NASA Technical Reports Server (NTRS) 19910020938: Preliminary performance and life evaluations of a 2-kW arcjet

31 Pages·12.6 MB·English

by  NASA Technical Reports Server (NTRS)

Iler ^ F/
4E-(^Oy3D
NASA Technical Memorandum 105149
AIAA-91-2228
Preliminary Performance and Life
Evaluations of a 2-kW Arcjet
W. Earl Morren and Francis M. Curran
Lewis Research Center
Cleveland, Ohio
Prepared for the
27th Joint Propulsion Conference
cosponsored by the AIAA, SAE, ASME, and ASEE
Sacramento, California, June 24-27, 1991
NASA
Preliminary Performance and Life Evaluations of a 2-kW Arcjet
W. Earl Morren' and Francis M. Curran'
National Aeronautics and Space Administration
Lewis Research Center
Cleveland, Ohio 44135
ABSTRACT
The first results of a program to expand the operational envelope of low-power arejets to
higher specific impulse and power levels are presented. The performance of a kW-class
laboratory model arcjet thruster was characterized at three mass flow rates of a 2:1 mixture
of hydrogen and nitrogen at power levels ranging from 1.0 to 2.0 kW. This same thruster
was then operated for a total of 300 h at a specific impulse and power level of 550 s and
2.0 kW , respectively, in three continuous 100-h sessions. Thruster operation during the
three test segments was stable, and no measurable performance degradation was observed
during the test series. Substantial cathode erosion was observed during an inspection
following the second 100-h test segment. Most notable was the migration of material from
the center of the cathode tip to a ring around a large crater. The anode sustained no
significant damage during the endurance test segments. Some difficulty was encountered
during start-up after disassembly and inspection following the second 100-h test segment,
which caused constrictor erosion. This resulted in a reduced flow restriction and arc
chamber pressure, which in turn caused a reduction in the arc impedance.
power processing units incorporating
INTRODUCTION high-voltage pulsed starting circuits have
been tested5-7. Extended, cyclic
Demands for high-performance auxiliary endurance tests on both laboratory $ and
propulsion systems on commercial 7
flight-type arcjet systems have been
communications satellites have driven an
completed. Other studies have focussed
intense effort toward the development of on the impacts of arcjet system
kilowatt-class arcjet propulsion systems. integration. Electron number densities
This is because the performance and temperatures have been measured in
advantages that these systems offer over both the near- and far-field arcjet plume
existing resistojet and chemical systems
using Langmuir probes9-12. The results
lead to significant reductions in north-
of these plume surveys have been used to
south stationkeeping propellant
model the effects of the slightly-ionized
requirements.
plumes on communications signals13,14
Testing of a flight-type arcjet system on a
During the 1980's arcjet system
spacecraft simulator directed toward the
development has focussed on meeting the
documentation of spacecraft arcjet system
technology goals necessary to bring these
interactions has been completed 15. The
systems to flight readiness. Stable and
culmination of these efforts has been the
reliable operation on hydrazine
selection of an arcjet system to provide
decomposition products at specific
stationkeeping on a new generation of
impulse levels of 450 to 500 s has been
commercial spacecraft16
demonstrated)-4. Pulse width-modulated
• Member AIAA
Copyright 0 1991 by the American Institute of Aeronautics and
Astronautics, Inc. No copyright is asserted in the United States under
Title 17, U.S. Code. The U.S. Govemmene has royalty-free license to
exercise all rights under the copyright claimed herein for Govemment
purposes. All other rights are reserved by the copyright owner.
2.0 kW. The accompanying text
Future specific impulse requirements are describes the test thruster, facilities, and
expected to increase, as are the limits of procedure. Also included are the results
power available to spacecraft propulsion of four performance characterizations
systems. Anticipating these requirements, conducted over the life of the lifetest
a program was initiated to expand the thruster, as well as the details of one
envelope of low-power arcjet operation interim and one post-test disassembly and
beyond the current state-of-the-art inspection.
(approximately 530 s mission-average at
1.6 kW into the thruster using hydrazine APPARATUS AND PROCEDURE
propellant) to 600 s and thruster power
levels of 2 to 5 M. The purpose of this Thruster, Figure 1 shows a schematic
paper is to present the results of the first diagram of the lifetest thruster, which
step toward that goal. was a modular laboratory design. The
cathode was a 3.18 mm diameter 2%
Arejet specific impulse is closely tied to thoriated tungsten rod with a cone of 300
the energy input per unit mass of half-angle ground on one end. The
propellant expelled. Increasing the state- anode, also fabricated from 2% thoriated
of-the-art specific impulse requires tungsten rod, incorporated a nozzle with a
increasing the ratio of input electric power conical convergent section with a 300
to propellant mass flow rate (referred to half-angle, a 0.55 mm diameter by 0.25
herein as specific power) and/or reducing mm long cylindrical constrictor, and a
the various losses inherent in arcjet conical divergent section with a 200 half-
operation. Thermal, frozen flow (i.e.,
angle. The expansion area ratio was 214.
energy not recovered as thrust), and
Figure 2 shows an exploded view of the
nozzle losses are dominate factors in cathode tip/anode constrictor region. The
arcjet efficiency. An investigation into arc gap (the minimum axial distance
means of reducing thermal losses is under
between the cathode tip and the anode)
way17 and an earlier effort showed that was set to 0.58 mm. This was
increasing nozzle expansion ratio accomplished by inserting the cathode
increased thruster efficiencytg. Additional into the thruster until it met the anode,
near-term programs are attempting to then withdrawing it by the desired gap
reduce frozen flow losses. This effort setting. The electrodes were contained in
seeks to enhance specific impulse by a housing fabricated from titaniated-
increasing the specific power. zirconiated-molybdenum (TZM). The
anode housing had a 0.25 mm thick layer
The first step toward expanding the of molybdenum powder plasma-sprayed
operating envelope beyond the current onto the exterior surface to enhance its
state-of-the-art was to identify the life- emittance.
limiting issues in the current thruster
technology at conditions beyond those The joint between the anode and the
typical of low-power arcjets. To this end housing was a tapered interference fit
a modular laboratory model arcjet thruster 6.35 mm in length with a 50 half-angle
was run for a total of 300 h in three and a minor diameter of 13.7 mm. This
continuous 100-h test segments at joint was lapped during thruster assembly
constant specific impulse and power to aid thermal conduction from the anode
levels of 550 s and 2.0 kW, respectively, to the housing and to minimize gas
on a mixture of hydrogen and nitrogen leakage. A TZM injector disk was located
simulating hydrazine decomposition immediately behind the anode. Two 0.38
products. Performance was measured at mm diameter holes in the disk injected the
propellant mass flow rates ranging from propellant into the are chamber, the small
3.0x10-5 to 5.0x10-5 kg/s and thruster plenum surrounding the cathode tip, so as
input power levels ranging from 1.0 to to set up a vortex flow field. This was
2
done because a vortex flow has been number 10 stainless steel bolts. Graphite
shown to improve starting and steady foil gaskets sealed this joint. Figure 3
state stability. However, no systematic shows a photograph of this design prior
study has yet been conducted to evaluate to assembly.
the effects of vortex strength on arcjet
stability. During operation, propellant was fed to
the thruster propellant feedthrough. The
Behind the injector disk was a boron gas flowed over the spring, through
nitride insulator (the front insulator in several axial channels in the boron nitride
figure 1). The front insulator served to ram and into a plenum between the front
provide electrical isolation of the cathode and rear insulators. From there the gas
from the anode-potential housing; to force flowed through sixteen axial channels,
the gas flow through several axial approximately 0.5 mm square in cross-
channels located next to the housing wall; section, cut into the outer surface of the
and to transmit the compressive force front insulator. This placed the gas in
imparted by the spring, located in the rear direct contact with the housing and
half of the thruster, to the front provided some regenerative cooling of the
insulator/injector and injector/anode forward end of the thruster. The gas then
joints. The compressive forces on the flowed through two 0.381 mm diameter
joints around the injector were required to holes in the injector disk and into a small
assure that the propellant flowed through plenum around the cathode tip (the arc
the injector ports and not around the chamber). The gas was injected into the
injector. Flowing the propellant next to are chamber so as to set up a vortical
the anode housing wall provided some flow. The gas then passed through the
regenerative cooling of the anode region. constrictor where it is heated by the arc
and is accelerated in the nozzle,
The upstream end of the thruster was producing thrust.
contained in the boron nitride rear
insulator. The propellant and cathode Propellant. The propellant used was a
feedthrough fittings were commercially- mixture of hydrogen and nitrogen at a
available compression fittings modified to mixture ratio of 2:1 by volume,
mate with flats machined into the rear simulating hydrazine decomposition
insulator. These joints were sealed with products. The pure (99.99% minimum)
gaskets made from 0.25 mm thick gases were stored in high-pressure
graphite foil. An inconel spring located bottles, then metered separately and
within the rear insulator provided the mixed upstream of the thruster. The
compressive force necessary to prevent mixture entered the thruster at room
leakage around the injector disk while temperature.
allowing for thermal expansion of the
internal thruster components. A boron Power Processing. Power was supplied
nitride plunger transmitted the to the thruster from a pulse width-
compressive force to the front insulator modulated power processing unit (PPU)
which in turn pushed the injector disk with high-speed current regulation. The
against the rear of the anode. Graphite PPU was capable of delivering up to 130
foil gaskets between the front insulator, Vdc and 50 A to the thruster. Starting
the injector disk, and the anode prevented was facilitated by a built-in, high-voltage
gas from blowing by the injector disk. circuit capable of providing a 4 kV pulse
Note that it was necessary for the front every second until breakdown occurred.
insulator to slide freely over the cathode Starting current surge protection kept the
within the anode housing for it to PPU from delivering maximum current
effectively transmit the springs force to until the arc had sufficient time to blow
the injector joints. The rear insulator was through the constrictor, seating in the
clamped to the anode housing by two 1.6 nozzle divergent section. Earlier efforts to
mm thick molybdenum flanges and four start without this surge protection often
3
resulted in constrictor damage due to transducer signals into engineering units,
momentary spot attachment in high- as well as calculation of several derived
pressure regions upstream of the parameters. These data were then
constrictor at high current levels. Further displayed in digital and graphical forms
details of this PPU have been detailed on the computer screen at a sampling rate
elsewhere19. of about 1.5 Hz. This system also
controlled automatic shutdown of the
Test Facilities. Two test facilities were thruster power and propellant flow in the
used to carry out these experiments. All event any monitored parameters deviated
performance characterizations were from designated ranges. This permitted
conducted in a cylindrical vacuum tank unattended operation of the thruster.
measuring 1.5 m in diameter by 4.5 m
long. This tank was equipped with four The arc voltage was measured at the
0.8 m diameter oil diffusion pumps power leads connecting the PPU to the
backed by a lobe-type blower and two test facilities. The are current was
rotary piston roughing pumps. The test measured using a shunt in series with the
cell background pressure was below 0.1 thruster. The current signal was
Pa for all performance tests. During processed using a low-pass analog filter
operation in this facility the thruster was with a time constant of about 0.5 s. The
mounted on a thrust stand in a horizontal filter was necessary to obtain a do-level
orientation, on the axis of the vacuum signal because the current had about 5%
tank. Figure 4 shows a photograph of an ripple at 20 kHz. Both voltage and
arcjet operating in the performance test current signals were then fed to the
facility. DACS through isolation amplifiers. The
DACS responses to the voltage and
Figure 5 shows a photograph of an arcjet current inputs were calibrated by
mounted in the endurance test facility, a removing the PPU from the electrical
cylindrical vacuum chamber 0.5 m in circuit, then applying reference voltages
diameter by 0.6 m high. This test cell was and currents to the signal conditioning
equipped with a rotary piston roughing system from a laboratory do power
pump capable of maintaining test cell supply.
pressures of approximately 100 Pa during
all endurance tests. During operation in Propellant mass flow rates were
this facility the thruster was mounted in a measured using commercially-available
vertical axis orientation and fired directly thermal conductivity-type mass flow
into the inlet of the pumping system. meters with operating ranges of 0 to 10
SLPM. These units maintained steady
Operational parameters monitored during propellant mass flow rates by using flow
all tests included arc voltage and current, sensor feedback to control integral
hydrogen and nitrogen mass flow rates, solenoid-operated flow control valves.
propellant feed pressure, and temperature The mass flow controllers were calibrated
of the exterior of the anode housing. in-situ at each of the test facilities using a
Thrust was measured only during volumetric standard, as is common
performance characterization. All data practice in the flow measurement
were monitored, reduced, displayed, and industry. The individual gases were
stored using a microcomputer-based data flowed through the respective mass flow
acquisition and control system (DACS). controllers, then into a cylinder of known
Data were stored every 30 seconds volume. The pressure and temperature in
throughout the endurance tests. The the cylinder were measured before and
performance and endurance test data after flowing gas into the cylinder, and
acquisition programs were written using a the length of time gas flowed into the
graphically-driven software system which cylinder was recorded. The mass added
facilitated real-time reduction of all to the cylinder was then calculated
assuming ideal gas behavior. This
4
standard had an estimated uncertainty of
about 1%, with typical repeatabilities Test Procedure. The objective of this test
better than M. series was to identify the life-limiting
issues of arciet thruster operation at a
Thrust was measured by a calibrated specific impulse and power level of 550 s
displacement-type thrust stand designed and 2.0 kW, respectively. The approach
and fabricated at the NASA Lewis was to conduct a series of 100-h
Research Center. This unit was equipped endurance tests with intermittent thruster
with numerous water cooling passages performance characterizations and
throughout the structure to minimize physical examinations. Figure 6 shows
thermal drift due to component expansion the test chronology time line. Prior to
during thruster operation. Thermal drift assembly the electrode masses were
was observed to be less than 1% of the recorded. The constrictor dimensions
nominal thrust level during all tests, and were determined by making a mold of the
was found to be always in the direction of constrictor region using a vinyl-
decreasing thrust. Therefore, all thrust polysiloxanc analogue (dental putty), then
measurements were corrected by inspecting the mold under a microscope
recalibrating the thrust stand while hot, equipped with a three-axis translation
making any residual thrust measurement stage assembly. The length of the anode
error conservative. Additional details of protruding from the end of the anode
this thrust measurement system are housing was also recorded.
available elsewhere20.
Immediately following assembly the
The propellant feed pressure to the thruster was installed in the endurance
thruster was monitored using a strain- test facility for a 20-h burn-in period.
gage type pressure transducer with a Laboratory experience has demonstrated
range of 0 to 1.4 MPa. It is important to the tendency with a new cathode for the
note that the measured pressure is that of arc voltage to increase substantially over
the propellant upstream of the injector the first 20 to 30 h of operation at
disk, which was estimated to be constant mass flow rate and current. The
substantially higher than the arc chamber purpose of the burn-in period was to
pressure. Furthermore, the pressure drop bring the electrode to a condition more
across the injector disk was a function of representative of steady state prior to the
thruster operating conditions. initial performance characterization of the
Nonetheless, the feed pressure was thruster. During most of this phase of
useful as an indication of the integrity of testing the thruster was operated at a flow
the flow path and pressure vessel during rate of 5.Ox 10-5 kg/s and a power level of
operation. 1.5 W. The last several hours of burn-in
included operation at 1.75 and 2.0 kW,
The temperature of the anode housing also at a mass flow rate of 5.0x10-5 kg/s.
surface (see Fig. 1 for location) was
measured using a two-color optical Following burn-in, the thruster was
pyrometer with a range of 700 to installed on the thrust stand to
14000C. This device used the ratio of the characterize thruster performance. This
energy emitted at two wavelengths (both was done to provide a baseline for interim
in the vicinity of 1 p m) to calculate the and post-test comparison, and to aid in
selecting the mass flow rate for the
target surface temperature. The output
endurance test. As shown in Table I,
reading was not sensitive to the absolute
performance was measured at each of
value of the emittance, but to the slope of
the variation of emittance with three mass flow rates (3.Ox 10- 5 ,
wavelength. 4.0x10-5 , and 5.0x10-5 kg/s) at up to 5
power levels (1.0, 1.25, 1.5, 1.75, and
2.0 kW). The boundaries of the operating
5
envelope were defined by a PPU output were cancelled. The thruster was
voltage limit of 130 V and an anode disassembled a final time, after which the
housing temperature limit of 1400°C. electrodes were sectioned for
metalographic analyses.
Following the initial performance
characterization, the thruster was installed RESULTS AND DISCUSSION
in the endurance test facility and run for
100 h at a power level and specific Performance Characterizations. Table II
impulse of 2.0 kW and 550 s (nominal), summarizes performance data acquired
respectively. Due to the current-regulating during this test series. Figure 7 shows the
nature of the power supply, it was measured specific impulse versus specific
necessary to reduce the current setpoint power (i.e., the ratio of electric power to
occasionally during the endurance tests to mass flow rate) for the test thruster
maintain a constant 2.0 kW as the arc following the 20-h burn-in period. These
impedance increased with time. At the data (labelled "0 h" in Table 11) provided
end of the first 100 h, the thruster was a baseline against which to measure any
reinstalled on the thrust stand and changes in thruster performance during
characterized to document any the course of the endurance test series.
performance changes. After the 100-h This test also showed that the mass flow
performance check the thruster was rate necessary to obtain the specified
moved back to the endurance test facility endurance test condition of 550 s at a
where it was run another 100 h at 2.0 kW power level of 2.0 kW was 5.0x 10-5
and 550 s. Upon completion of the kg/s. Note that the specific impulse at a
second 100-h test segment, the thruster given specific power was dependent upon
performance was checked, following mass flow rate. This has been observed
which it was disassembled for inspection. previously in lower-power arcjets21. To
date there are insufficient data to
The third 100-h test segment was understand this phenomenon. Also
preceded by a performance shown in Fig. 7 are performance data
characterization. For reasons which will obtained in previous arcjet tests2.21,22.
be discussed later, graphite foil gaskets Agreement with the data of the current
were not used in the injector disk seals effort is generally good. The Ref. 2 data
during this test segment. An anomaly were acquired at two different test
during the first restart resulted in some facilities. The uppermost point was
constrictor damage. This necessitated measured in an industrial test cell at a
another disassembly and inspection, mass flow rate and power level of
during which the surfaces of the injector 4.4x10-5 kg/s and 2.0 kW, respectively,
disk were cleaned and the inconel spring using hydrazine propellant. The other
was stretched to increase the preload on Ref. 2 datum was measured at a
the injector seals. The performance
government facility at 2.8x10-5 kg/s and
characterization was then completed and a
1.0 kW on a 2:1 hydrogen-nitrogen
third 100-h test segment was conducted.
mixture. Although the Ref. 21 data were
obtained at a variety of mass flow rates,
After completing a total of 300 h at the
they were all measured at specific power
nominal operating conditions the thruster
levels below 15 MJ/kg. At that point the
was disassembled for inspection a second
effects of mass flow rate variations on
time. No performance characterization
specific impulse began to disappear. The
was performed prior to this inspection.
Ref. 22 data were measured at LeRC on a
The thruster was reassembled after the
hydrogen-nitrogen mixture at a mass flow
electrodes were examined so that the
rate of about 4.1x10-5 kg/s. Both points
performance could be measured.
are within approximately 3% of the
However, starting difficulties were
4.0x 10-5 kg/s data acquired during the
encountered, and the performance tests
current effort. Other performance data at
6
comparable conditions indicate good later in this section. The second and third
agreement as well18,23 unintended shutdowns occurred after
approximately 173 and 230 h of thruster
Figure 8 shows the results of the operation, respectively. In each case a
performance characterizations conducted facility interlock for the thruster power
during the test series at a flow rate of supply was violated, causing a loss of
5.Ox 10-5 kg/s. Figures 9 and 10 show the power to the PPU. The test was restarted
corresponding performance at flow rates within about 45 minutes of the second
shutdown, and immediately following the
of 4.Ox 10-5 kg/s and 3.Ox 10-5 kg/s,
third shutdown. Each time the thruster
respectively. No degradation of thruster
restarted without difficulty. Neither of
performance is indicated at any of the
these shutdowns was caused by any
operating conditions tested. This is
parricularly interesting in consideration of irregularity in thruster operation.
an anomaly experienced during the first
The arc voltage increased during the first
restart of the thruster after the 200-h
disassembly and inspection. An apparent 200 h of operation at an average rate of
35 mV/h. As shown in Fig. 11, the
leak around the injector disk resulted in
thruster experienced periods of increasing
substantial erosion of the constrictor.
and decreasing voltage, although all
This caused the arc chamber pressure at a
voltage values fall within a band of ±2.5
given mass flow rate and power level to
V from the nominal value at that point in
decrease, which lead to a decrease in arc
the test. This represents a variation of
impedance at all operating conditions.
Further details of this anomaly will be about 2.2% of the average arc voltage
given in a later section. The most measured during the first two 100-h test
segments. Arc voltage fluctuations of this
important result was that thruster
performance and stability were not magnitude have been observed in
impacted by the constrictor damage. previous endurance tests8,23. During the
third 100-h test the voltage rose at a
Endurance Tests, Figures 11, 12, and 13 substantially lower rate. This test segment
show the are voltage, propellant feed was also relatively free of the voltage
pressure, and anode housing temperature, excursions typical of the first two
respectively, at 30-second intervals over endurance tests. However, the voltage
the endurance test series. As discussed during this test was several volts lower
earlier, the test plan called for a series of than during the first two test segment due
100-h continuous endurance test to the constrictor damage sustained prior
segments with intermittent performance to this test segment.
characterizations and physical
examinations. Three unscheduled During most of the endurance tests the
shutdowns also occurred, one during propellant feed pressure remained steady.
each of the 100-h test segments. The first There were two instances, however,
occurred about 20 h into the first test, when the feed pressure declined gradually
caused by interruption of the two-color over periods of about 20 h and 50 h
pyrometer signal to the DACS. This during the early portions of the first and
violated the lower shut down setpoint for second test segments, respectively (see
this parameter, which in turn shut down Fig. 12). Either of these trends could
the propellant flow and PPU. The test have indicated a leak from the thruster
was restarted after about one hour pressure vessel, an increase in constrictor
without difficulty. None of the thruster diameter, or a leak past the injector disk
operating parameters was out of its into the arc chamber. The most serious
specified range prior to this shutdown, leak would be from the thruster pressure
although a declining feed pressure trace vessel because this would have been a
(see Fig. 12) was causing some concern. failure resulting in the loss of
This will be discussed in greater detail performance and, perhaps, the failure of
7
the thruster. Concurrent symptoms of a 630 kPa. Between 100 and 135 h the
pressure vessel leak would have been voltage fell from an initial value of 112 V
declining arc voltage and increasing to a minimum of about 110 V and the
anode housing temperature (due to an anode housing temperature rose to a
increase in the effective specific power). maximum of nearly 1400°C. These
The symptoms of an increase in trends suggested the possibility of a leak
constrictor diameter are similar to those of from the thruster pressure vessel or
a leaking pressure vessel: concurrent constrictor erosion. However, the voltage
decreases in arc voltage and feed began a strong recovery and the anode
pressure. Leakage past either of the housing temperature fell sharply at about
injector disk seals would have allowed 135 h, while the feed pressure was still
propellant to bypass the two injector falling. At about 150 h the feed pressure
ports. This would result in a reduction in and arc voltage increased in a step change
feed pressure, accompanied by irregular to 680 kPa and 115 V, respectively. The
arc voltage excursions, although no arc voltage, feed pressure, and anode
general trend toward decreasing arc housing temperature were relatively
voltage would be expected. The arc steady during the remainder of the second
voltage excursions could be due to rapid test segment.
movement of the arc seat about the
cathode tip in the absence of sufficient Thruster Inspections. As shown in Fig.
vortex stabilization in the arc chamber. 6, the thruster was disassembled for
inspection following the second 100-h
During the first 20 h of the first 100-h test endurance test. The overall thruster length
segment the feed pressure gradually was measured prior to disassembly. This
decreased from its original value of 680 was necessary for determination of the
kPa to about 640 kPa. Over the same arc gap later in the disassembly. The
period the anode housing temperature overall length was found to have
increased by about 30°C to 1330°C and increased by about 0.56 mm. The length
the arc voltage remained steady at 109 V. of portion of the anode protruding from
Following the first unintended shutdown, the anode housing was also measured.
the feed pressure and the anode housing The anode had moved within the tapered
temperature returned to their original seat in the anode housing as indicated by
values of 680 kPa and 1330°C , a 0.56 mm increase in the length of the
respectively. At that time it was protruding portion. Because the cathode
speculated that the front insulator might was tied rigidly to the rear of the thruster,
have been sticking to the cathode or movement of the anode within the
anode housing prior to the shutdown. housing would result in an increase in the
This would have defeated the spring in arc gap. The movement of the anode
the rear of the thruster, allowing the within the anode housing could have
compressive force on the injector seals to contributed to the pressure variations
be relieved, and resulting in leakage past observed during the first two test
the injector. The thermal cycle associated segments by allowing propellant to leak
with the shutdown could then have freed past the tapered seal or by relieving some
the insulator, allowing the thruster to of the compressive force on the injector
return to its original condition. No seals.
significant pressure variations were
encountered for the remainder of the first The arc gap was determined by
100-h test segment. measuring the overall thruster length with
the cathode pushed in until it met the
During the first half of the second 100-h anode, then subtracting this length from
test. The feed pressure gradually the overall length measured previously.
decreased from an initial value of The are gap was found to have increased
approximately 680 kPa to a minimum of by 0.013 mm, which was about 0.55 mm
8
less than was expected based on the cracks. Differences in the operating
increase in the length of the anode histories of these two cathodes include a
protruding from the housing. lower are current and anode temperature
(by at least 4000C), and substantially
The discrepancy between the measured greater number of starts on this cathode
arc gap and the forward movement of the than that of the current work.
anode within the housing was explained
when the cathode was removed from the Inspection of the graphite gasket between
thruster for inspection. The tip of the the insulator and the injector disk revealed
cathode had formed a crater nearly 1.3 some holes, although the actual sealing
mm in diameter with a bulged rim (see surface was intact. However, the gasket
Fig. 14). The material on this rim lay between the injector disk and the anode
outside the original conical envelope by had been completely consumed, leaving
an amount approximately equal to the no visible trace. This had been observed
distance moved by the anode. The axial in a previous life test conducted by the
distance from the original tip location to author during which the anode had
the crater rim was approximately 0.58 operated at temperatures some 500°C
mm. Figure 15 shows a sketch of the higher than in the current tests. As was
cathode tip with the original conical tip discussed earlier, some degradation of the
shape superimposed for comparison. Of injector sealing was believed to be
particular interest was the migration of responsible for the pressure fluctuations
material from the center of the cathode tip observed during the first two 100-h
to a region outside the original conical endurance tests. The lack of any remnants
shape of the cathode tip. This could have of the forward graphite foil gasket
resulted in shorting of the cathode to the confirmed those suspicions. Of particular
anode had the anode not moved forward interest was the fact that the thruster
within the housing. The measured resumed stable operation after each of the
cathode mass loss was approximately pressure excursions, suggesting that the
2x 10-6 kg. graphite foil gaskets were unnecessary
components.
Figure 5a of reference 24 shows a cross
section of a 2% thoriated tungsten The anode was not removed from the
cathode tip operated for 9 h in three housing at this time to avoid disturbing
thermal cycles at a current of 11 A. The the tapered seal between these two
overall diameter of the crater in that components. However, examination of
cathode tip was only about 0.5 mm the constrictor exit under a microscope
(versus 1.3 mm for the current revealed no substantial degradation.
specimen). However, the movement of Some metal deposits lined the nozzle wall
material to a region outside the original near the constrictor exit, but the
conical shape of the tip bears a striking constrictor was still circular and its
resemblance to the current test results. diameter had increased by only 0.015 mm
This suggests that the observed cathode (to 0.57 mm).
erosion was driven by some phenomenon
which was not time dependent. The thruster was reassembled and its
performance was characterized in
The tip of another cathode run for a total preparation for a third 100-h test
of 1000 h and 500 thermal cycles at a segment. The only deviation from
constant current of 11 A displayed a previous thruster assembly procedures
distinctly different type of erosion$. The was the deletion of graphite gaskets
results of that test showed a crater slightly around the injector; this in an effort to
smaller than that of the current test, but return the thruster to a condition as close
the rim of the cathode was not bulged. as possible to that prior to the
Instead, the crater rim had several axial disassembly. The arc gap was set to 0.60
9