AI AA-94-2856

Performance Evaluation and Life Testing of the SPT-100

Charles E. Garner*, John R. Ilrophy,** James E, Polk*, and [Link] C. Plcss+”
Jet Propulsion laboratory
California lnslitute of Technology
Pasadena, Ca 91109

indicated that the acmal performance of the thruster appears

CIOSC to the claimed performance of 50% efficiency at a
A cyclic endurance test of the Russian 1.35 kW Stationary specific impulse of 1600 s.] ‘Studies indicate that for
Plasma Thruster SIT- 100 is described, The cndurancc test is north/south s[ation keeping and Earth orbit raising
schcdulcd for 6,000 on/off cycles and 5,000 hours of applications of electric propulsion, the optimum speeific
operation at an inpui power to the thruster of 1.35 kW. impulse is in the range of 1,000-2,000 SCC.12 The
Cycles arc 50 minutes of thrusler on-time and 23 minutes of combination of the flight heritage of the SPT-70 and the
thruster’ off-time, To date 5,000 cycles and 4165 hours of availability of thrusters and thruster data, has led to increased
thruster on-time have been completed. Thruster efficiency interest in these thrusters by Western spacecraft
manufacturers for primary and auxiliary propulsion
decreased from 50% to 45% as the tfrrustcr aged over the firs[ applications,
1,000 hours; efficiency increased slowly over the next 1,000
hoursand has remained essentially unchanged bclween 2,000- Loral is presently flight-qualifying 1.35-kW, SPT-
4,000 hours of opcraling time, despite considerable erosion 100 thrusters for nordl/south station keeping and Earl]] orbit
of the discharge chamber walls and the downsmcam face of raising applications and plans to provide these thrusters on
the thruster, There are no known cases where the thruster their spacccraf[. ] 3 At Design Bureau “Fakcl”, lcwated in
failed to start in the cyclic endurance test. The cndurancc test Kaliningrad (Baltic coast) a sleady-state life test is being
is being performed under a cooperative program bctwccn pcrforrncd 14, and performance, plume and EM I/RFI
Space Syslcms/Loral, JPL, and the Ballistic Missile Dcfcnsc evaluations are being conducted at NASA Lewis Research
Organization (BMDO). Center. 15- 18A kcy aspect of the SPT-1OO evaluation
program is characterization of the long term operating
I)IJU . ION behavior of the thruster. Typical mission applications of
interest require operating times of several thousand hours.
Stationary plasma thrusters (SPT) arc gridlcss ion Potential usc of the thruster for norlh-south stationkecping
thrusters that were originally developed in the U.S. in the of commercial communication satcllilcs will also require the
early 1960s.1’3 Although efforts in the U. S, to develop capability for several thousand on/off cycles. To address
high thrust cfficicncics failed, efforts in the former U.S.S.R. these objectives a cyclic cndurancc testis being performed at
were quite successful; the SPT was successfully dcvclopcd JPL under a cooperative program between Space
during the 1960’s and 1970’s by Morozov4 and others5~6 Systems/Loral, JPL (under the JPL Affiliates program) and
with a unique combination of specific impulse and the BMDO. The endurance testis scheduled for 6,000 on-off
efficiency. It has been reported that more than 50 SPT-70 cycles and 5,000 hours of operation at an input power of
thrusters have flown in space, starting with the Meteor 1 in 1.35 kW. A previous paper dcscribcs the results of the
1969 -1970.7~8 More reccmtly, 8 SPT-100 thrusters were endurance test for the first 800 cycles. 19 This paper
dcscribcs the preliminary results of this cyclic enduranc.c test
flown on the Russian GALS communication satellite.9
through the first 5,000 cycles.
Reportedly, all thrusters arc performing rouline
stationkecping functions,
~Y
In 1991 a team of electric propulsion specialists
visited the former U.S.S.R. to experimentally evaluate the The Russian SPT-100 plasma thruster used in this
performance of a 1.35-kW SPT at the Scientific Research cndurancc test is shown in Fig, 1. “f’his thruster as tested
Institute of Thermal Proeesscs in Moscow and at “Fakcl” comes with two redundant hollow cathodes and a xenon flow
control system (XFC) cncloscd in a box directly behind the
Enterprise in Kaliningrad, Russia, ] ‘~] 1 The evaluation SPT- 100 (Fig. 2). The outer insulator which forms the
outside [Link] chamber wall is approximately 100,8 mm-
* Member of the Technical Staff, Advanced Propulsion dia.
Technology Group.
+ Member of the Technical Staff, Electric Power Systems The SPT-1OO thruster cndurancc test is being
Section. performed in a 3. l-m dia. x 5.1-m stainless steel vacuum
** Supervisor, Advanced Propulsion Technology Group. chamber equipped with three each, 1.2-m diameter helium
cryopumps. The minimum no-load tank pressure was
observed to bc 3,2 x 10-8 Torr. The nominal pumping speed
for the lhrcc pumps combined is 81,000 liters/s on xenon, and is loca(cd directly behind it in the vacuum systcm.
however the mue pump speed on xenon is approximately Thcrcforc pressure in the propcl]ant tubin~ is above
50,000 I/s,lg atmospheric pressure up to the SPT XFC. ‘Any xenon
contamination should then kc duc to contaminants picked up
The thruster is mounted near onc cnd of the from tubing walls, propellant system components, or the
vacuum tank, directly facing a cryopump which is positioned xenon itself, The purity of the xenon used by the SPT was
at the other cnd of the vacuum tank. The discharge chamber tested from the xenon bottle. Purity data indicate that the
surfaces of the SPT thruster consisl of insulator materials, specifications (99.999Yo) were excccdcd.
and it has been demonstrated that the performance of SFT
thrus[crs can be significantly affected by the deposition of a A thermal mass flow meter was used to rncasurc
conducting coating on the insulator.20 To protect the propellant flow rate. The flow meter was calibrated on both
cryopump and minimize the amount of material sputtcrccl nitrogen and xenon by the manufacturer using a primary
back to the thruster, the facility was lined with graphite, and calibration strrndard, and at JPL using a bubble volrrmctcr.
a graphite beam targcl was constructed, The beam targcL The nitrogen calibration pcrformccl at JPL agrees with the
consisls of 6.4-mnl thick graphite panels arranged in a primary standard calibration to within 1 YO for all flow rates
chevron configuration and placed as shown in Ref. 19. tested. The bubble vohrmctcr data on xenon were curve fit
Graphite was selcctcd as the targcl material bccausc of its and the curve fit was incorporated into the SPT data
low sputter yield at the ion energies expcctcd from the acquisition and control program.
SPT.2 ] The chevron configuration results in Iargc angles
(typically greater than 50 degrees) between the cxpcctcd ion A xenon recovery systcm is used to recover and
trajectories and the direction normal to the graphite surface, store xenon consumed by the SPT- 100. The rccovcry
which may both rcducc the graphite sputter rate and rcducc systcm is uscxl in the event of a cryopump faihrrc or when
the amount of sputtered material directed back towards [he 3000 liters of xenon have been consumed. A mechanical
thruster. A photograph of the SPT-1OO mounted inside the pump in the rccovcry system pumps the xenon out of the
vacuum chamber is shown in Fig, 2. vacuum tank and into a holding tank chilled with liquid
nitrogen; the vacuum tank is pumped by the rccovcry systcm
Duc to the beam divergence characteristics of the pump to ICSS than 0.4 Torr pressure. Oil traps prevent
SPT- 100, material can bc sputtered from the vacuum tank rccovcry systcm pump oil from entering the vacuum tank.
sidewalls and deposited onto the thruster. Thcrcforc the
cylindrical side walk of the vacuum chamber were also lined A probe rake, consisting of 25 Faraday probes of
with graphite panels. The graphite beam target panels were diameter 2.3 cm, mounted on a semicircular arc 2.4 m ‘in
baked out at high temperature (approximately 2300 K) prior diameter was used to examine the thruster exhaust plume.
to installation in order to minimize the potential for The probe butlons were biased to -23 volts when used to
hydrocarbon contamination. The side panels, which arc measure the plume shape. The probe rake was positioned
subject to a much lower thermal load during thruster such that the thruster is at the center of the semicircular arc;
operation, were not. Glass slides were placed 21 cm to either the axes of the rake were aligned with the plane formed by
side of the SPT thruster such that material back sputtered to the outer insulator of the SPT-100 such that the rake can bc
the SPT could be quantified and characterized. pivoted around a line normal to the thruster axis, This
configuration enables complete hemispherical profiles of the
Tank pressure was measured using two ion gauges. exhaust pltrmc to bc obtained. When the probe rake is not in
Onc gauge tube was mounted directly to the outer wall of USC, a motor rotates the rake to a position of approximately
vacuum tank; the other tube was mounted inside the vacuum 90 degrees with respect to the thruster axis.
tank, approximately 0.51 m above and 0,58 m behind the
SPT-1OO, This tube was calibrated on xenon and nitrogen The SPT-1OO is mounted (o an inverted pendulum
using a spinning rotor gauge that is traccablc to N] ST, data style thrust stand of the type dcvclopcd at NASA LcRC;22
from this calibration were used to generate a curve fit that in this design, thrust is indicated by a Iincar voltage
was used to calculate tank pressure. displacement transducer (LVDT). The thrust stand is
surrounded by a water-cooled housing to minimize
The propellant systcm for supplying xenon to the tcmpcraturc cffccls on the measured thrust; the housing can
SPT-1OO thruster is dcs@ed in Ref. 19. The system was bc seen just below the SPT-1OO in Fig. 2. The thrust stand
constructed from 0.64 -cm-dia stairdcss-steel tubing that was inclination is adjusted continuously by computer to improve
scrubbed with acctonc and alcohol before assembly. Prior to the accuracy of the thrust measurement over extended test
operating the SPT, the propellant systcm was checked for times. Thrust stand calibrations are performed in-situ
dcw point, hydrocarbon contamination, and particulate throughout the life tcs[ in the vacuum chamber using a set of
contamination. weights deployed with a motor. Normally 15 or more
Pressure was indicated by a capacitance manometer calibrations arc pcrforrncd in order to obtain a large enough
that was calibrated to an accuracy of f0,25% at 249.94 kPa, sampling for statistical analyses, Statistical analyses of the
the nominal working pressure for the SPT- 100. It should bc calibration data was also used to dctcrminc the thrust stand
noted here that the pressure is dropped to the level rcquirti by inclination. Repeatability of the calibrations were normally
the cathotlc and discharge chamber within the xenon flow 1.0% or bct[cr.
control system (XFC) that is an integral part of the SPT- 100

2
 The thruslcr is orwatcd
. with a bread board r)owcr
‘
PCU stop command to the PCU, opens a relay bctwccn the
conditioning unit (PCU) dcvclopcd by Space S yskmM/l,oral. PCU and its power source, and activates a tclcphonc dialing
Preliminary data imply that the bread board PPU is machine (autodialcr). The computer sends a change-oDstate
approximately 9390 efficient. The SPT- 100 discharge. current signal every 15 seconds to an electronic timer (heartbeat
and magnetic field currcnl arc adjuslcd by supplying the box); in the event of a computer failure, the timer activates a
appropriate voltage input signal to the PCU. The PCU series of relays to turn off PCU power, xenon flow, and
ou[put voltage is fixed at approximately 300 V. Propcllanl activates the autcdialcr,
flow rate was controlled by lhc SPT-1OO flow control unit
which is part of the thruster; the mass flow rate was The data acquisition program averages and stores
dctcrmincd by the discharge current selling. The PCU was PCU lclcmctry for discharge current, cathode healer current,
turned on and off by supplying the appropriate digital voltage capillary current, discharge voltage, floating voltage and flow
pulse to the PCU. Once the engine paramclcrs were adjusted rate. PCU tclcmctry for various SPT- 100 currents were
and the PCU was hrrncd on, theSPT-100 thruster was started calibrated to values obtained from calibrated current shunts,
and operated automatically by the bread board I’CU. The The discharge current calibration was dctcrmincd using a
PCU sequencing is described in Table. 1. voltmeter which averages the direct-current value of the
discharge current shunt voltage drop over a period of four
Table 1, Power Conditioning Unit Scqrrcncing seconds. Oscillations in the discharge current were obtained
— with an inductive probe placed on the discharge current
cable CIOSC to the vacuum tank fccdthrough, The discharge
Time Action voltage ripple was measured at the vacuum tank fcedthrough
m as well, using a combination inductive/Hall effect probe,
. Cabling length bctwccn the fccdthrough and SPT is
approximately 6 m. After cycle 663 of the cyclic life test
0 PCU start command received from the RMS vahrc of the discharge current was measured using a
computer true RMS voltmeter.
10 Cathode heater, capillary currcn[, The thruster was photographed (from an off-axis
magnetic field currcn[ on view) periodically through a window in the vacuum system
to document the condition of the thruslcr. Insulalor
170 Cathode igniter vohagc applicxi; cathode thicknesses arc determined from photographs by measuring
heater off. the ratio of insulator width to the length of the outer edge of
the outer insulator or inner edge of the inner insulator. An
171 SPT achicvcs 1.5 A disehargc; run time off-axis view of the SPT-1OO is shown in F3g. 3. The SPT-
clock started, begin on-phase 100 can be seen operating in lhc endurance tesl facility in
Fig. 4.
180 SPT achicvcs 4.5 A discharge
J’ROCI 11) RFL
3180 SPT tumcd off by stop command issued
by compulcr; begin off-phase The SPT- 100 was purged with argon when the
mechanical pumps were used to pump the vacuum tank from
4380 PCU starl command rcccived from atrnosphcrc to 50 mTorr; as of this writing the WI’- 100 life
computer for next cycle test thrusler has not been exposed to atmosphere since it was
—— installed on June 22, 1993. The SPr-100 is operated only
if the vacuum tank pressure indicated by the calibrated ion
Steady-state thruster operation, slart-up and gauge reads below 2 x 10-7 Torr. During a facility shutdown
shutdown sequencing are controlled by a PC based da[a (cryopumps of~ the SPT- 100 is purged with xenon.
acquisition (DAC) systcm. This syslcm also monitors the
vacuum facility enabling unattended operation. A total of 56 A cycle is defined as any time the thruster achicvcs
channels that include thiusl, xenon mass flow rate, anode a discharge current of >1.5 A. The first 25 cycles of the life
voltage and current, floating voltage, magnet current, cathode test were used to test the data acquisition and control
program, the facility, and the probe rak~ in these cycles the
heater current, thcrrnothrotde current, SPT inlet xenon thruster was opcra[cd for varying time periods, from ICSS than
pressure, tank pressure and various other facility components onc minute to over 60 minutes, and at varying discharge
are monitored and recorded as a function of lime. The data arc currents.
averaged and tic averaged values are displayed on a monitor
scrccn, The data arc recorded on the computer hard disc drive The computer performs the task of starting and
every 60 seconds, and prinlcd every 120 seconds, stopping the thruster, taking data, and monitoring the
facility. Life test cycles arc nominally 50 minutes on and 20
The computer is responsible for issuing the PCU minulcs off, with an additional three minutes for cathode pre-
start and stop commands. If certain engine or facility heating. Oscilloscope traces of current oscillations in the
parameters cxcccd spccifica[ions, the computer sends the discharge current and a,c. ripple in the discharge voltage are

3
obtained approximamly every 10 cycles. Approximatc]y
every 200 hours the life tcsl was intcrruptccl for a short
period of time 10 photograph the lhrustcr, measure flow
meter zero drift, and to rc-calibrate the thrust stand (T/S). Tab]c 11, Test/shutdown descriptions
Approximately every 200 hours a probe scan was obtained -. —.. —— ________ .
using the probe rake. . - — _________ —
Shutdown Description Number of Shutdowns
(,’
B~suI 7S AND DISC[JSSION Cycle stopped, thruster 2
cmitLing sparks
‘1’hc SPT- 100 has complclcd 5,000 starts and 4165
Cryopump failures 20
hours of operation; of these, 86 cycles totaling 26.64 hours
were not opcralcd for the standard 50 minutes bccausc of PCU/PCU input power 18
testing requirements, compulcr/facility failures, opcralor supply failures
error, or to computer-commanded shutdowns. All shutdowns
have occurred duc to facility/hardware failures, PCU or PCU 2
Power grid transients
power input failures, operator errors, or computer software
failures, There. have never been any shutdowns required duc
to SPT operation- in all cycles lhc SPT has always sLarlcd Operator errors 6
and performed to its nominal operating condition of 1.35 7
kW. Compu[cr/DAC errors
or hardware fai lures
A glow in the non-operating calhodc was observed Thruster/t)AC tcs[ing
in cycle 1 and all subsequent cycles. The glow may bc duc 30
10 a propellant leak in the SPT XFC. This SPT-100 To”I’AI. 8S
thruster utilizes a propellant systcm with no absolute flow
——-. —
shut-off to the unused cathodc,23 a design feature present in
newer SPT thrusters, The cathode glow occasionally flickers
on engine start-up and periodically disappears complctc]y for shorl lime; the partial pressure was mostly xenon
approximately 3 seconds before reappearing. The glow in evaporating off of the pump surfaces, Xenon flow through
the unused cathode is visible in a photograph (Fig. 5) of the the SPT was maintained when a cryopump was turned off,
SPT- 100 in operation. The brightness of the glow may have and tank pressure never excccxl 0.1 Pa.
decreased with time.

Starting with cycle 26 theSPT-100 was cycled for

50 minutes on and 20 minutes off, with 3 minutes for TEST DATA FOR CYCLES 26-500Q
cathode pm-heating. Four cycles were “double-booked) so the
total actual number of cycles reported herein should bc Engine parameters such as efficiency, discharge
incremented by 4 (after C822). In 85 of the 5,000 cycles current and voltage, etc., for cycles 26-5,000 are shown in
accumulated to date the thruster was opcratd for lCSS than the Figs. 6-9. The data rcprcscnt values for 242 individual
usual 50 minutes for a variety of reasons, The shutdowns Cycles. Data through cycle 2,000 were calculated by
arc categorized in Table 11, averaging computer dala for the last tcn minutes of each
cycle. Discharge current and voltage data obtained using
During cycle 93 a sof[warc error resulted in voltmeters arc more consistent compared to data obtained
operation of the SPT thruster for approximatcl y 30 minutes from the data acquisition systcm, bccausc the DAC or the
without sufficient xenon flow; no data for this cycle were PCU telemetry arc affcctcd by thruster operation, Thcrcforc
printed or stored. The software error was corrected and tested after cycle 2,000 values for discharge current and voltage read
during cycles 94-97. Cycle 4657 was operated for 5.6 hours from voltmeters were used in Figs, 6-9.
duc to a computer/software error, During most of this cycle Thruster data were analyzed to determine cycle-to-
lhc thruster operating characteristics were not monitored and cyclc changes in thruster operating characteristics. Thrust
data were not stored by the computer. The thruster was was determined by subtracting the LVDT voltage four
operating at nominal values of discharge voltage and minutes after the SPT was turned off from the LVDT voltage
discharge current when the problcm was discovcrcd, obtained from the ten-minute average, and multiplying by
the appropriate thrust stand calibration factor. Efficiency
20 shutdowns were duc to cryopump failures. Most and specific impulse were calculated using the values for
of the shutdowns were commanded by computer when a thrust, mass flow rate, and engine power, averaged fc~r the
cryopump began to fail and vacuum tank pressure excccdcd last 10 minutes of the cycle.
0.013 Pa; tank pressure did not exceed 0.04 Pa during these
Large variations in thruster parameters such as
failures. In two of the faihrrcs the computer commanded the floating voltage and thrust occurred in the first 900 hours of
cryopumps off and tank pressure approached 0.1 Pa for a thruster operation, In most of the cycles the variations were

4
incluccd by applying supplementary current to lhc magnc~ slipped on the chain holding the weights and is in the wrong
coils to investigate thruster performance- and oscillations in posi[ion for a calibration.
the discharge current and vohagc. Most of [hc variations in
thruster performance evident in o[hcr cycles arc related This scenario should not affect the calibrations of
dircclly 10 the fact that when the thruster is shut off for more the other two masses. 1 t will bc noted here, however, that
than a fcw hours and then restarted, for a fcw cycles aficr the carlhquakc the LVDT voltage diffcrcncc over time
following the start discharge currcnl and thrus[ increase, and rcvcrscd its direction and began to incrcasc in magniurdc
current/vohagc oscillations dccrcasc. This behavior has been (Fig. 11). Scvcrc shaking from the carlhquakc may have
observed rcpcatcdly, including those limes when the tank changed the spring constant of the thrust stand systcm.
pressure did not cxcecd 10-5 Pa. Calibrations have con[inucd to have standard deviations of
Icss than 1.25Y0.
Discharge volt&!&
Efficiency and St)cc ific Impti
The dccreasc in discharge voltage bc{wcxm run hours
1S0-900 was duc to failures of onc of the PCU discharge 13c[ween run hours 1-1,000 thrust and efficiency
supply modules; without all modules functioning the engine dccreascd, [hen incrcascd bctwccn run hour 1,0002,000.
discharge loaded down lhc PCU and the oulput voltage Since approximately run hour 2,500 thruster performance has
dropped. All modules bccamc functional after run hour not changed significantly, cxccpt bctwccn run hours 2,650-
1400, Duc to noise in the DAC voltage mcasurcmcn~s af[cr 2,850, where thrust incrcascd by approximately 3 mN; it is
run hour 2000 were obtained from voltmeters. not understood how or why the thrust incrcasc occurred, but
analyses of the computer data indicate that the thrsulor most
Mm Flow Rate and Discharizc Cu rrcnl of these cycles was greater than ncwmal. These dam arc
surprising; however, the general shape of the efficiency vs
Mass flow increased 3% and discharge current lime curve shown in Fig. 9 matches WCII with the data
dccrcascd 1% [Link] run hour 8-4,000. The thermal mass obtained from a steady-state life test being performed in
flow mclcr calibration drifted approximately 1% bctwccn run Russia]4 at the Design Bureau Fakel. Quantitatively, the
hour O-1 ,740; in the data prcscntcd in Figs. 6-9, the mass thrust measured at JPL is approximately two mN greater
flow indicated by the flow meter was adjusted with the than thrusi measurements obtained at Fakcl. Theoretical
assumption that the drift rate was uniform bclwccn run hour analyses of the thrust incrcasc with time will be prcscntcd in
0-1,740, Variations in discharge current arc duc in part to Ref. 14.
shifts in PCU tclcmctry output,
Pischarflc Currcnt Oscillations and Voltage Ripple
3’hrust Mcasu rcmcnts
Oscillations in discharge current and voltage arc
Thrust measurements arc obtained from an invcrtcd- shown in Figs.12-l 5, The probes used to measure discharge
pendulum thrust stand, In-situ calibrations were performed current and voltage oscillations were positioned close to the
periodically to obtain the thrust stand calibrations shown in vacuum tank fccdthrough. There is approximately 6 m of
Fig. 10. In Fig. 11 the voltage difference of the LVDT is cable length between the tank fccdthrough and the thruster.
plotted as a function of tirnc for the three masses used for the
calibrations. To increase repeatability of the thrust In Ref. 19 rcduccd current and voltage oscillation
measurements the thrust stand plalform which supports the amplitudes were associated with improved thruster
SPT-1OO was wiggled for five or tcn seconds during performance. Initially, discharge. current oscillation
calibrations and at the start and cnd of each cycle Typically, amplitudes were generally less than 2 A p-p, and voltage
calibrations arc bctlcr than onc pcrccrm LVDT drif[ in the ripple Icss than 2.5 V p-p, By cycle 1,100 (run time 887
thruster off-phme was studied; for all thrust measurements 30 hrs) current oscillation amplitudes up to 10A p-p, and up to
mV was subtracted from the LVDT vollagc difference 15 V p-p ripple in the discharge voltage could be found.
between the on and off phases to account for LVDT drift, Oscillation amplitudes began to decrease at approximately
Thrust stand tilt values for the on and off phases were also cycle 2,500 (2,043 hrs run time); on cycle 5,000 current
rccordcd; if the till varied bctwccn the on and off phases, a oscillation amplitudes were 6 A p-p maximum and 7 volts p-
correction factor to the LVDT voltage diffcrcncc was applied p for the discharge voltage.
to account for changes in LVDT vohagc duc to diffcrcnccs in
thrust stand inclination. Insulator Erosion

Until approximately run hour 2,600 the LVDT Erosion of the discharge chamber insulator surfaces
voltage diffcrcncc cxhibiled a predictable change over [imc was documented photographically and is plotted in Fig, 16 as
(Fig. 10); On January 17, 1994, about 10 minutes after a function of run time. The photographs indicate that by
cycle 2,674 (run hour 2122, 14) ended, a magniludc 6,9 app. run hour 1,200 the edge of the outer insulator was
earthquake struck Los Angeles. As a result of [Ilc scvcrc eroded cornplctcly away; by run hour 2,600 the downstream
shaking the LVDT off-phase voltage 13Yo; in addition, it was Pdcc of the SPT-100 was being eroded by ion bombardment
no longer possible to obtain thrust stand calibrations of the bccausc the outer insulator was worn flush with the thruster
second of three weights; it is possible that the second mass surface. The inner insulator has been rcduccd in thickness to
17% of the original thickness. Preliminary results indicate
an erosion rate that is approximately the same as reported in
Ref. 24, cxccpt for data oblaincd” al JPL on [hc inner
insulator at run hour 4,000. Thruster performance continues 1. Lary, E.C., Mcycrand, R.G., Jr., and Salz, F., “Ion
to bc very acceptable despite the cxtrcrnc wear of the Accclcration in a gyro-dominakxl neutral plasma--theory and
insulators and the downstream face of the thruslcr. cxpcrimcnl,” Bul. Am. Phys. Sot. 7 441 (1962),

Wear characteristics of lhc SPT-100 arc shown in 2. Scikcl, G. R., and Rcshotko, E,, “Ilall-current ion
Figs, 17-22, The dark material on certain locations of the accclcralor,” Bul. Am. Phys. Sot, 7414 (1962),
SPT-100 may be uncrodcd deposits of graphite spuucrcxi
from the graphite beam targc[, The unused calhodc (bottom 3. Jones, G. S., Dotson, J., and Wilson, T.,
cathode) has eroded considerably. The orifice of (1IC “Elcctrostalic accclcrdtion of neutral plasmas -- momentum
functioning cathode has incrcascd in diarnctcr and is tmnsfcr through magnc[ic field,” Proceedings of the Third
evolving into an ellipsoid with onc end, the side facing away Symposium on Advanced Propulsion Concepts (Gordon&
from [hc SPT- 100, eroding the most. Breach Science Publishers, Inc., New York, 1963) pp. 153-
175.
CON~LL!MQNS
4. Morozov, A.1, ct al. , “P1asn~a Accelerator with
An cndurancc test of an SPT-1OO is schcdulcd for Closed Electron Drifl and Extended Accclcration Zone,”
6,000 on/off cycles and 5,000 hours of operation at an input Soviet Physics -- Tech. Physics, Vol. 17, No. 1, July 1972.
power of 1.35 kW. The cndurancc test was ini[iatcd July 1,
1993 and has accumulated 4,165 hours of operation and 5. Bugrova, CL al., “Physical Processes and
5,000 on/off CYCICS as of this writing. The nominal cycle Characteristics of Stationary Plasma Thrusters with Closed
duration is 50 minutes on and 23 minutes off, including Electron Drift,” IEPC-91-079, October 1991.
nearly three minutes of cathode preheat time. Thruster
efficiency dccrcascd, from 49% to 44$?0 as the thrusbm aged; 6, Bobcr, A, ct. al,, “State of Work on Electrical
lhrustcr efficiency incrcascd and has remained rcla[ivcly Thrusters in U.S.S.R:’ IEPC-91-003, October 1991.
constant after approxima[cly 2,000 hours of operating Lirnc.
7. Artsimovich, L, A., “The Dcvclopmcnt of a
Variations in thruslcr performance from cycle to Stationary Plasma Engine and its Test on Mckor Artificial
cycle appear related primarily to the whether or not lhc Earth Satellite,” Space Explorations, Vol. 12, No. 3, pp
[Link] was not operated for any significant time; this 451-468 (1974)
phenomena appears unrelated to tank pressure during the off-
phase. The insulators forming [he discharge cham~cr have 8, Arkhipov, B. A. CL al., “SPT Electric Propulsion
been heavily eroded, as has the downstream face of the System for Spacecraft Orbit Maneuvering,” Paper RGC-EP-
insula~or, Current and voltage oscillations incrcascd, then 92-07, Ist Russian-Gcmlan Conference on Electric
decrcasti 6 A p-p maximum at cycle 5,C00. Propulsion, March 1992.

9. Personal communication, Mike Day of Space

S yslcn]s/Loral, April 1994,
The authors thank Mr. Alison Owens and Mr.
Robert Toomath for their efforts in support of lhc lesting 10. Brophy, J.R., “Stationary Plasma Thruster
described in this paper. The authors gratefully acknowledge Evaluation in Russia,” JPL Publication 92-4, March 15,
the support of Dr. Lcn Caveny, Innovative Science and 1992.
Technology office of the Ballistic Missile Dcfcnsc
Organization. 11. Brophy, J,R. et al., “Performance of the Stationary
Plasma Thruster: SPTIOO,” AIAA-92-3155, 1992.
The wori described in this paper was performed by
the Jet Propulsion Laboratory, California Ins[itutc of 12. Patterson, M.J. ct al. “Experimental Investigation
Technology, and was sponsored by the Ballistic Dcfcnsc of a Closed-Drifl Thruster,” AIAA-85-2060, 1985.
Missile Organization/Innovalivc Science and Technology
office through an agreement with the National Aeronautics 13. Cavcny, L. H., Curran, F. M,, and Brophy, J. R.,
and Space Administration and Space Systems/Loral, Palo “Russian Electric Space Propulsion Evaluated for Usc on
Alto, California, American Satellites,” Paper RGC-EP-93-D2-4, 2nd
Russian-Gcnnan Conference on Electric Propulsion,
Moscow, Russia, July 1993,

Arkhipov, B. A. ct al., “SPT- 100 Module Lifetime

~’~~1 Results”, to bc presented at AIAA-94-2854, 1994.

6
15. Sankovic, J. c1 al., “Pcrfom~ancc Evaluation of d]c
Russian SPT-100 lhrustcr at NASA LcRC”) lflPC-93-094,
1993.

16. Meyer, R.M, and Manzclla, D. H., “SPT Thruster

Plume Charackx-istics”, lEPC-93-096, 1993.

17. Manzclla, D. H,, “SPT Thruster Plume Emissions”,

IWC-93-097, 1993.

18. Pencil, E,]., “Preliminary Far-Field Plume

Sputtering of the SPT-100, IEPC-93-098, 1993.

19. Gamer, C,E. et al., “Pcrformancc Evaluation and

[Link] Tcsling of the SPT-I 00, IEPC-93-091, 1993,

20. Vladimir Kim, personal communication, Moscow

Aviation Instimtc, June 26, 1993.

21. Matsunami, N. ct al,, “Energy Dcpcndcncc of the

Yields of Ion-Induced Sputtering of Monatomic Solids,”
IPPJ-AM-32, September 1983.

22. Haag, T.W. and Curran, F. M., “Arcjct Starling

Reliability: A Mullistart Test on Hydrogcrr/Nitrogen
Mixtures,” AIAA-87-1061, May 1987. (NASA TM-
898867).

23. T. Randolph, personal communication, Space

Systems/Loral, June 24, 1993.

24, Absalamov, S. K. et al., “Mcasurcmcnt of Plasma

Parameters in the Stationary Plasma Thruster (SPT- 100)
PIumc and its Effect on Spacecraft Components,” AIAA-92-
3156, July 1992,

25. Arkhipov, B. A. ct al., “Anomalous erosion of an

insulator under action of a stream of plasma,” Phys, Plazmy
18 124) -1244 (Sept. 1992).
Fig. 1. New SPT-1OO as tested. Fig. 2. SPT-1OO instaIIed on water-cooled thrust stand. Fig. 3. Xew SPT- lWl form off-axis view.

Fig. 4. SpT- 100” in operating in the JPL life test facility. Fig. 5. LTnused (bottom) cathode glow.
 ‘“’~-’o 4.60 7.0

4.55
g
6.5
: 4.50
L
3
u 4.45 6.0
al
u)
g 4,40 5
.—u) 5.5 ~
0
4.35 ~

2“~—’6 o 1000 2000 3000 4000

4.30
0 1000 2000 3000 4000
5.0

SPT-100 Operating Time (hrs) SPT-100 Operating Time (hrs)

Fig. 6. Oischarge and floating voltage for the SPT-l(XI cyclic endurance test Fig. 7. Discharge current and mass flow rate for the SPT-100 cyclic c

95~0”5’ 1700
0.60
0.50
90 1600
Z- 0.45 0.55
g n?
~
0.40 ~ 1500
2 0.50
0.35 %
80 1400
0.30 0.45

0.25 1300
7 I i I I
I

0 1000 2000 3000 4000 0 1000 2000 3000 4000

SPT-100 Operating Time (hrs) Thruster Operating Time (hrs)
Fig. 8. Thrust and efficiency for the SPT-100 cyclic endurance test. Fig. 9. Efficiency and lsp for the SPT-100 cyclic endurance test.
 ,,, ,., . . x
,,- “ ““—+
! +
-=.-----+-+- .
~:—-. --=&_
. . . . . . . . . . . . . .
, .,.
-=q=~
+. .
,, y---~

,: ,. ,.,
.;
,. + ,,, ,..
--.4
. ..~g
-~~-s
:>
---&=-
i “:”
. . . . ~:. ..-
..+ 4=?-
—-
““”s~”
,. . s
.73
. . . , ,.
k===
~ .,. . . . . . . . . . . . ..~.... ..; ,.,,..;.,,; ..~; ..; ,;., . :, ,—
. .; ., . , -
~

j q?
.:.
-=$=
~;
r-
-..
-- .j==+=&- s-
.>3=’
5i3=-
$;”’”
ccl

.T-
0)
-_~+=-
-. i w
q’””
,: ix
r-
.

;sE-=-

.3
.. .. .

(A) asuodsq-j PLMS lSnJLll

e m
o-l
, \
.
\
u)

0/
1 I
 :. .:

o
0
e

.,. o
0
0
C7

>
q
>
m
. ,,,. ., , . . . . ., -.
-==i-+------ *

& “%
. +. . .
0
0
“:”
(’ --

0
i “,’ r-

j
~,.
i +-s”-
3
.:.
!

. .s ‘-s
. -+ . . x“
“1”

-
‘+=4
..,. J . . . . . . . . .

~ i ““””””’-”’ ““”’’ ”-”

‘
,0
\ “’ 0 co * 0

. ,.: ., , .,
,.!. -,,.7 ..,.. 03 N
‘-=f=f=~
T
P,..:!..!

>~
T-

(Ullu) sswyqql Jo]131nsu[

. . ‘. . .
3!. ~.
““-’””f---
;*> .?
. ~., ., ,. , . , , . r.... ,
( &
i= },
~+.t +-
*>
.-++ , : “--- .
j:,
!,. ,

,.. .,.. .. ?..., ; , ... .

= %

. . . . ..-
A. . . . . . . . . . . . . . . . . . . . . . . . . . . ..2
—
Fig. 17. SPT-1OO after 1100 cycles and 887 hours Fig. 18. SPT.-1OO after 2200 cycles and 1795 hours Fig. 19. SPT-1OO after 3816 cycles and 3169
of operating time at 1.35 kW. of operating time at 1.35 kW. of operating time at 1.35 kW.

Fig. 20. SPT-1OO after 3816 cycles and 3169 hours Fig. 21. SPT-1OO after 4701 cycles and 3879 hours Fig. 22. SPT-100 after 4710 cycles and 387
of omxatinsz time at 1.35 kw. of operating time at 1.35 kW. of operating time at 1.35 kW.