QUALIFICATION 971
IEPC-93-106
FLIGHT OF AN 18-mN XENON ION THRUSTER' J.R. Beatie,** J.D. Williamst, and R.R. Robsontt Hughes Research Laboratories Malibu, California 90265 Jnk rlc mm = = = = = = = = = = = = = = = = neutralizer keeper current, A cathode flow rate, mA main flow rate, A neutralizer flow rate, mA power spectral density, g 2/Hz thruster input power, W vacuum chamber pressure, Pa accelerator voltage, V beam voltage, V cathode keeper voltage, V discharge voltage, V neutralizer coupling voltage, V neutralizer keeper voltage, V thruster electrical efficiency, % total propellant utilization efficiency, % thruster efficiency, %
Abstract In this paper, we describe an 18-mN xenon ion propulsion subsystem (XIPS). The XIPS consists of a 13-cm-diam thruster, a power processor, and a xenon storage and control unit. The thruster produces 17.8 mN of thrust at a specific impulse of 2585 s, with an input power of about 439 W. The power processor contains only about 400 parts in its 7 individual power modules: screen, accel, discharge, two keepers, and two heaters. The power processor is designed to operate from either a 29- to 34-V or a 49- to 53-V power bus and achieves an overall efficiency of 88 to 90% over these ranges. The xenon propellant is stored at an initial pressure of 7.6 MPa (1100 psia) to give a tankage fraction of only about 12%. We will flight qualify the XIPS thruster by subjecting two flight units to qualification testing, including performance, thermal-vacuum, and vibration tests. We will also subject the qualification thrusters to a cyclic life test where they will accumulate at least 12,000 h and over 6,000 ON/OFF cycles. Nomenclature BOL dB/oct EOL F g Hz Isp JA Jb Jc Jd JE beginning of life decibels per octave end of life thrust, mN root-mean-square acceleration in g's Hertz specific impulse, s accel electrode current, mA beam current, A cathode keeper current, A Sctde eer current, A = decel electrode current, mA = cathode emission current, A = = = = = = = = = =
hn
PSD PT Pa VA Vb Vck VD Vg Vn Te Tim
TT
* Work performed under company funding. ** Manager, Plasma Sources Department, Plasma Physics Laboratory. Senior Member AIAA. t Member of the Technical Staff, Plasma Sources
Department, Plasma Physics Laboratory. Member
1. Introduction In previous publications, 1-3 we described the development and testing of a 25-cm-diam XIPS thruster intended for North-South stationkeeping (NSSK) maneuvers on large spin-stabilized satellites. A more recent paper 4 describes the development of a 13-cm-diam thruster. The smaller thruster, which produces about 18 mN of thrust, exhibits near-optimum characteristics for performing NSSK maneuvers on three-axis-stabilized and smaller spin-stabilized communication satellites. In this paper, we describe the 18-mN XIPS, which consists of a thruster, a power processor, and a xenon storage and control system. We present performance characteristics of an engineering-model XIPS thruster, and we also describe the testing that two flight-model thrusters will be subjected to in order to qualify the XIPS for space applications. The qualification program includes performance, thermal-vacuum, and vibration tests, followed by a cyclic life test. In addition to the flight qualification work, we describe a long-term cathode life test, which will be conducted for an indefinite time in order to establish the ultimate lifetime of the XIPS cathodes. 2. Xenon Ion Propulsion Subsystem A schematic of the XIPS is shown in Fig. 1. This compact arrangement of a 13-cm-diam thruster, power processor, and propellant tankage and control unit represents one-half of a fully redundant propulsion system intended for use on small spin-stabilized satellites. For three-axis-stabilized satellites, a typical arrangement 1
t
AIAArtment, Plasma Physics Laboratory. Member S' Manager, Power Electronics and Controls Department, Plasma Physics Laboratory. Member AIAA.
I IEPC-93-106
9112-04-032
972
the application (slightly different circuitry is used for the all the timing and control required to start, stop, and operate the thruster, as well as to detect and implement corrective actions in the event of over-current conditions in the screen or accelerator power modules. The screen module produces the maximum output voltage of only 750 V to give xenon ions an exhaust velocity equivalent to about 3 390-s of specific impulse (uncorrected for propellant utilization efficiency and thrust loss due to beam divergence and doubly charged ions). The low output voltage of the XIPS power processor simplifies spacecraft integration and voltage isolation. We recently completed two working-model systems containing the essential features of the XIPS shown in 1. They consist of qualification-model thrusters, breadboard-model power supplies, and a flight-prototype pressure regulator. We present detailed discussions of the individual elements of the XIPS in the sections that follow. 2.1 Thruster Figure 2 shows a cutaway drawing of the 13-cm-diam thruster. The thruster, scaled from the 25-cm-diam XIPS design, was derived from the original ring-cusp configuration developed by Sovey. 5 The discharge chamber employs a central hollow cathode equipped with an enclosed-type keeper electrode. Xenon enters the discharge chamber through an annular plenum, which distributes the propellant uniformly throughout the ionization volume. Three rings of SmCo permanent 5 magnets located on the cylindrical side wall and circular end wall produce the magnetic field used to confine the discharge electrons. PROPELLANT ELECTRICAL ISOLATOR ISOLATOR PROPELLANT PLENUM MAGN C RETURN PATH 92-04-R
two ranges). The completely self-contained unit includes
PROPELL
TANK
PRO T
PRESSURE
REGULATOR
HONEYCOMB PALLET
POWER PROCESSOR
THRUSTER SUFig. Fig. 1. Xenon Ion Propulsion Subsystem (XIPS). consists of four thrusters and power supplies, and two propellant tanks. Table 1 presents a mass breakdown for this latter configuration. Table 1. XIPS Mass Estimates. Table - XPS Mass EstimaXIPS Unit Mass No. Per Unit (kg) Spacecraft Ion Thruster Power Processor Unit Xenon Tank Pressure Regulator Regulator Other Feed Components Gimbal Structure Structure Total Mass 5.0 6.8 2.0 0.8 3.5 2.2 2.8 2.8 4 4 4 2 2
Mass Per Spacecraft 20.0 27.2 4.0 1.6 3.5 3.5
2-a
67.9
NEUTRAUZER SUBASSEMBLY
With an input power of 439 W, the XIPS thruster produces 17.8 mN of thrust at a specific impulse of 2585 s, resulting in a thrust-to-power ratio of 40.6 mN/kW and a thruster efficiency of over 51%. We believe this to be the highest level of performance ever reported for an ion thruster of this size operated at a discharge voltage of only 28 to 30 V. The xenon propellant is stored initially as a high pressure gas, with a BOL density nearly twice that of water. The propellant feed system passively controls the flow of xenon into the thruster using a pressure regulatorMBLY to maintain a constant pressure on the upstream side of flow restrictors located in the propellant lines leading to the discharge chamber, its cathode, and the neutralizer cathode, The power processor requires an input power of about 500 W and operates over a spacecraft bus voltage range of either 29- to 34-V or 49- to 53-V, depending on 2
PERMANENT
MAGNETS ELECTRODE APE URES (3145) ELECTRICAL GROUND SCREEN CATHODE/ KEEPER MASK I-T IONEXTRACTION ELECTRODES (3) PERMANENT MAGNETS
M
Fig. 2. 13-cm-diameter XIPS thruster. A three-grid ion extraction assembly extracts, focuses, and vectors the 3145 individual beamlets that form the thrust beam. The discharge power supply maintains most of the interior of the discharge chamber at anode potential, which is 28- to 30-V positive of the cathode potential.
973 IEPC-93-106 The cathode, keeper, and screen electrode are the only surfaces that are not maintained at anode potential. The neutralizer assembly, also scaled from the 25-cm-diam XIPS thruster, provides electrons for neutralizing the positive ion beam. Experimental Performance Data Table 2 contains measured electrical characteristics of the XIPS thruster, and Table 3 summarizes calculated performance parameters. In addition, Table 4 lists the sensitivity of thruster performance to induced variations in critical power processor outputs and xenon flow rates. The data contained in Table 4 indicate that the sensitivities in thrust and specific impulse arem considerably less than 1%per percent change in operating conditions. Table 2. Nominal Operating Conditions of EngineeringModel XIPS Thruster. Operating Parameter Value Beam Jb, A 0.405 Vb, V 751 JA, mA 0.97 VA, V 299 Jd, mA 0.49 Discharge E, A 3.38 VD, V 30.0 Keeper Jck, A 0.87 Vck, V 18.6 Jnk, A 0.50 Vnk,V 16 22 Vg, V Propellant Flow Rate m m ,A 0.446 mc,mA 35.3 mn,mA 33.5 Tank Pressure
Ptk, Pa 1.3 x 10 .4
Table 4. Thruster PerformanceSensitvity. Performance Sensitivity. %/% Operating Parameter Powr Supply Output Power Supply Output Vb VA
E
Thrust +0.5 +.03 +0.7 +0 +0.7
Specific Impulse +0.5 0 +0.7 -.2 -0.4 312-1-0o26
Xenon Flow Rate
me 10 ,
-0.3
1
j(rl2)
-((1-co0S)) 2 A -4.60 A n - 0.923 -74.21
0.1
I I I 10 20 ANGLE. 0, deg. Fig. 3. Far-field Faraday probe data.
1.o
0.01
0
I 30
9312-19
Table 3. Performance of Engineering-Model XIPS Thruster. Operating Parameter F, mN
isp, s
z 0.8 -
o.
r,
---
"
Value 17.8
2585
PT, W Tm, % Oe. %
TIT__%
439 78.7 69.3
51.3
0.6
_j
o
z
0.4
Figure 3 shows far-field Faraday probe measurements made -26 beam diameters downstream of the XIPS thruster. We fit a point source model 6 to the data and used the resulting equation to generate Fig. 4, which shows the fraction of the total beam contained within a given half angle. Figure 4 demonstrates the high degree of beam collimation by showing that virtually all of the beam is contained within 240 and that 95% of the beam is contained within 180.
0.2
o
0
10
20
I 30
I
40
50
Fig. 4.
ANGLE. ,deg. Angular dependence of XIPS ion beam.
3
IEPC-93-106 Preliminary Thermal Vacuum Test Data Figure 5 contains a photograph of an engineering model XIPS thruster undergoing thermal-vacuum testing. It shows a stainless steel radiation shield that simulates adjacent satellite components by covering about 70% of the cylindrical side wall, and a quartz lamp that simulates solar radiation. 20969-11
974
To simulate a worst-case situation, the quarti lamp at two times 30%solar flux at wall ground screen structure illuminated the of the side GEO. In addition, radiation a a t t wo bl e s the so la r uxnnt G E I addton, radiation s sh e ld s b lo c k ed h e remaining 7 0 % o f th e ground screen i structure. The thruster temperatures equilibrate in approximately four hours from an initial temperature of -0C. This represents twice the maximum operating time in a typical NSSK application. Under transient conditions. we observed a 400 C margin on our 3000 C upper limit on the iron magnetic return structure at 1.9 h of operation. (Before initiating the transient measurement, the nonoperating thruster temperatures equilibrated under the 2-sun, 70%-blockage condition.) Figure 7 shows the measured facility temperatures corresponding to the equilibrium thermal test mentioned in the preceding paragraph. This defines the radiative heat transfer environment of the thruster.
9312-19-02R1
Fig. 5. Photograph of 13-cm XIPS Engineering-Model thruster under thermal-vacuum testing. Figure 6 shows the steady-state temperatures obtained with the engineering-model thruster operating at nominal conditions. The highest temperature measured on the iron magnetic circuit was 281*C; 19*C lower than our self-imposed upper limit of 300*C. (The manufacturer of the SmCo5 magnets suggests an upper limit of -325 0 C.) The stainless-steel anode liner exhibited a maximum temperature of 361C, but this surface is shielded from directly radiating to the side wall and end wall magnets by stainless steel magnet retainer structures.
ouARTZ LAMP
27.5'C RAD.
SHIELD
-133"C loN OUARz20. LAMP C
THRUSTER
-128C
8.5
2
Ia
26.C
-131C
-133C
-132C
Fig. 7.
Facility temperatures during thermal-vacuum testing of 13-cm XIPS Engineering-Model thruster.
9*112o0A
Erosion of XIPS Three-Grid Ion Optics System
SETAT 2-SUNS GROUND SCREEN
SET AT 2-SUNS
1S231C 314*c
91 * 256C 2 -2 3Sc
C
ADDITIONAL
70% BLOCKAGE
COVERI NG GROUND SCREEN
_
_ a51 c 3ae*C 361*C 2impingement C 14-c
Fig. 6.
'Thermal-vacuum test results under standard Thermal-vacuum test results under Fig. 6. da operating conditions and severe thermal environmental conditions. 4
Figure 8 presents the fraction of sputtered accel material escaping through the XIPS apertures decel as a function of the accel aperture diameter. It shows two curves corresponding to the 13- and 25-cm XIPS optics systems 7 obtained using configuration factors given by Howell. (The 13-cm XIPS accel grid is 2x thicker than the 25-cm XIPS grid.) The configuration factor remains relatively invariant over the expected life time range of accel aperture diameter. Using the configuration factors shown in Fig. 8 and experimental data 3, we estimate that the effective sputter yield of the impingement ions striking the accel grid is 0.193. We believe that the effective sputter yield is less than the yield for 300 eV xenon ions8 because (1) the ions probably do not strike the surface with normal incidence, (2) the charge-exchange ions are created at a location whose potential is between 0 and the -300 V potential of the accel grid, and (3) the actual sputtering process may exhibit some specular behavior which preferentially directs some of the sputtered atoms upstream. Assuming the same effective sputter yield applies to the 13-cm XIPS, we predict an overall mass
975
IEPC-93-106
9312-19-024
0.14
Z0 0.14
t 0.13
-j
K3
2.3 Xenon Storage and Control Unit Figure 10 contains a block diagram of the propellant tankage and control unit configured for a typical fourthruster arrangement on a three-axis-stabilized satellite.
9112-04-089R2 suJS + l soucTE
I I
m 0.12 < ACCEL GRID STHICKNESS z 0.11 0 0.508 mm 9 D 0.254 mm I. 0.
10
Ai.
D
1.1
1.3 1.4 ACCEL APERTURE DIAMETER, mm
1.2
1.5 *, *V
m
SQC
Toa S lMrt
RFF
Fig. 8.
Fraction of accel material escaping vs. accel aperture diameter for the 13- and 25-cm XIPS.
ma
V, w LM
-a
t
Ma r
A-L c
w am
NATER
----
loss from the ion optics system of 1.0 g over a period of 12,000 h. The actual mass lost from the accel grid would cause a typical accel aperture diameter to enlarge from its initial value of 1.14 to 1.30 mm. The above discussion supports three benefits directly related to the thicker accel grid of the 13-cm XIPS optics system. The first one concerns a lower view factor between an accel barrel and a decel aperture exit plane. The second benefit is lower accel aperture diameter dimension changes for a given amount of net material loss. The third benefit of a thicker accel grid is its larger self-viewing configuration factor. 2.2 Power processor A block diagram of the breadboard-model power processor is shown in Fig. 9. The power processor contains the seven power modules (screen, accel, discharge, two keepers, and two heaters) and simple control logic required to operate the thruster. It employs a dedicated series resonant inverter (SRI) as a dc-to-dc voltage converter for powering the screen module, which processes about 90% of the total power. The remaining six modules (which, with the exception of the accel, have current-regulated outputs) operate from the output of a second SRI that is used to convert the dc bus voltage into an ac current source. This scheme has the advantage that the current regulation is performed only once (in the ac current source), which leads to a significant simplification in the power processor design and reduction in its parts count. The measured efficiency of a 28-V-input breadboardmodel XIPS power processor is 88% at the nominal thruster operating point. Table 5 summarizes the individual power-module losses that correspond to this overall efficiency. We estimate an efficiency of about 90% with a 50-V input voltage.
M]W
A"
-,
RATI
--
<T
sor
Fig. 9. Block diagram of power processor. Table 5. Power processor Losses. Module Power Dissipation Screen 22.6 14.0 ac Inverter 2.3 Accel 5.7 Discharge Cathode Keeper 2.1 Cathode Heater (Off) 1.2 Neutralizer Keeper 2.1 Neutralizer Heater (Off) 1.2 Housekeeping 2.7 Internal Wiring 3.9 Output Wiring 0.6 Miscellaneous _9 59.3 W Total Efficiency at nominal power - 88% Xenon is stored at a moderate pressure of 7.6 MPa (1100 psia), regulated to a low pressure of 68.9 kPa (10 psia), and then expanded through flow restrictors that are sized to the flow rate requirements of the discharge plenum and the discharge and neutralizer cathodes. With an initial storage pressure of 7.6 MPa, the estimated tankage fraction is about 12%.
5
IEPC-93-106 Redundancy is provided for the pressure regulator, which is the critical element of the xenon storage and control system. We have used the flight-prototype pressure regulator for testing XIPS thrusters since 1985, accumulating well in excess of 5000 hours of operating time. In initial testing of the regulator, we determined that it regulated its output to 68.9 0.25 kPa (9.85 0.04 psia) over an inlet pressure range of 28.9 MPa to 689 kPa (4200 to 100 psia). The output pressure of the regulator has remained extremely well regulated throughout the eight years of accumulated testing.
9112-04-83R2
976
power, in addition to documenting their behavior over a wider range of power and xenon flow rates. Specific tests include discharge-chamber performance evaluation, ionoptics perveance and backstreaming measurements. sensitivity to power supply output and xenon flow rate variations, as well as beam envelope and thrust-loss measurements. This comprehensive performance mapping and characterization will define the nominal operating point for each thruster, including xenon flow rates and cathode emission current (the only independent performance variables).
20089-9-5
XENON
XENON
0 TO 2000 psia
0 TO 30 psia
Ppu
PPU
uI
PPU
PPIJ
Fig. 11. Qualification-Model thrusters. 3.2 Environmental Environmental testing will verify the capability of the XIPS thruster design to withstand the thermal environments anticipated in transfer and final orbit, and the vibration environment characteristic of launch by typical commercial launch vehicles. Details of the thermal-vacuum and vibration testing are presented in the
that follow.
ION THRUSTERS (4)
H LATCH VALVE FILLANDDRAN VALVE PRESSURETRANSCER
SPRESSURE TRANSDUCER
U]
REGULTOR, ITO 10
NORMALLYCLOSEDSOUS VALVE FILTER 2MICRON Ssections
a FLOW IMPEDANCE
POWER LINES
NOTE: GIMBALS NOT SHOWN
Thermal Vacuum Fig. 10. Block diagram of propellant storage and control a unit 3. Thruster Qualification Tests Qualification testing of the XIPS thruster will be accomplished in accordance with standard industry procedures for qualifying space hardware. The two qualification thrusters shown in Fig. 11 will be subjected to performance and environmental testing as part of the qualification. Later, the thrusters will be subjected to a cyclic life test Details of the qualification testing are presented in the sections that follow. 3.1 Performance The qualification plan will subject the thrusters to a comprehensive performance evaluation using lab-type power supplies. This testing validates the performance capability of the thrusters operating at their nominal 6 Thermal-vacuum testing will subject the individual qualification thrusters to a 1-h cold soak at -53 0 C and then a total of three temperature cycles between the cold and hot environmental limits of -51*C and +65 0 C, respectively. The rate-of-change of temperature during the transitions will be maintained at greater than lC/min. The thrusters will be OFF during the transition and will undergo a 12-h "soak" at the temperature extremes. The thrusters will then be operated for a nominal 2-h period after the first cold soak and the first hot soak. Thruster performance variables will be recorded throughout the ON period. In addition, each thruster will be subjected to twenty temperature cycles wi...