent-voltage traces.

Another concern was the

possibility that ignition of an arcjetand the resultant formation of the

weakly ionized plasma plume wouldinduce an elecrostatic discharge

(ESD) between charged surfaces and

ground. Spacecraft surfacesexposed to the ambient conditionsin geosynchronous orbit can buildup multi-kilovolt potentialdifferenences between spacecraftsurfaces or between spacecraftsurfaces and spacecraft ground.1 t

If the charging voltage exceeds thebreakdown threshold, an

electrostatic discharge, or spark,can occur. ESD's can causeinterference in spacecraft

electronics ranging from simplelogic switching to complete systemfailures. Significant long-termdegradation of exterior surfacessuch as optical solar reflectors mayalso result from repeated ESDevents.12 In order to investigate

discharge phenomena, an electron

beam gun was used to chargeseveral spacecraft material samples,such as a silicon solar cell array, anoptical solar reflector (OSR), and aS13GLO thermal paint sample. Sincesample ESD rate can have a directrelationship to the outgassingrate,13 samples were mounted withadhesives used in satellite

manufacturing in order toaccurately simulate someoutgassing rates. Exposure ofcharged surfaces to the arcjetplume allowed investigation of thedischarge phenomena and possiblesurface degradation that resulted.

Although issues regardingEMI for an arcjet system in steadystate operation were previouslyaddressed ,9 there remainedunexamined areas of interest.These included conducted andradiated transients associated with

arcjet starting as well as radiatedemissions in special communicationbands during steady state operation.Selected EMI measurements were,therefore, performed whichfocused on these untreated issues.

For comparison purposes, anattempt was also made toincorporate some overlap withprevious investigations 9 byexamining low frequency (< 50MHz) radiated emissions using twodifferent PPU's. An array ofantennas was used to measure arcjetradiated emissions in frequencybands which included 50 kHz to 50

MHz, 160 to 500 MHz, and specialUHF, S, C, Ku, and Ka bands at adistance of 1 meter from the

thruster exit plane. Startupradiated transients were captured

using two antennas of the arraywhich were calibrated for acombined 20 to 500 MHz frequency

range. Conducted startup transientswere characterized using lead-acidbatteries instead of a commercial DC

power supply as the primary powersource for the arcjet system.

3

Hardware and Facility

Arc iet System and Interfaces

A flight-type arcjet systemconsisting of an arcjet thruster,power processing unit (also knownas power conditioning unit), andtriax interconnect cable was used

throughout this test program. Thearcjet was operated on N2:'2H2

mixtures simulating fullydecomposed hydrazine, whicheliminated the need for a gasgenerator. The thruster was a 1 to 2kW design developed at RRC undercontract to NASA. A cross-sectional

view and description of thisthruster were reported elsewhere.1 4Two different PPU's were used to

run the arcjet during testing. Thefirst unit, PPU A, developed under aprior NASA program by Watkins-Johnson and later modified by PED,was used in previous tests.9 Thisunit was supplied with 1.4 kW inputpower and produced a 1.26 kWregulated output to the arc jetthruster. PPU A was used in

contamination testing and forcomparative low frequency EMImeasurements. The second unit,

PPU B, developed by PED, was anengineering development modelused for space qualification testingunder a program between MartinMarietta and RRC. PPU B is identical

in circuit layout and packaging tothe Series 7000 PPU's. This unit

converts a maximum input power of1.8 kW to a nominal output power of1.63 kW at an arcjet operatingvoltage range of 90 to 140 Vdc.15PPU B was used in the EMI and ESD

portions of the test.Figure 1 shows an electrical

schematic of the arcjet system withsupporting interfaces. Under mostcircumstances, primary power tothe arcjet system was provided by acommercial, phase-control,regulated DC power supply. Duringconducted transient EMI

measurements, eight 12V, 200 Amp-

hr lead-acid batteries connected in

series were used as a power sourceto simulate satellite battery outputimpedance. At startups, power wasswitched to the PPU auxiliary, thenmain inputs by manual activationof electrical relays. Arc ignitionand stop command pulses (+10 V, 750Ixs) to the PPU were subsequentlysupplied by a command (CMD)interface pulse generator. Arccurrent and voltage telemetry(TLM) signals were monitored witha computer data acquisition system.Cabling between the PPU andprimary power, CMD, and TLMinterfaces consisted of unshielded,

twisted pair or bundle pair,approximately 8 m in length. Allnegative return leads weregrounded to the vacuum tank wall.The arcjet anode was also groundedto the tank wall due to the triax

cable configuration and PPU

chassis grounding.

Vacuum Facility

All experimental testing wasperformed in a 4.6 m diameter by19.5 m long, metal vacuum facility.The pumping system for thechamber included twenty 0.8 mdiameter diffusion pumps followedby blowers and roughing pumps.16With a propellant mass flow rate of42.9 rag/s, the facility backgroundpressure was 0.01 Pa (1 x 10-4 torr).

Several small and isolatable

test ports were available for accessto the main vacuum chamber. A

retractable rod assembly supportedsome of the spacecraft materialsamples and was inserted via a 0.9 mport. By retracting the assemblybehind an isolation gate valve, thesamples were removed and testedfor degradation without venting theentire chamber.

Diagnostic Apparatus and

Plum_ Contamina_iQn

4

TestThe arc jet and PPU were

mounted to a water-cooled plate inthe main portion of the vacuumchamber as shown in Figure 2.

Figure 3 shows the location ofstationkeeping arcjets on asimplified schematic of the Series7000 satellite built by MartinMarietta. Spacecraft samples thatwere used in the contamination

tests included solar cells, opticalsolar reflectors, thermal blankets,

and several paints which aredescribed in more detail in Table I.

These samples were positioned tosimulate two regions of the satellite.The first region was in thebackflow of the arcjet thruster

representing surfaces of the mainspacecraft body closest to thethruster exit plane. In this region,samples (1-6) were permanentlymounted to the PPU thermal

interface control plate inside themain vacuum chamber as shown in

the schematic of Figure 4 and the

photograph in Figure 5. The secondregion was the solar array panel.In this case, several samples weremounted on a retractable arm

assembly which allowed forrepeated measurements of sampleproperties with minimal testinterruption. These samples (7-11)were attached to a mounting plate

whose normal was pointed directlyat the arcjet exit, thus maximizingthe incident flow as shown in

Figure 4.

Test ProcedureDue to concerns about

contamination of samples bydiffusion pump oil, the sampleswere exposed for 40 hours to thevacuum environment with cold gas

flow from an unpowered arc jet.After the control exposure, sampleson the retractable arm assemblywere moved to an isolatable 0.9 m

diameter port. The gate valve wasclosed and the port vented to

atmosphere. Only the samples onthe retractable arm were removed

for visual inspection and surfaceproperty measurements. Thesamples were remounted andexposed to the arcjet plume for 40hours. After this exposure, all

samples were removed and theirsurface properties were tested.

Surface Property

M¢asurements

The degradation of materialperformance due to exposure to thearcjet plume was quantified by themeasurement of surface properties.Table I describes each of the

spacecraft materials and thesurface properties measured foreach sample.

A spectrophotometer with a60 mm diameter barium sulfate

coated integrating sphere was usedto measure the surface reflectance

spectral response. Total solarabsorptance over the wavelengthrange between 250 and 2500 nm wascalculated for an opaque sample byconvoluting the reflectance spectrawith the solar spectral irradianceaccording to Equation 1. Theaccuracy of this measurement wasestimated to be within + 2% and the

repeatability of the measurementswas + 0.005.17

a = 1 - p = 1 - {l p(_.) I(_,) dX}/{I l(k) d_,} ( 1 )

An infrared reflectometerwas used to measure the total room

temperature reflectance over thewavelength range between 5 and 25gm. This unit used dual rotatingcavities that compared radiationwith both a room temperaturesource and a heated blackbody.Total reflectance, the resultant

alternating energy component, wasthen converted to emittance using

Equation 2. The accuracy and theprecision of the measurement were

5

estimated as + 2% and + 0.005,

respectively.

e = 1-p (2)

Surface resistance was

measured with a high resistancemeter. Surface resistance wasmeasured across each sheet of the

material from one 7.6 cm long edgeto the opposite edge. In the case ofthe OSR sample (1), the resistancewas measured from the samplesurface to the mounting plate. Thepoor accuracy of the instrument for

resistances greater than lxl01 0ohms, due to electrical noise

coupling to the probe cables, was alimitation of this measurement.

Therefore, all resistances greaterthan lxl010 ohms were reported byorder of magnitude only and anaverage of three measurements wasused as the reported resistance.

Current-voltage curves ofthe solar cell array were obtainedusing a filtered xenon arc lamp thatsimulated the solar spectralirradiance in Earth orbit. A bipolarpower supply was used as a powersink by incrementally changingthe voltage drop. The cell responsewas characterized by a voltagemeasurement across the array andcurrent measurement obtained

based on the voltage across atraceable resistor in series with the

array. Variation in illuminationdue to fluctuations in the xenon arc

lamp was corrected by using astandard cell. The current of thestandard cell was measured

simultaneously with the testspecimen's current and voltage."The specimen's current was thencorrected by the ratio of thecalibration current to the measuredcurrent of the standard cell. The

precision of the voltagemeasurement was + 0.1%. Theillumination variation limited the

repeatability of the currentmeasurement to + 1%.

Electrostatic Discharge

Characterization

Concerns about the surface

property degradation caused by ESDled to the investigation of thedischarge phenomenon of chargedsamples during arcjet ignition.Spacecraft material samples ofsilicon solar cell array, optical solarreflector, and S13GLO paint(samples 11, 7, and 10) wereindividually mounted on aretractable rod assembly whichsimulated the relative position ofthe solar array. Each sample wascharged by a 20 keV electron beamgun with a 17 ° divergence beamthat was mounted in a 0.9 m side

port of the test chamber as shownin Figure 4. For these tests thesamples were aligned to face theelectron beam gun and weremounted to a Kapton coated plate ona movable assembly. A Hall effectcurrent probe measured anyinduced current in the ground wirethat connected the back of the

samples to the tank. The output ofthe current probe was monitored bya digital oscilloscope. The triggerlevel of the oscilloscope was set atO. 1 A to sense electrostatic

discharges, or sparks, whichtypically result in induced currentsof at least 1.0 A.18 Once the

oscilloscope was triggered, severalparameters, such as time, amplitude,and frequency were recorded andstored by computer. The chargingvoltage of the samples taken beforeand after each exposure, weremeasured by an electrostaticvoltmeter and high voltage probewhich was attached to another

retractable rod assembly as shownin Figure 6.

To determine the impact ofcharge time on sample charging

voltage, electron beam exposuretime for the solar cell array wasvaried from 3 to 18 minutes.

6

Exposure times of 3 to 5 minuteswere found to maximize the

charging voltage without causingexcessive sparking. Aftercharging, the samples were movedunder the high voltage probe formeasurement of the potential with

respect to tank ground. Sampleswere then moved out into the main

section of the chamber for exposure

to the arcjet plume. After exposurethe samples were again movedunder the high voltage probe for a

second potential measurement.The charged solar cell array

was exposed to the arcjet plume fordurations ranging from 1 to 120seconds, in order to determine the

time required to discharge samplepotential. The optical solarreflector and S13GLO paint samples

were repeatedly charged andexposed to the arcjet plume for 1second to determine any shot-to-shot variations in the observed

discharges. Concerns about chargedecay over the time necessary forsample movement and surfacepotential measurement, led to theexamination of several operating

conditions. Charged samples wereplaced in the vacuum chamber for15 minutes to determine the charge

decay associated with the movementof the samples on the retractablerod assembly and the exposure tobackground vacuum. Chargedsamples were then exposed to thecold gas plume of an unpoweredarcjet for 2.25 and 15 minutes todetermine if the discharges werecaused by neutral species. The 2.25

minute exposure time simulated theconditions of the 1 second powered

arcjet exposure, with the exceptionof thruster ignition, by accountingfor time needed to pressurize

propellant lines.Surface degradation of the

solar cell array via ESD's wascharacterized by changes in thecurrent-voltage characteristics.Changes in the OSR and S13GLOsamples were quantified by the

measurement of the surface

properties of absorptance andemittance.

Electromagnetic InterferenceCharacterization

Radiated emissions of the

arc jet were measured using anarray of antennas located a distanceof 1 meter from the arcjet exit planeas shown in Figures 5 and 7. Theantennas and their corresponding

frequency ranges were: active rod,or monopole, (10 kHz to 60 MHz),biconical (20 MHz to 300 MHz),

broadband dipole, or BBD, (160 to500 MHz), log periodic dipole, orLPD, (1 to 18 GHz), and horn (26 to40 GHz). Steady state emission

spectra were obtained byconnecting selected antennas to a50 kHz to 26.5 GHz superheterodyne-

type spectrum analyzer. Anexternal mixer allowed extension of

the analyzer upper frequency limitto 40 GHz when used with the horn

antenna. Arc ignition radiatedtransients were captured by

connecting the biconical or BBDantennas to the 50 ohm input of a500 MHz bandwidth digital storageoscilloscope. Antenna cables wereof the coaxial type RG58C/U for theactive rod, biconical, and BBDantennas. RG214/U was used withthe LPD and RG223/U with theexternal mixer of the horn

antenna. A short segment offlexible waveguide connected thehorn antenna to the external mixer.

Conducted voltage transients weremonitored with a matched pair of

passive, xl0 attenuation voltageprobes using a digital storageoscilloscope (DSO). These gave aneffective measurement bandwidthof DC to 200 MHz. Differentialmeasurements were obtained bywaveform subtraction.

7

Procedure

Table II shows the frequencybands for which steady stateradiated emission measurements

were taken. Spectrum analyzersweeps of each band were acquiredusing positive peak detection withthe (predetection) resolutionbandwidths indicated. These

measurements were performed forboth arcjet/PPU off and onconditions to allow discrimination

of background or ambient signalsfrom arcjet/PPU generatedemissions. Data were saved byhardcopy plots of the analyzerdisplay. These were later digitizedso that frequency dependentconversion factors could be appliedto obtain narrowband and

broadband field strengths at theantennas.

Prior to capturing thetransient radiated emissions

produced at arc ignition the arcjetwas cycled while reducing the DSOtrigger level until reliabletriggering was obtained on arcstart. A similar procedure was usedfor acquisition of conductedemission transients during thethree arcjet/PPU startup stages ofauxiliary power application, mainpower application, and arc start.These conducted transients were

measured with short time scale (2 to5 _ts), maximum bandwidth DSOsweeps. Longer time scale (20 ms)sweeps were also used todiscriminate relay closure andbounce spikes from the powersurge transients of interest.Voltage probe compensation wasperiodically checked with the DSOsquare wave calibration source toassure accurate transient

representation.

ResulI8 and Discussion

Plume Contamination

Surface properties of samples7-11, mounted in the simulated solar

array region, taken before andafter exposure to the cold gas froman unpowered arcjet are listed inTables III(a) and III(b). Thedifference between initial and final

properties was defined to be withinexperimental error when thedifference was less than twice the

uncertainty. This definitionaccounts for overlap ofmeasurement error bars betweenboth measurements. In the case of

absorptance and emittance, ameasureable change, occurredwhen the difference was greaterthan 0.01. With the exception of theabsorptance of Z93, sample 9, thechanges in absorptance andemittance of all samples werewithin the experimental accuracyof the measurements. The

measurable change in theabsorptance of Z93 might have beendue to a coating of backstreamingdiffusion pump oil. Minor variationin the surface resistance was

measured for both samples 8 and 9.These variations were small relative

to the uncertainty of themeasurement.

The solar cell array current-voltage curves taken before andafter exposure to the cold gas arcjetplume are shown in Figure 8(a).Changes in the solar cell array(sample 1 1) current-voltagecharacteristics are listed in Table

llI(b). Comparison of the initialand final values of short circuit

current and open circuit voltageshowed changes of less than 0.5%and 0.05%, respectively. Both werewithin experimental error. Otherparameters included in Table III(b)are defined as follows in Equations 3and 4:

FF = Pmax / (Isc Voc) (3)

11 = Pmax / Pinput = Vmax / (CsA) (4)

8

The changes in the derivedparameters of maximum power, fillfactor, and efficiency were all lessthan 0.6%, which was within

experimental error.The surface properties of the

samples, measured before and afterexposure to a powered arcjet plume,are listed in Tables IV(a) and IV(b).

Absorptance decreased in bothcases where the change was greaterthan experimental error. Forsample 1 it was not clear whetherthe difference was due todifferences in measurement

techniques or some effectattributable to the arc jet. The

change in absorptance for sample 9led to possible conclusions thateither the arc jet evaporated pumpoil contamination on the Z93 or the

post-control absorptance wasmeasured on a different portion of

the sample resulting in a falsedegradation. The change inemittance was greater than

experimental error in only onecase, where the emittance of thethermal blanket (sample 4)decreased significantly. Again, itwas not certain whether this

change was due to arcjet exposureor a discrepancy in measurement

techniques.The resistance of one Z93

paint sample (5) decreasedsignificantly during the exposureto the arcjet. However, decreasingsurface resistance would tend tolower differential charging

through an increase of chargeleakage to spacecraft ground. Theresistance increased for only one ofthe OSR samples (1) and one of theZ93 white paint samples (9). Anestimate of the end of life resistance

was calculated using a linearextrapolation based on the changein resistance after 40 hours. The

ratio of the test exposure time to the

total thruster operation time on thesatellite gave the extrapolationfactor of 16.25. The final

resistances after 650 hours of

exposure to the arcjet for the OSRand Z93 samples were calculated tohe 4.8x10 s ohms and 3.7x10 s ohms,

respectively. Both extrapolatedresistances were lower than themaximum of lx109 ohms, typically

specified for spacecraft.1 lTable IV(h) lists the

important characteristics of thesolar cell current-voltage curves

shown in Figure 8(b). The firstmeasurement of the current-

voltage curve measured a decreaseof 56 mV in the open circuit

voltage, but only a 4 mA increase inthe short circuit current. Anincrease in the cell operation

temperature is known to cause aslight increase in the cell current,at a rate of 0.03%/°C, and a

significant decrease in the voltage,at a rate of 2.2 to 2.3 mVFC. 19 An

increase in the operating

temperature of 3 to 6°C would haveaccounted for the differences

previously noted for a 4 by 4 solarcell array. The measurement wasrepeated the next day undercontrolled thermal conditions. Theresults are listed in Table IV(b) and

plotted in Figure 8(b). Both opencircuit voltage and short circuitcurrent repeated the originalmeasurements within experimentalerror. This information suggestedthat variation of the measurement

procedure caused the solar cell

array to be heated by the xenon arclamp which probably caused thenon-repeatable change in current-voltage characteristics. 2° Thechanges in the derived parametersof maximum power, fill factor, andefficiency were all less than 0.6%,based on the final current-voltagetrace, which was within

experimental error.

Electrostatic Discharge

Characterization

9

Sample charging times werevaried from 3 to 18 minutes in orderto determine the maximum

attainable charging voltage foreach sample. It was found that eachsample quickly reached a potentialbeyond its breakdown thresholdand an arc would form causing adecrease in the charging voltage.As a result of spark discharges, theoscilloscope was always triggeredand sample potentials neverreached ground potential. In onecase the solar cell array potentialincreased from -8900 to -5800 volts

when the sample sparked to thehigh voltage probe as it was beingretracted. After the initial

discharge event, a cyclic pattern ofdischarges occurred on a fairlyregular interval as long as theelectron beam charging processcontinued. Maximum potentialdifference or charging voltage wasobtained, typically after 3 to 5minutes, by turning off theelectron beam gun before the nextanticipated spark. The maximumcharging voltage varied for eachsample and depended on surfaceresistance. The maximum charging

•voltages for the solar cell array,OSR, and S13GLO samples were -9700,-11,600, and -300 volts, respectively.

Examinations of severalcontrol conditions were done with

the solar cell array due to concernsabout charge decay caused bymovement of the samples with theretractable rod assembly, the timedelay in the voltage measurement,and exposure of samples to the coldgas arcjet. The solar cells werecharged to -5100 volts and insertedinto the test chamber. After a 15

minute exposure to vacuum, thesample potential was -4600 volts.Since the oscilloscope was nottriggered during this period, somecharge dissipation mechanism,perhaps a low-current Townsenddischarge or charge leakage to thesupport, caused some decay in thenegative potential. Charge decay

due to the exposure of the solar cellarray to the cold gas arcjet for 2.25minutes was also investigated. Itwas found that the potentialchanged from an initial -7900 voltsto -7000 volts. Like the vacuum

exposure, a reduction in thepotential occurred, but the samplewas still negatively biased to tankground. Finally, all three sampleswere exposed for 2.25 minutes to acold gas arcjet with the PPUauxiliary power on. All samplesexperienced changes in voltagepotential without sparks during theexposure, but none reached groundpotential during the exposure.

The solar cell array wasexposed to the powered arcjet forexposure times of 1, 6, 60, and 120seconds. During every exposure the

initial negative potential, whichranged from -8100 to -7000 volts,was raised to ground potential (0 + 2volts) without triggering theoscilloscope. It was believed thatsparking was not the dischargemechanism, because the

oscilloscope was not triggered inthe process and charging voltagewas raised to zero. The OSR and

S13GLO samples were charged 5 and2 times, respectively, and exposedfor 1 second to the arcjet withouttriggering the oscilloscope. Thecharging voltages ranged from11,600 to -5800 volts for the opticalsolar reflector, while the pre-exposure potentials were -300 and -200 volts for the S13GLO paintsample. The potentials of allcharged samples were benignly andconsistently raised to groundpotential without triggering theoscilloscope during arcjet exposure.

The changes of the opticalproperties of absorptance andemittance for the OSR and paintsamples resulting from the ESD testsare listed in Table V(a). All changesin the surface properties werewithin experimental error of theinstrumentation. This result was

not completely unexpected based on

10

the limited number of charging

cycles for each sample. However, itshould be stated that any optical

degradation due to ESD damagewould not be induced by the

ignition of the arcjet thruster.Current-voltage characteristics forthe solar array measured beforeand after the electrostatic discharge

testing showed little variation.Table V(b) lists the before and after

exposure current-voltagecharacteristics for the solar array.Traces are shown in Figure 8(c).

All measured final properties

repeated original measurementswithin experimental error.

Electromagnetic InterferenceCharacterization

Nearly the complete set ofEMI data is included within Figures9-19. The discussion below

concentrates on the prominentfeatures of the data and, where

possible, their comparison withother work.9

Steady state radiated emissionsignals and/or thresholds recordedwith the spectrum analyzer werecorrected to give correspondingnarrowband and broadband field

strengths at the antennas. Detailsof this data reduction are discussed

in the Appendix.Figures 9(a) shows the

results for narrowband emissions at

ambient conditions, while Figures9(b) and 9(c) show narrowbandemissions during arcjet operation

using PPU's A and B, respectively.Discontinuities which appear in the

various plots were due to analyzerbandwidth and attenuation changesas outlined in Table II. Comparison

of the two PPU cases against theambient plot demonstrated theradiated emissions attributable to

operation of the arcjet system. Thenarrowband spikes apparent from50 kHz to 1 MHz in Figures 9(b) and9(c) were separated by the

corresponding PPU switchingfrequencies, about 16 kHz for PPU Aand about 20 kHz for PPU B. These

"switching harmonics" continuedout to around 20 MHz but were

unresolved by the larger 30 kHzresolution bandwidth used above 1MHz. In both PPU cases the

narrowband spikes associated withPPU switching exceeded the MIL-STD-461C limit by up to 20 dB forfrequencies below 1 to 2 MHz.When compared to previousresults,9 in which the same arcjetand PPU were used, the levels of

this test appeared to be about 20 to40 dB lower. However, as discussedin the broadband analysis below,

the signal levels for bothnarrowband and broadbandemissions measured in this test withthe active rod antenna (50 kHz to 50MHz) were likely beingunderdisplayed, or compressed, dueto saturation of the antenna

amplifier by a high level of arcjetassociated low frequency broadbandnoise. This also accounted for

suppression of ambient signalsbetween 1 and 50 MHz in both PPUcases, which is apparent when

comparison is made to Figure 9(a).As a result of the active rod

saturation, quantitative emissionlevel comparisons between arcjetoperation for PPU's A and B werenot considered appropriate.

Nevertheless, in qualitative terms,it may be said that the spectralsignatures were similar and in bothcases exceeded the MIL-STD-461C

limit by at least 5 to 20 dB forfrequencies less than 2 MHz.

Two other points concerningthe narrowband emission plotsshould be noted. First, the reducedlevel in the 10 to 50 MHz range of

the PPU B plot compared to theambient or PPU A plots was a resultof a 10 dB difference in analyzer

input attenuation and, therefore,noise measurement threshold.

Secondly, the elevated narrowband

spike at 455.6 MHz in Figures 9(b)

11

and 9(c) relative to the ambientlevel in Figure 9(a) should not beattributed to the arcjet or the PPU,as the magnitudes of such ambientemissions were often found to varyfrom measurement to measurement.

Figures 10(a) 10(c) showthe broadband emission results forconditions similar to those of the

narrowband plots in Figures 9(a)9(c). As indicated in Table II,however, larger resolutionbandwidths were used in someinstances to achieve better

sensitivity, or lower threshold, tocoherent broadband emissions. The

most striking feature evident whencomparing Figures 10(a)- 10(c) isthe increased level of low

frequency broadband noise duringarcjet operation. This emissionexceeded the MIL-STD-46 IC

specification for frequencies up to300 kHz. Howeve, r, it also exceededthe active rod antenna saturation

limit of +105 dB_V/m/MHz.Although it appeared that thissaturation occurred only forfrequencies less than or equal to150 kHz, the actual effect was a

compression of signal level acrossthe entire 50 kHz to 50 MHz activerod band, as mentioned earlier.

Avoidance of this problem wouldhave required use of an active rodwith pre-gain stage attenuation or apassive rod antenna. Due to timeconstraints, neither was available

during testing.A slight difference in noise

threshold between Figures 10(a)and 10(c) for the 10 to 50 MHz rangewas due to a 10 dB change inattenuation and use of a smaller

resolution bandwidth in the PPU Bcase. Likewise, the lower thresholdfor 10 to 50 MHz in Figure 10(b)compared to 10(a) or 10(c) was theresult of switchout of analyzerattenuation. Emission peaks at 11.6MHz and 15.7 MHz in Figure 10(c)were examined closely and found tobe 95 kHz harmonics and 1.3 Hz

impulses, respectively. These were

believed to have been intermittent

ambient signals. Though notpresent in Figure 10(a), the peakbetween 212 to 215 MHz in Figure10(b) was also found to be atemporary ambient signal.

Overall, comparison of thebroadband results in Figures 10(b)and 10(c) with previous work,9revealed a general similarity inspectral profile. In both tests, ahigh level of broadband noiseexceeding MIL-STD-461C wasapparent for low frequencies. Thisnoise rolled off by 20 to 30 dB perdecade with increasing frequency.Since the communication bands ofinterest are well above this

frequency range, no radiated EMIproblems are foreseen for steadystate arcjet operation on acommercial communicationssatellite.

To verify that arcjet radiatedemissions were not a problem atcritical satellite communication

frequencies, a set of sweeps wasconducted which focused onselected communications bands.

The results for UHF, S, C, Ku, and Kabands are displayed in Figures 11-15, respectively. No ambient orarcjet related emissions were foundfor any of these frequency rangesto the sensitivity levels indicated inthe figures. Improved sensitivity,that is, lower noise measurementthresholds were desired, but

required impractically long sweeptimes (using smaller resolutionbandwidths) for the narrowbandcase and/or use of more sensitive

measurement equipment in thebroadband case. Nevertheless, it

was clear from Figures 11-13 thatarcjet narrowband emission levelsin the UHF, S, and C bands werebelow the MIL-STD-461C

narrowband specifications. Sincethis narrowband limit terminates at10 GHz, it does not appear in the Kband plots of Figures 14 and 15.Arcjet broadband emissions in the

UHF band investigated may also be

12

concluded to be within the MIL-

STD-461C broadband limits as Figure11 shows. Because this specificationextends only to 1 GHz, it is notshown in Figures 12-15. It is also

noteworthy that measurementsensitivity tends to become poorerat higher frequencies as the "step"feature around 6.2 GHz in Figure 13

highlights. This discontinuity insensitivity level reflects a changein the spectrum analyzer localoscillator harmonic and

corresponding increase in analyzerinternal noise.

Arc ignition radiatedtransients were observed using thebiconical and broadband dipoleantennas. Figures 16 and 17 showthe resulting triggered time domainsignals received from theseantennas during arc jet ignition.Although not unfolded fromantenna factors, these waveforms

give some indication of how longtransient high frequency

components may take to reachsteady state levels. From Figures16(a) and 17(a), it would appear thisoccurred within 1 to 2 its for the 20to 500 MHz antenna band. However,effects of test chamberreverberation and antenna cable

internal reflections or ringingwere difficult to distinguish whenanalyzing such transients. Thesefactors would, however, prolong thedecay of the observed transientrelative to that which actually

would occur in an openenvironment as in space. Figures16(b) and 17(b), which areexpanded segments of 16(a) and17(a), show risetimes of 15 to 40 nsand dominant oscillations of 50 to

200 MHz. To obtain a better pictureof the spectral content and radiatedemission field levels represented bythe time domain pulses, the trace ofFigure 16(b) was numericallyFourier analyzed. Antenna factorswere then applied to yield Figure 18which shows the transientbroadband radiated emission levels

over the 20 to 300 MHz biconical

antenna band. The large dips in the

spectrum are an artifact of thetruncation of the pulse in Figure16(b) at approximately 160 ns. Forreference, an error in amplitude of

two percent makes the digitizationnoise floor in Figure 18 about 34 dBbelow peak, or about 32dBI, tV/m/MHz. At 60 MHz thespectrum shows a peak emission of66 dBlaV/m/MHz which was stillwithin the steady state MIL-STD-461C broadband limit. The high

frequency rolloff in emission levelof approximately 7 to 10 dB/decadeappears to maintain compliancewith the steady state MIL-STD-461Climit for frequencies above 60 MHz.However, the transient emission

levels appear significant(measurable) to at least 300 MHz and

thereby warrant more extensiveinvestigation.

Voltage transients observedon the primary power, command"ON", and arc current telemetrylines at the stages of PPU/arcjetstartup are outlined in Table VI.Included for comparison are thetransient levels or amplitudes fromPPU B switching, captured duringsteady state arcjet operation. Figure19 shows an example of a transienton the primary power lines at themoment of battery power

application to the main power inputof PPU B. The MIL-STD-461C CE07

specification calls for such depower line transients to not exceed+50% or -150% of the nominal line

voltage. For the +96 V primarypower voltage here, thiscorresponds to upper and lowerlimits of +144 V and -48 V. As can be

seen for the example of Figure 19and from the values of Table VI, the

auxiliary on, main on and arc startevents result in power linetransients which are within the

appropriate limits except for aminor +7 V violation of the upperlimit in the main on case. The MIL-

STD-461C CEO7 specification does not

13

apply to the command andtelemetry signal lines which do nothave fixed line voltages, but 0 to 5 Vranges. Transient peak to peakamplitudes were neverthelessrecorded and durations found to be

less than 2 _ts. Effects of CMD/TLMinterface impedance characteristicson the CMD/TLM lines were not

investigated here. In priorintegration testing,9 no changes intelemetry data were observedduring arcjet ignition.

Conclusions

In order to address residual

user integration concerns in theapplication of hydrazine arcjets oncommercial communications

satellites, an integration test wasperformed under a Space Actagreement between Martin MariettaAstro Space, the Rocket ResearchCompany, and NASA Lewis ResearchCenter. In this test, plumecontamination on spacecraftmaterial samples, electrostaticdischarge phenomena, andelectromagnetic compatibility wereinvestigated.

Potential contamination of

spacecraft surfaces wasinvestigated by positioningspacecraft material samples relativeto the arcjet thruster in order tosimulate both the satellite body andsolar array regions of a typicalcommunications satellite.

Contamination was quantified bythe measurement of surface

properties both before and after theexposure. The samples in thesimulated solar array region wereexposed to the cold gas arc jet plumefor 40 hours to address concerns

about contamination bybackstreaming diffusion pump oil.With the exception of one sample,no significant changes were

measured in absorptance andcmittance within experimentalerror. Surface property

measurements taken before and

after the exposure to a poweredarc jet plume revealed severalthings. Absorptancc decreased intwo cases where only minor

changes were measurable. Thedecrease in emittance of a thermal

blanket sample was the onlymeasurable degradation of thisexperiment. Measurable changesin resistance yielded acceptable endof life characteristics. Thecontamination of a silicon solar cell

array was quantified by themeasurement of the current-voltage characteristics both beforeand after exposure to the cold gasarcjet and powered arcjet plume.No measurable change in thecurrent-voltage characteristicsoccurred with the exception of anon-repeatable shift in onemeasurement believed to be a

temperature effect and not acontamination issue.

Concerns about the surface

property degradation caused byelectrostatic discharges led to theinvestigation of the dischargephenomenon of charged samplesduring arcjet ignition. Shortduration exposures of chargedsamples demonstrated that thepotential differences wereconsistently and completelyeliminated within the first second

of exposure to the weakly ionizedplume. The spark dischargemechanism was not the dischargephenomenon since the charging

voltage was completely dissipatedand the discharge process did nottrigger the oscilloscope with asignal from the current probe,which measured the induced

current in the sample ground strap.In contrast, spark discharges werefound to trigger the current probe,but not completely dissipate

charging voltages. Exposure tocontrol conditions did not cause a

significant dissipation in chargingvoltage. These results suggest that

14

the arcjet could act as a chargecontrol device on spacecraft.

Steady state radiatednarrowband and broadbandemissions were measured for

various frequency ranges between50 kHz to 40 GHz. Comparison ofresults for arcjet operation on twodifferent power processing unitsshowed similar spectralcharacteristics for bothnarrowband and broadbandemissions. Broadband emissionsexceed the MIL-STD-461C below 0.3

MHz while previous work hasshown that the upper frequency ofexcessive emissions extended to 40MHz. The difference between

results may be explained bysaturation of the active monopoleantenna. Sweeps of special UHF, S,C, Ku, and Ka bands showed nonarrowband or broadbandemissions above the measurement

thresholds, which were below theMIL-STD-461C standards.

Acknowledgements

This project has beencosponsored by Martin MariettaAstro Space Independent Researchand Development and NASA LeRC.The authors would like to thank

Rocket Research Company andPacific Electro Dynamics forproviding the PED engineering

development model PPU and theinvaluable expertise of Peter Skelly.The authors would also like to thankJames Armenti for his technical

assistance, particularly in theassembly and operation of theelectrostatic discharge test

equipment. The authors would liketo recognize the insightful adviceof Charles Bowman in association

with these activities. Finally, theauthors would like to acknowledge

the dedicated support provided bythe Test Installations Division of the

Electric Power Laboratory.

References

1. Curran, F. M. and Haag, T. W., "AnExtended Life and Performance Testof a Low Power Arcjet," AIAA 88-

3106, July 1988, (NASA TM 100942).

2. Gruber, R. P., "Power Electronics

for a 1-Kilowatt Arcjet Thruster,"AIAA 86-1507, June 1986, (NASA TM

87340).

3. Knowles, S. K., "Arcjet ThrusterResearch and Technology, PhaseII," NASA CR-182276, RocketResearch Co., Redmond, WA, March1991.

4. Carney, L. M. and Keith, T. G.,"Langmuir Probe Measurements ofan Arcjet Exhaust," Journal of

Propulsion and Power, Vol. 5, No. 3,May-June 1989, pp 287-294.

5. Carney, L. M. and Sankovic, J. M.,"The Effects of Arcjet OperatingCondition and Constrictor Geometryon the Plasma Plume," AIAA 89-

2723, July 1989, (NASA TM 102284).

6. Sankovic, J. M., "Investigation ofthe Arcjet Plume Near Field UsingElectrostatic Probes," NASA TM103638, October 1990.

7. Carney, L. M., "Evaluation of theCommunication Impact of a Low

Power Arcjet Thruster," AIAA 88-3105, July 1988, (NASA TM 100926).

8. Ling, H. et al., "Effect of ArcjetPlume on Satellite Reflector

Performance," IEEE Transactions onAntennas and Propogation, Vol. 39,No. 9, September 1991, pp 1412-1420.

9. Zafran, S., "Hydrazine Arcjet

Propulsion Integration Testing,"IEPC 91-013, Proceedings of the22nd International Electric

Propulsion Conference, October1991.

15

10. Bogorad, A., et al.,"The Effects ofArcjet Plasma Propulsion Systemson Spacecraft Charging andSpacecraft Thermal ControlMaterials," IEEE Transactions on

Nuclear Science, Vol. 39, No. 6, pp1783-1789, December 1992.

11. Purvis, C.K., et al., "DesignGuidelines for Assessing andControlling Spacecraft ChargingEffects," NASA TP 2361, 1984.

12. Bogorad, A., et al., "ElectrostaticDischarge Induced Thermo-OpticalDegradation of Optical SolarReflectors (OSRs)," I E E ETransactions on Nuclear Science,Vol. 38, No. 6, pp 1608-1613,December 1991.

13. Bogorad, A., et al., "RelationBetween Electrostatic DischargeRate and Outgassing Rate," IEEETransactions on Nuclear Science,Vol. 36, No. 6, pp 2021-2026,December 1989.

14. Morren, W. E. and Lichon, P.J.,"Low-Power Arcjet Test FacilityImpacts," AIAA 92-3532, July 1992.

15. Skelly, P. T., Fisher, J. R., andGolden, C. M., "Power ConditioningUnit for Low-Power Arcjet FlightApplication," Pacific ElectroDynamics, Redmond, WA, AIAA 92-3529, July 1992.

16. Sovey, J., et al., "Test Facilitiesfor High Power ElectricPropulsion," AIAA 91-3499,September 1991, (NASA TM 105247).

I7. Dever, J. A., et al., "Evaluation ofThermal Control Coatings for Use onSolar Dynamic Radiators in LowEarth Orbit," AIAA 91-1327, June1991, (NASA TM 104335).

18. Bogorad, A., et al., "AmplitudeScaling of Solar Array Discharges,"IEEE Transactions on Nuclear

Science, Vol. 37, No. 6, pp 2112-2119,December 1990.

19. Agrawal, B.N., Design ofGeosynchronous Spacecraft,Prentice-Hall Inc., Englewood Cliffs,NJ, 1986.

20. Brinker, D.J.: Privatecommunication.

21. Engelson, M. and Telewski, F.,Spectrum Analyzer Theory andApplications, Artech House,Dedham, MA, 1979.

Appendix

Data displayed by thespectrum analyzer reflected asignal (noise) level, whetherinternal to the analyzer or externalfrom an antenna, which wasreferenced to the spectrumanalyzer input. This level wasexpressed in logarithmic terms asthe equivalent rms voltage to apower dissipated in its 50 ohm inputimpedance. That is, signal wasshown in decibels relative to lgV, ordBgV where

• (dBgV) = 20 log[_(gV)/l_tV ](5)

Spectra were recorded by plottingout the spectrum analyzer display.These plots were later digitized sothat frequency dependentcorrections such as antenna factor

and cable loss could be applied toyield corresponding electric fieldstrengths at the antennas. For thenarrowband type analysis thefollowing equation was used on alldata:

NB(dBgV/m) = ¢_(dBI.tV) +AF(dB/m) + CL(dB) (6)

Conversion loss of the externalmixer for the horn antenna is not

included in Equation 6 because it

16

was accounted for by the spectrumanalyzer. The antenna factor andcable loss factor were obtained frommanufacturer supplied ANSI C63.5antenna calibrations and calculated

cable attenuations, respectively.Broadband type noise

analysis was accomplished by firstselectively deleting clearlyidentified narrowband type signalsfrom the tabulated data generated

with Equation 6. Because broadbandnoise levels may depend on thereceiver bandwidth used, anadditional bandwidth normalization

factor was then applied to yieldbroadband emission levels as:

BB(dB_tV/m/MHz) =NB(dBlaV/m) + BNF(dB) (7)

This corrected noise levels from

those observed with the spectrumanalyzer 3 dB Gaussian resolutionbandwidth to those observed with a

standard 1 MHz impulse bandwidth.The correction assumed (worst case)

coherent type noise i.e. noise forwhich a xl0 change in bandwidthresults in a 20 dB change in noise

level. Since the equivalent impulsebandwidth of a Gaussian shaped

filter is approximately 1.5 times theGaussian 3 dB bandwidth21 the

normalization factor was given by:

BNF(dB) = 20 log{(1MHz)/[1.5 RB(MHz)]} (8)

Table I - Spacecraft materials description and placement.

SampleNumber

Description

Four Indium-tin oxide (ITO)coated CMX OSR's on an

aluminum plateS13GLO white paint on

aluminum

SimulatedLocation on

SpacecraftBackflow

Backflow

Size

(cm x cm)

7.6 x 7.6

7.6 x 12.7

3 MH21SLO black paint on Backflow 7.6 x 12.7aluminum

4 0.13 mm Dupont Kapton®, Backflow 15.2 x 15.2second-surface-aluminized

thermal blanket material

5 Z93 white paint on aluminum Backflow 7.6 x 12.7

6 Z306 black paint on aluminum Backflow 7.6 x 12.7

7 Solar Array 7.6 x 12.7

Solar Array

Six fused silica Optical SolarReflectors on a 1.3 cm

honeycomb panelCarbon-loaded Kevlar® 7.6 x 12.7

9 Z93 white paint on aluminum Solar Array 7.6 x 12.7

1 0 S13GLO white paint on Solar Array 7.6 x 12.7aluminum

1 1 Solar Array 10.2 x 12.7A four element by four elementsilicon solar cell array circuit

on a Dupont Kevlar® skin

honeycomb panel with fusedsilica coverglasses

PropertiesEvaluated

a,E,R

cx,c,R

ct,_,R

Gt,£

ct,c,R

Gt,£

IX,E

a,c,R

Current-

voltage trace

17

Table II - Antenna types and spectrum analyzer bandwidths for the frequency ranges

investigated during steady state conditions.

Antenna Band

Active Broad

Rod Range

MonopoleActive Broad

Rod Range

MonopoleActive Broad

Rod Range

MonopoleActive Broad

Rod Range

MonopoleBBD Broad

Range

BBD UHF

LPD S

LPD C

LPD K u

Horn Ka

Resolution Bandwidths* (Hz)Narrowband Broadband

Frequency Ambient PPU A PPU B Ambient PPU A PPU B

Range (I-Iz)50-250k 100 100 100 100 100 100

250k-lM 300 300 300 300 300 300

1-10M 30k 30k 30k 30k 30k 30k

10-50M 10k** 10k** 10k 300k** 300k 10k

160-500M 10k 10k 10k 300k 300k 300k

240-255M 10k 10k 300k - 300k

2.6-2.7G 10k 10k 300k - 300k

5.9-6.4G 10k 10k 300k - 300k

14.0-14.5G 10k 10k 300k - 300k

27.5-30G 10k 10k 300k - 300k

* Video (post-detection) bandwidths _> resolution bandwidths for peak detection.

** Analyzer input attenuation = 10 dB.

Table III- Initial and final property measurements for 40 hour exposure to controlconditions.

Table III(a) - Initial and final surface properties of samples exposed for 40 hours to the

cold gas from an unpowered arcjet.

Sample Initial Final Initial Final Initial Final

Number Absorptance Absorptance Emittance Emittance Resistance Resistance

(+ 0.005) (+ 0.005) (+ 0.005) (:1: 0.005) (ohms) (ohms)

7 0.038 0.042 0.804 0.802

8 0.939 0.940 0.894 0.892 0.9x107 1.2x107

9 0.127 0.138 0.918 0.915 2.8x107 2.4x107

10 0.198 0.201 0.901 0.899

18

Table III- Initial and final property measurements for 40 hour exposure to controlconditions.

Table IIl(b) - Solar cell array property changes over exposure to 40 hour cold gas from an

unpowered arcjet.

Characteristic

Short Circuit Current

(+ 1%)

Units Initial PropertymA 1176

Final Property1186

Open Circuit Voltage mV 2217 2216(± 0.1%)

Maximum Power mW 2000 2010

Fill Factor % 76.8 76.4

Efficiency % 11.4 11.4

Table IV - Initial and final property measurements for 40 hour exposure to powered arcjetconditions.

Table IV(a) - Initial and final surface properties of samples exposed for 40 hours to an

arc jet

Sample Initial Final InitialNumber Absorptance Absorptance Emittance

(± 0.005) (± 0.005) (± 0.005)

I 0.107" 0.090** 0.810"

2 0.198" O.I92"* 0.900"

3 0.970* 0.974** 0.900*

4 0.339* 0.332** 0.540*

5 0.127" 0.128"* 0.920*

6 0.960* 0.959** 0.915"

7 0.042 0.042 0.802

8 0.940 0.940 0.897

9 0.138 0.127 0.915

10 0.201 0.200 0.899

Final Initial Final

Emittance Resistance Resistance

(± 0.005) (ohms) (ohms)

0.800** 0.5x106t 3.0x1077

0.905** 101° 101°

0.908** 1011 1011

0.518"*

3.1x1070.917"* 2.0xlO s

0.913"*

0.798

0.895 1.2xlO 7 1.1xlO 7

0.915 2.4x10 v 4.7x10 _

0.897 10 l° 10 l°

* Typical optical properties of samples.** Measured after both control and powered arcjet conditions.

? Measured from sample surface to mounting bracket.

Table IV(b) - Solar cell array property changes over 40 hour exposure to arcjet.

Characteristic

Short Circuit Current

(+_ 1%)

Open Circuit Voltage

(+ o.1%)Maximum Power

Fill Factor

Efficiency

* No thermal control

** With thermal control

Units Initial Property1186

Final Property*1190mA

mV 2216 2160 2216

mW 2010 1930 2010

% 76.4

11.4%

75.2

11.0

Final Prooerty**

1180

76.8

11.4

19

Table V - Initial and final property measurements for electrostatic dischargecharacterization tests.

Table V(a) Initial and final surface properties of electrostatic discharge

characterization test samples.

Sample

Type

OSR

S13GLO

Initial Final Initial Final

Absorptance Absorptance Emittance Emittance

(± 0.005) (± 0.005) (± 0.005) (+ 0.005)

0.042 0.042 0.798 0.797

0.200 0.203 0.897 0.897

Table V(b) - Solar cell array property changes for electrostatic discharge characterization

testing.

Characteristic

Short Circuit Current

(:1:1%)Open Circuit Voltage

(_+0.1%)

Maximum Power

Fill Factor

Efficiency

Units

mAInitial Property

1180Final Property

1182

mV 2216 2218

mW 2010 2000

% 76.8

11.4%

76.4

11.4

Table VI - Conducted voltage transients.

Primary power CMD "ON" lineEvent

AuxiliaryOn

Main On

Arc start

PPU

switchingtt

Level (V)Max Min

+123 +78

+15.1 -14

+98.7 +93.7

+98.5 +93.5

Duration*

(_s)

Amplitude**

(v)

Duration*

(ITS)

Arc current telemetryline

Amplitude**

(v)Duration*

(ITS)

<5 2.3 <2 2 <2

<5 8 <2 14 <1

<2 3.5t <2

<3 5.6 <3 6.3 <3

* Spike duration measured from 20% leading edge to 20% training edge amplitude points.** Peak to peak amplitudes.

t First power processing unit switching spike.

tt Measured using DC power supply for primary power.

20

_rc_et

R_d

V I OI'F ON

i Tank

Wall

+ I Relay,,,,u,y t- Box

/

\

+1 Batzueriesi_ - I _:po,,_r

, Oscilloscope /

I Da_ IAcqui- I

s_e_ I

Figure 1 - Arcjet system electrical schematic.

Figure 2 - Arcjet and PPU mounted to cold plate inside vacuum chamber.

21

Figure 3 - Diagram showing location of arcjets on Series 7000 communications satellite.

High voltage

Electronhelm gun

Gate valve

Cold plate

9 D

4.6 m

Corn

Figure 4(a) - Schematic top view. Figure 4(b) - Schematic cross section.

Figure 4 - Spacecraft material contamination test schematic.

22

i!

Figure 5 - Test setup showing spacecraft materials and antenna array.

Figure 6 - High voltage probe measuring surface potential of an optical solar reflector.

23

Co_ _,t

f m

/IIII

\Cold,

mt_

\

B_

f'_'_/ R - lm

BICO_CIII

Figure 7(a) - Schematic tank cross section. Figure ?(b) - Schematic tank side view.

Figure ? - Schematic of antenna array for electromagnetic interference tests.

1.5

m

1 m

0.5

I I I I

0 0.5 1 1.5 2 2.5

Pre-exposure

Post-exposure

Voltage, V

Figure 8(a) - Pre and post cold gas arcjet exposure.

Figure 8 - Current-voltage curves for 4 x 4 silicon solar cell array.

24

t_t_

1.5

1 m

0.5

0

I

I I I I

0 0.5 I 1.5 2 2.5

Pre-exposure

Post-exposure(no thermal control)

Post-exposure(with thermal control)

Voltage, V

1.5

Figure 8(b) - Pre and post powered arcjet exposure.

.xe__t

a..

_ w

I

0.5-

0 l

0 0.5I I I

1 1.5 2 2.5

Pre-exposure

Post-exposure

Voltage, V

Figure 8(c) - Pre and post electrostatic discharge test exposure.

Figure 8 - Current-voltage curves for 4 x 4 silicon solar cell array.

25

E

c

z.z.J

!00

_o IIIIIII,o IllUll_o fltlill __o iilllll;,,

l_o )tlllll40--i-.i_,_

IIIIIII-_o lllllili

IIIILII,__,_

10-2 .10-1

IIIIIIIII IIIItlUl illllllIIIIIIIIIIIIIIIII l illlltlIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIit111l ill_

111111l i IllilllJ.'t'gllllUIUL.J_ .l,_llil7"._ ,lllZ.,.B_IIIIII IIIIIII

100 101 102 103Frequency,

E

>=t_

.

125

115

105

95

85

75

65

55

45

35

10"2

, ......iSillill......I IIIItU

: II IIIII1• : :::::

-_',,'f,(,I IIIll

_L......k_ "....... IIrrllll_

IIIIIIII: :::::: , i ,.,,. :'-::::: : :

- 100 10110I

Frequency,k(Hz

; : :::::

: : :'.::: : :

: :::::: : ;

IUIIII "

• " IIII1111ii Illlil

I _1IIIIII) il IIIIII

!!!!!!!102 103

Figure 9(a) - Ambient conditions (ArcjerJPPU off).

E

c

.o

_0

80

70 tl

40 -_

20

tO

0 . •

IlilllIIIIIIII IIIIIIIIIllillllIIit1111IIIIIIIIIIIIlliIIIIIIIIIIIIIIIIJ lllllllIIIlilI lllllll_ IIIIII IIIIIIIIIIIIIIIIIIIIIIi llllllli ):lilt,, IIIIIIIIIIIIIIII

__ 1t__IUtlI.,ItlIIIIIll_iillllii__.,_,IH, , , IlH. " ",,,,! ................

10"1 100 101 102 103Frequency,MHz

Figure 10(a) - Ambient conditions (Arcjet/PPU off).

E

w

L_5

125

115

105

95

85

75

65

55

45

35

10-2

IIIIIII llillli( iiiiiii

i :::::::i........ ; : :::::: I

....... I

Jllii['_ii............ '" I lifl! I: : : : :::

,- _11 ....... IIIIIIII N.IIII[""_ "IIIIIIII II_.,,. IIIIIII-j"IIIIIII IIIIII1_,.1111111iIIIIIIII IIIIII "tI,IUJLi

: )!!!!!!!!!!!!!!).n.!._.;10"1 100 101

Frequency,MHz

II IIIIIII_1IIIIIIII IIIIIIIIII IIIIIIIi IIIII111I ltlL_

I1 LP_tlIIII

102 103

Figure 9(b) - ArcjerJPPU A emissions. Figure ]O(b) - Arcjer/PPU A emissions.

lO0

9O

8o

70

60

5o

- 40

30

2O

10

0

I "llllil I IIIIIII,!!IIIIIIII Illlll

i !i_H_,iilit,IIIIIIII IIIIU

_H JCi_I,d-HINUIIIII,,Ilmi,IJJlll

IIIlli_ :::..... '"":" _

•-- i iiii,, IIIIii,iiiJi '_

......... 95ii till ......... ,,.,_

1 "• - - IHI ................

IIIIII.......: : :::::

Jtln_,i i iii

IIIlll _1111illlll I1_Illlll. IIIIIliIIIIIIIIIIIIII!!!!!!)!!UU!

10-2 I0-1 100 101 102 103Frequency,k(..Iz

I)-2

iiiiiiii )llllI llJllll........ ,11.11

iiiiiiii IttHtll lllttili iiiii II IIIIII

It11!1II IIIIIli

!!_I Ii!!! !)!!!!!!!10"1 100 101 102 103

Frequency,I,,#Hz

Figure 9(c) - Arcje_PPU B emissions.

Figure 9 - Narrowband radiated emissions.

Figure 10(c) - Arcjer/PPU B emissions.

Figure 10 - Broadband radiated emissions.

26

go

8O

7O

5O

2O

tO

024O 245 2S0 2_

Freq., I_

Figure 11(a) - Ambient narrowband condidons.100

gO

8O

7O

5O

._ 4O

245 25O 255

Frequency,

i 3O2O

10

0240

Figure 1 l(b) -

9O

m_mJ:rt

10

024O 245 25O 2_

Frequency,li_

Figure 11(c) - Ambient broadband conditions.100

9O

8O

SO

2O

I0

0

Arcjet/PPU B narrowband emissions. Figure 1 l(d) -Arcjer./PPU B broadband emissions.

Figure 11 - Narrowband and broadband radiated emissions in the UHF band.

E

gO

8O

4O

JO

2O

10

0

u. _I_14 :lqS

2.60 2.65 2.70

Frequer_,

Figure 12(a) - Ambient narmwband conditions.

._ 4o

2O

tO

0

Figure 12(b) - PPU B narrowband conditions.

i".I-

Figure 1

]52.60 _5 2.70

Frequ_'y,

2(c) - Ambient broadband conditions.

135

t_

106

_6

85

55

55

45

2.50 _ 2.70

F_, GHz

Figure 12(d) - PPU B broadband conditions.

Figure 12 - Narrowband and broadband radiated emissions in the S band.

27

100

gO

7O

W50

j°3O

20

10

05.9

Figure 13(a)100

9O

a0

7O

m _a i

J5.0 6.1 5.2 5.3 5.4

Frec_, Gt'lz

- Ambient narrowband conditions.

5.O 5.1 5.2

' Frequency,

5_ 6.4

Figure 13(c) - Ambient broadband conditions.

W

125

115

j 5555

45

355_

50

j°3O

2O

10

05.g f_O _,! 52 53 5.4

Figure 13(b) - Arcjet./PPU B narmwband emissions.

100

Figure 14(a)IO0

gO

_0 5.1 6.2 53 5,4

Frequency.

Figure 13(d) - Arcjet/PPU B broadband emissions.

Figure 13- Narrowband and broadband radiated emissions in the C band.

W

J° j_0

20 55

10 45

0 3514_ 14.1 142 14.3 14.4 14_ 14.0

Frequency,GI_

- Ambient narrowband conditions.

14.1 14.2 !43 14.4 14.5

Frequency,

Figure 14(c) - Ambient broadband conditions.

20

10

014.0 14.1 14.2 143 14.4 14.5

Frequency,

Figure 14(b) - Ar_jet/PPU B narrowband emissions.

t_

511 05

J° :i143 14.2 1¢.3 14.4 14.5

Frequency,

Figure 14(d) - Arcjet/PPU B broadband conditions.

Figure 14- Narrowband and broadband radiated emissions in the Ku band.

28

tO

gO

80

tO_0

j°30

20

10

027.5

• • • I .... I .... I

Frequency,

, , , [

ZU

Frequency,

Figure 1S(a) - Ambient narrowband conditions.

tO

gO

g0

tO

.j°111

0 , I

Figure 1S(b) - ArcjerdPP.U B narrowband emissions.

_S

115

1

l°ir5

:B

Figure

65

55

45

351 .... I ,. .I • , ,27_ 28.0 28.5 29.0 29.5 _.0

Frequency,

1S(c) - Ambient broadband conditions.

115

.l

.la

35

Frequency,GI_

Figure 1S(d) -Arcjet/PPU B broadband emissions.

_>

tt_

Q;tm

Figure 15 - Narrowband and broadband radiated emissions in the Ka band.

......._.......hr-ii-I.......I.......I......................................,llitllIILt,,l,,I,,,.,I,I,,

.,_...... .JiU|iil_i_lhdltta|iliA,iLl _.[_,_,_,i,Ji,.,,,,z,_.,I,.,_L,_,,,-,,o-

_llll'IrlItlllI_I_,I-_' '

___....,,,,,,,,,__,-_________:...........I

0 Time, 200 ns/div

Figure 16(a) - Arc ignition transient over extended time scale.

_o- i i /1/'V_ , -: / , _0 , | ,

..... . ...............I

0 Time, 20 ns/div

V

Figure 16(b) - Arc ignition transient expanded time scale.

Figure 16 - Arc ignition transient captured with biconical antenna (20 to 300 MHz).

29

| ...,,. i.

I'L't'll"|h imt_l_

_Pd) I'

0 Time, 100 ns/div

Figure 17(a) - Arc ignition transient over extended time scale.

I

I

'\ /o-,_,,_,\_, _1 I/_ _f_,_" V _ _ _J _ ;

..... J -- --_ .................. I

I

,,!

i

0 Time, 20 ns/div

Figure 17(b) - Arc ignition transient expanded time scale.

Figure 17 - Arc ignition transient captured with broadband dipole antenna (160 to 500 MHz).

?0-

65

_ 55-: 50-

45-

40 _

35

10 lOO lOOO

Frequency,MHz

Figure 18 - Fourier spectrum of arc ignition transient shown in Figure 16(b).

3O

96-

|I i

!

I!

iI i

---4 ....... _'-

gin

!

I i!

I !I

! i

1 . l .

0Time, 200 ns/div

Figure 19 - Startup conducted voltage transient on primary power line for main power application to PPU B.

31

Form Approved

REPORT DOCUMENTATION PAGE OMBNo. 0704-0188

Public.rel_xting.1_. en. for this collection of informationis estimated to .average 1 hourper .respor_., includingthe timefor revlewi_.lnp,mJctiofls,tearching existing¢la_asoumes,_llhe .ffilg_ malnt_nlflg the data needed, _ co_..ing _ reviewingthe coltectlorlof mlormatlon. Send oo.n_ts regardingthe. burdenel_imale or any othor aspect of lhiscolk)ct)onof infon'r_ion, includingsugge6tionsfor reduang thm burden, to WashingtonHea_uaners Setwces. Directoratefor InformationOperationsand Reports, 1215 JelfetsonDavis Highway. Suite 1204, Arlington.VA 22202-4302, and to the Offto9o( Management and Budget.Pape_ork Red_clionProject (0704-0188), Wash_gtofl, DC 20503.

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE

June 1993

i 4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

Low Power Arcjet System Spacecraft Impacts

6. AUTHOR(S)

Eric J. Pencil, Charles J. Sarmiento, D.A. Lichtin,

J.W. Palchefsky, and A.L. Bogorad

7. PERFORMINGORGANIZATIONNAME(S)ANDADDRESS(ES)

National Aeronautics and Space AdministrationLewis Research Center

Cleveland, Ohio 44135-3191

9. SPONSORING/MONITORINGAGENCYNAME(S)ANDADDRESS(ES)

National Aeronautics and Space Administration

Washington, D.C. 20546-0001

3. REPORTTYPEAND DATESCOVERED

Technical Memorandum

WU-506-42-31

8. PERFORMING ORGANIZATIONREPORT NUMBER

E-8048

10. SPONSORING/MONITORINGAGENCY REPORT NUMBER

NASA TM- 106307AIAA-93-2392

11. SUPPLEMENTARY NOTES

Prepared for the 29th Joint Propulsion Confta, ence and Exkibit cosponsored by the AIAA, SAE, ASME, and ASEE, Monterey,

California, June 28--30, 1993. Eric J. Pencil and Charles J. Sarmiento, NASA Lewis Research Center, D.A. Lichlin, J.W. Palchefsky,

and A.L Bogorad, Martin Marietta Astro Space, Princeton, New Jersey 08543. Responsible person, Eric J. Pencil, (216) 433-7463.

12a. DISTRIBUTION/AVAILABILITY STATEMENT

Unclassified - Unlimited

Subject Category 20

12b. DISTRIBUTION CODE

13. ABSTRACT (Maximum 200 words)

Application of electrothermal arcjets on communications satellites requires assessment of integration concernsidentified by the user community. Perceived risks include plume contamination of spacezzmft materials, inducedarcing or electrostatic discharges between differentially charged spacecraft surfaces, and conducted and radiatedelectromagnetic interference (EMI) for both steady state and transient conditions. A Space Act agreement betweenMartin Marietta Astro Space, the Rocket Research Company, and NASA's Lewis Research Center was established toexperimentally examine these issues. Spacecraft materials were exposed to an arcjet plume for 40 hours, representing40 weeks of actual spacecraft life, and contamination was characterized by changes in surface properties. With theexception of the change in emiuance of one sample, all measurable changes in surface properties resulted in accept-able end of life characteristics. Charged spacecraft samples were benignly and consistently reduced to groundpotential during exposure to the powered arcjet plume, suggesting that the arcjet could act as a charge control deviceon spacecraft.Steady state EMI signatures obtained using two different power processing units were similar toemissions measured in a previous test. Emissions measured in UHF, S, C, Ku and Ka bands obtained a null resultwhich verified previous work in the UHF, S, and C bands. Characteristics of conducted and radiated transientemissions appear within standard spacecraft susceptibility criteria.

14. SUBJECT TERMS

Arcjet thruster;, Electromagnetic interference; Spacecraft materials; Integration testing;Electrostatic discharge; Electric propulsion

17. SECURITY CLASSIFICATION

OF REPORT

Unclassified

18. SECURITY CLASSIFICATION

OF THIS PAGE

Unclassified

NSN 7540-01-280-5500

19. SECURITY CLASSIFICATIONOF ABSTRACT

Unclassified

15. NUMBER OF PAGES

3216. PRICE CODE

A0320. LIMITATION OF ABSTRACT

Standard Form 298 (Rev. 2-89)Presctioed by ANSI Std. Z39-18298-102