WEAR CHARACTERISTICS FROM THE EXTENDED LIFE TEST OF THE DS1 FLIGHT SPARE ION THRUSTER

Anita Sengupta, John R. Brophy, Keith D. Goodfellow

Jet Propulsion Laboratory California Institute of Technology ,4800 Oak Grove Drive, MS 125-109, Pasadena, CA, 91109, US

Anita.Sengupta@jpl.nasa.gov A~STRACT

An on-going life test of the Deep Space One flight spare ion thruster, is being conducted at the Jet Propulsion laboratory. The thruster has operated for over 27,290 hours and processed in excess of 219 kg of xenon propellant. To date several potential failure modes and wear processes have been observed and identified. Severe discharge cathode keeper erosion was first observed during and after operation at the 1.5 kW power point. Keeper erosion has continued at a steady rate, at the 2.3 k W power point, fully exposing the cathode heater and orifice plate. At the full power point, significant erosion of the accelerator grid apertures and grid webbing, has occurred, with the rate of hole enlargement increasing during the last full power test segment. At present, the thruster can no longer be operated at the full power point due the resultant increase of the electron backstreaming limit beyond the voltage capability of the flight accelerator grid supply. During operation at the minimum power point, there was a significant increase in neutralizer keeper voltage. Video data revealed the presence of deposits within the neutralizer orifice, resulting in a loss of neutralizer (minimum flow rate) margin, from plume-spot mode transition. Subsequent operation at the full power point, removed the deposits, and neutralizer operation has returned to nominal levels. Another notable erosion process is the degradation of electrical isolation between neutralizer keeper and common, and the both components to ground. Thruster performance has not degraded after 27,290 of operation, with the levels of thrust and specific impulse relatively unchanged from the beginning of test values. The thruster currently continues to perform nominally, and is running at 1kW in part to investigate neutralizer performance at the low end of the throttle range.

INTRODUCTION

The Deep Space One (DS1) mission was launched in October of 1998, on a mission to the Asteroid Braille and Comet Borelly. DS1 was a technology validation mission, flying a 30-cm-diameter Xenon Ion Engine as its primary propulsion system. The ion thruster successfully completed the mission in December of 2001, processing a total of 82 kg of propellant and accumulating 16,265 hours of operation in space. Although the mission was a success, future NASA science missions utilizing solar electric propulsion will require lifetimes and propellant throughput in excess of 20,000 hours and 200 kg. As a result, assessing the ultimate service life capability of the technology is vital, requiring extensive ground testing and data analysis. Details on the mission performance can be found in references [ 1-31.

Two flight unit thrusters were fabricated for the DS 1 mission. FT1 was mounted on the spacecraft, and l T 2 was designated the flight spare. The flight spare ion thruster has been the subject of an extended life test, at the Jet Propulsion Laboratory, since the fall of 1998. The thruster was started just prior to the launch of DSl, and has been under vacuum, to the present day. Thruster performance data has been collected and analyzed over the past 3.5 years, to determine and characterize potential failure modes, wear mechanisms, and performance degradation over time.

TEST PLAN

The objectives of the extended life test of FT2 are to identify and characterize unknown failure modes, determine how engine performance changes with operating time, and determine the ultimate service life capability of the NASA 3Ocm-diameter ion thruster technology. The initial test objective, to demonstrate 150% of the DS1 mission throughput capability, or the processing of 125 kg of Xenon, was accomplished in December of 2000. As this level of throughput was accomplished with no signs of engine performance degradation or significant

1

wear, the test was continued to demonstrate higher throughput capability needed for future NASA science missions.

Table 1. NSTAR Thruster Throttle Table The thruster is throttle-able to maximize use of available solar array power in space. Table 1 is the throttle table for the ground test, where the designation TH is given for each operating point, with a power range of 0.5 kW (THO) to 2.3 kW (TH15). The beam current and voltage are controlled to provide different fixed levels of thrust and specific impulse. For each throttle point, flow rates are set to account for tank ingestion, and to maintain a propellant efficiency of approximately 90%. F r 2 has been operated at several throttle points for approximately 5000 to 6000 hour intervals, to better understand wear as a functioa of power level. TG date the thruster has been operated at TH15 for approximately 14,000 hours, TH8 for 5500 hours, THO for 5700 hours, and has been running at 1kW (TH5) for the past 1500 hours. There has been an emphasis on full power operation to maximize propellant throughput, and as accelerator grid wear is most significant at this level. At 25700 hours, full power operation was terminated, as the electron backstreaming limit exceeded the capability of the flight PPU. Operation at TH5, 1 kW, was chosen to provide neutralizer performance data at the lower power levels, and to obtain operational experience for the upcoming Dawn Mission, which has a large portion of its mission profile at the 1kW operating point.

Performance measurements are taken every 100 to 200 hours, measuring thrust, doubles and singles current, the electron backstreaming and perveance limit, and screen grid transparency to ions. The test is fully automated and computer controlled, with thruster electrical parameters and facility data recorded every 5 seconds, and thrust vector data recorded every 300 seconds. Throttle tests, neutralizer characterizations, and sensitivity tests are taken every 2000 to 3000 hours, to investigate engine performance and wear over the full throttle range. Video and photographic data of the neutralizer and discharge cathode assemblies, and downstream face of the accelerator grid are taken every 1000 to 2000 hours, to monitor and quantify erosion processes.

THRUSTER DESIGN

The flights engines were fabricated by Hughes Electron Dynamics. The 30-cm discharge chamber employs a three-ring cusp magnetic field design, and a hollow cathode as the electron source. A two-grid molybdenum optics system focuses and electro-statically accelerates the ionized Xenon propellant, to produce thrust. The neutralizer hollow cathode assembly supplies electrons to charge neutralize the ion beam. Details on the 30-cm thruster operation can be found in reference [4].

TEST FACILITY

The test is conducted in a 3-m by 10-m-long vacuum chamber with a total xenon system pumping speed of 100 kL/s [12]. The vacuum system provides a base pressure of less than 4x10-6 Torr at the full power flow rates. The pumping surfaces are regenerated after accumulation of 10 kg Xenon, during which time; the engine is exposed to a mostly Xenon background pressure of 1 Torr. The vacuum chamber is lined with graphite panels to reduce the amount of material that is back sputtered onto the engine and test diagnostics. A quartz crystal microbalance (QCM) is located next to the engine, and provides real time measurements of back sputtered material in the plane of the grids. The propellant feed system consists of two mass flow meters for each of the cathode,

2

neutralizer, and main lines. The downstream meters are used to measure the flow, and the upstream meters as flow controllers. Laboratory power supplies, with similar capabilities to the flight PPU, are used to run the thruster. The ground electrical design references facility ground, and the reference potential is neutralizer common. A computer data acquisition system is used to monitor the engine and test facility. It records and stores test data, and is programmed to shut down the thruster in the event of a facility problem or out of tolerance engine parameter. Details of the flow system and electrical system can be found in reference [5].

TEST DIAGNOSTICS

Several diagnostics are used to measure the ion beam characteristics as well as general engine performance parameters. A modified version of the GRC inverted pendulum thrust stand is used to make direct thrust measurements.2 A thrust vector probe, consisting of a series of current collecting graphite rods, located downstream of the thruster, provides current density measurements across the beam. An EXB probe, also located downstream of the thruster, measures the double and single ion beam current. A faraday probe that sweeps over the diameter of the engine is used to measure the radial beam current density profile. Two cameras mounted on a three-axis positioning system in side the tank, allow detailed photography and video recording of the discharge cathode, neutralizer, and downstream surface of the accelerator grid A laser profilometer is also mounted on the positioning system, for detailed measurements of accelerator grid erosion. Specific details on the operation and design of the diagnostics can be found in references [5-81.

EXTENDED LIFE TEST RESULTS

Accelerator Grid Erosion

Towards Outer Edge -

Fig. 1 Accelerator grid wear (a) at 25542 hrs over radial section, (b) higher resolution of grid aperture at 21306 hrs

Video and images of the downstream face of the accelerator grid are taken at regular intervals over the course of the test. Prior to 23,770 hours, measurements of accelerator grid aperture diameter were taken at 4 different radial locations, using the camera-positioning system inside the chamber. At 23,770 hours, the z-axis of the positioning stage failed, preventing detailed focusing on the downstream surface of the accelerator grid. Detailed images of accelerator grid erosion are now obtained with a high-powered telephoto lens camera, focusing through a tank port window. From the image data and previous testing, it can be seen that accelerator grid erosion is characterized by enlargement of the grid aperture or barrel section, and the formation of pits and grooves in the grid webbing surrounding each hole.3 Figure l(b) is an image of a hole in the center of the accelerator grid at 21306 hours, taken normal to the grid plain. Pits are located between every three adjacent apertures, and grooves between each two neighboring apertures. Figure l(a) shows a larger section of the grid from the carbon deposition region (outer edge) towards the center of the grid at 25542 hours. The pit and groove erosion pattern surrounding each aperture can also be discerned in these images, showing a different depth perspective as the images are not normal to the grid plane. In figure l(a), the region towards ~~- the outer edge of the . ed efibits d n i m a f p h aGd-koove erosion, Gd-appeGs darker incolor due to re deposition of carbon from the tank walls Beyond this region, a hexagonal erosion pattern is discerned in the grid webbing region. The

3

1.50 erosion pattern around each aperture becomes circular, towards the left E 1.45

hand side of the image, corresponding 1.40

to the center of the grid (beam). After 25,542 hours of operation, only a thin shell of material separates the aperture 1.30

wall from the pits and grooves in the 1,25 grid webbing. Operation when these

areas have intersected is not desired, g 1.20

as the geometry of the barrel would I l.,s

change dramatically, potentially

- 1.35

e,

- -

compromising the state of the grids, 1.10

: TH15 : THE : TH15 : THO : TH15 : TH5

0 :

A

0 Center Hole G2.9 cm radius A5.7 cm radius 08.6 cm radius

" 4000 8000 12000 16000 20000 24000 28000 M P T Run Hour for erosion measurements that will be

taken at the conclusion of the test. Fig. 2 Accelerator grid aperture erosion versus run hour During the third TH15 test segment, the electron backstreaming limit increased at a rate higher than the previous two full power segments. It is likely that the rate of aperture enlargement also increased, resulting in the profile in figure l(a). Operation at TH5 has resulted in no measurable enlargement of the apertures or increased erosion of the webbing, over the past 1500 hours of operation.

Figure 2 is a plot of barrel erosion at four different radial locations on the grid, for the first 21,306 hours of operation. Each vertical line marks the beginning of a new test segment. The data indicates that barrel erosion is most significant at TH15, and negligible at THO. TH15 is associated with the highest beam current density and total accelerating voltage, and as a result the most significant erosion. The data also indicates the most significant barrel erosion occurs towards the center of the grid. Faraday beam profiles indicate the beam profile is peaked, with maximum current density along the centerline of the thruster, -240

therefore leading to higher erosion rates at p 200 the center of the grid.

Charge exchange ions are likely 5 160 $ responsible for the erosion on the s 120

downstream face of the accelerator grid t and the observed barrel enlargement. Y 80

Computational results indicate the sites of 3 $ 4 0 charge exchange birth are both between

- A

the screen and accelerator grid apertures 0

? : 0 Fr2 Cathode Orifice Area .Initial Cathode O r i i i Area ' -_ 0 .

and downstream of the accelerator grid, 0 4000 B o o 0 12000 16000 Zoo00 24000 28000 MPT Run Hour leading to the accelerator aperture

enlargement and pits and groove erosion, Fig. 3 FT2 Discharge cathode keeper orifice area

respectively. l1 In addition, the higher neutral density during ground testing, may lead to a larger production of charge exchange ions further aggravating barrel erosion.

Discharge Cathode Keeper Erosion Discharge cathode keeper erosion is a potential failure mode for ion thrusters, as demonstrated by the previous 8,200-hr test of an engineering model thruster and the current test of 171-2 .~~~ After 27,300 hours of operation, the keeper on FT2 has

0 Hours 26675 Hours

Fig. 4 FT2 Discharge cathode keeper erosion

4

eroded to fully expose the cathode heater and cathode orifice plate. Keeper erosion was first observed during the TH8 test segment, following a short between cathode keeper to cathode common. Cathode keeper erosion has continued at the subsequent full power segments, but at a significantly lower rate during THO operation. Figure 3 is a plot of cathode keeper inner area versus run time. The last TH15 segment exhibited a similar rate of increase in diameter as the previous full power segment, suggesting the erosion mechanism is power level dependent. Figure 4 is a comparison of the beginning of life condition of the cathode keeper, and the most recent image at 26675 hours. The cathode heater and orifice plate do not appear to exhibit surface erosion or damage, in spite of 15,000 hours of operation since the onset of keeper erosion. The heater power, however, began to degrade following the short between cathode common to keeper. This is critical to cathode life, as sufficient heater power is required to initiate electron emission of the cathode insert. To date the heater power has decreased -lo%, from the beginning of life value. A possible explanation for the heater power reduction is radiant heat lost to the environment as the orifice enlarges. The cause and throttle level dependence of cathode keeper erosion is not yet well understood One possible explanation for the onset of keeper erosion at the TH8 power point is the increase in production of multiply charged ions, due to a proportionately low cathode flow rate, as compared to THO and TH15. For a fixed space charge current limit from the cathode, lowering the flow will result in an increase in the charge of the ions. Multiply charged ions may have sufficient energy to sputter erode the keeper material, resulting in the erosion ongoing since 5850 hours. As the keeper orifice enlarged, neutral density in the orifice reduced, potentially increasing the production of multiply charged ions, for a fixed current level. This may explain why keeper erosion continued at the subsequent TH15 test segments, when none was observed during the fnst 5000 hours of operation at TH15.

30000

In spite of the severe keeper erosion and degradation in heater performance, the cathode has experienced no noticeable change in its ignition characteristics. This suggests that sensitive cathode heater components, the cathode orifice, and the cathode insert have not been compromised. As a result, keeper erosion is not expected to cathode failure in the near term, which sets a new precedent for the usable life of the hollow cathode technology. Up to now, the life test of the Space Station plasma contactor hollow cathode has served as the benchmark of hollow cathode technology usable lifetime. The operation of the plasma contactor cathode sharply changed after only 23,776 hours of operation, requiring a 400 V increase in the start voltage to ignite the cathode,. The cathode end-of-life was officially established at 27,800 hours, when the cathode failed to ignite with an applied voltage of 1050 V.13 The FT2 cathode has run in excess of 27,290 hours, with a start voltage of 250 V. Performance of the cathode has not degraded after 292 starts, suggesting that the usable life of hollow cathode technology is far greater than 27,000 hours. The improved F"2 cathode lifetime is likely due to differences in the operating temperatures, of the two cathodes. F"2, with a larger orifice, has operated at a lower tip temperature than that of

temperature difference is likely to have accelerated barium depletion of the the plasma contactor cathode. The space station cathode insert.

Ion Optics Wear In addition to the visual measurements of the downstream face of the accelerator grid, wear of the ion optics system is characterized from perveance, electron backstreaming, and screen grid transparency measurements taken every 100 to 200 hours. These measurements indicate the ability of the system to extract, focus, and accelerate ionized propellant, to produce thrust.

The perveance limit provides

-21 0

I I

-160

-110 I I

-60

t, B :/-

4; 5600 10060 15060 20000; 250& -10 8

I . 1 100 Runhour

Fig. 5 Electron Backstreaming and Perveance Limit versus run hour a

measure of how defocusedthe beam can be before direct ion impingement on the accelerator grid occurs. Direct 5

ion impingement, with ions energies equivalent to the total accelerating voltage, can lead to rapid accelerator grid wear and eventual structural failure. The perveance limit is also an indicator of changes in the intra-grid electric field, and possible disturbances in the neutralization plane. For the ground tests, the perveance limit is defined as the screen grid voltage where a 1 V reduction in screen voltage, results in a 0.2 mA increase in accelerator grid impingement current. Perveance data is shown in Fig. 5 versus run hour, where vertical lines delineate each test segment. The perveance limit decreases with accelerator grid aperture enlargement, as the beamlets must become increasingly defocused to impinge on the surface of the grid. As a result, the rate of perveance limit is highest during full power operation. During the test, the rate of change has been close to linear, with the exception of the first 1000 hours, shifts at 13933 and 25566 hours, and during the last 2000 hours of the third TH15 segment. At the beginning of the test, it is likely that wear of manufacturing cusps accounted for the rapid change in perveance limit, and subsequent leveling off period. The shifts in perveance limit are not fully understood, although each discontinuity followed an extended shutdown period, where the thruster was exposed to a 1 Torr background pressure for several days. It is possible that during this exposure, adsorption of oxygen and water vapor present in the chamber, may have reacted with carbon deposits on the grids, resulting in a change in the grid material thermal emissivity. An increase in the hot grid-gap, would lower the intra-grid electric field, and could account for the measured shift in the perveance. The increase during the last TH15 segment is not fully understood, although it is related in part to an increase in the rate of accelerator barrel erosion. At TH5, the perveance limit has remained constant over the past 1500 hours of operation. This indicates that accelerator grid barrel erosion is also minimal at 1kW.

The electron backstreaming limit versus run hour is also shown in Fig. 5. Electron backstreaming occurs when the potential at the center of the accelerator grid apertures is not sufficiently negative to prevent beam- neutralizing electrons from streaming into the discharge chamber. Electrons backstreaming into the discharge chamber are indistinguishable from ions accelerated out of the thruster, to the beam power supply. Therefore at a fixed beam current, electron backstreaming results in a decrease in discharge current, reducing ionization, thereby significantly decreasing thruster perfonnance. To prevent this, the accelerator grid is biased to a sufficiently negative voltage to prevent electrons from getting into to the discharge chamber. For the ELT, the backstreaming limit is determined by reducing the accelerator grid voltage until a 1% reduction in discharge loss occurs. Discharge loss is the ratio of the energy to produce beams ions to the extracted beam current. Results indicate the backstreaming limit is most negative at TH15, and least negative at THO. As the thruster is throttled to higher power levels, the beam current and voltage are increased, along with the positive space charge. As a result the aperture potential increases, requiring a more negative accelerator grid voltage to prevent electron backstreaming. Similarly, as the accelerator grid holes enlarge, the applied voltage required to prevent backstreaming becomes more negative. Therefore, the rate of increase of the electron backstreaming limit is most significant at TH15, and negligible at THO. The rate of increase of backstreaming limit was relatively linear for the first 23,000 hours of operation at each of the test segments, however, 2000 hours into the last TH15 segment, the backstreaming limit began to rise rapidly. Initially it was thought that the accelerator grid barrel sections had begun to intersect the pits and grooves in the grid webbing, resulting in a change in material erosion rate. However, detailed imaging and telescopic inspection of the downstream face of the accelerator grid does not indicate that operation in this regime has occurred. Another possibility is rogue hole formation in the grid webbing, although this is not apparent in the most recent images. At present the electron backstreaming limit at TH15 has reached the upper range of the accelerator grid power supply. Although the thruster is fully operational below TH15, the thruster can no longer be operated at the full power point, demonstrating that accelerator grid wear is the critical life limiting mechanism at full power operation. Significant shifts in the electron backstreaming limit on the order of 5 to 10 V occurred at 13,993 hours and 25548 hours, respectively. As mentioned previously, these shifts may be due to changes in the hot grid-gap, as a result of extended exposure to a 1 Ton background pressure. It is most probable that the shifts are a result of changes to the intra-grid electric field, as both the backstreaming and perveance data exhibit similar discontinuities. The backstreaming limit at TH5 has not changed over the past 1500 hours of operation. This also suggests that aperture enlargement is minimal at 1kW.

6

Discharge Chamber Performance Discharge voltage, discharge current, and thruster efficiencies are monitored to track discharge chamber performance and wear as a function of runtime and power level. In addition, measurements of doubles and singles current are taken every 100 to 200 hours, to prevent operating in a regime that can lead to accelerated discharge chamber wear.

21 0 DischargeCurrent - Figure 6 is a plot of discharge current

30 7 0 I 18

- 22 { / - Dikharge Voltage 7 1,

: n o s : m : n o s : m : m I s : m

current with run time at the full power point, most noticeable over the first two TH15 segments. As the accelerator apertures enlarge, more neutral xenon is lost from the discharge chamber. As beam current is proportional to neutral density, the discharge current would have to increase to maintain the fmed beam current. This may be a factor in discharge current variations observed in FT2. The TH8 segment exhibited a net loss in discharge current after 5509 hours of operation. Although the cause is unknown, the period of

'?

discharge current decrease coincided 0 5000 10000 15000 20000 25000 30000 with a short that developed between cathode keeper and cathode common, and a marked increase in the discharge voltage. Virtually no change in discharge current occurred over the 5663-hour THO segment. This may correspond to the absence of accelerator grid enlargement at this power level. The most recent TH15 segment was relatively constant, until 24,000 hours, when the discharge current began to decrease. This is unusual as this test segment experienced a rapid change in electron backstreaming, characteristic of accelerator aperture enlargement. It was first thought the thruster was operating whilst electron backstreaming, resulting in the lowered discharge current. However, extensive electron backstreaming testing and thrust measurements did not reveal this to be the case. The decrease in discharge current is on the order of a tenth of an ampere, but the reduction does suggest improved ionization efficiency, in spite potentially increased neutral loss due to aperture enlargement. Operation at TH15 was terminated at this point, and the thruster was throttled to TH5 operation. Operation at TH5 has also lead to relatively constant discharge current over the past 1500 hours.

Runhour Fig. 7 Ratio of doubles to single current versus run hour

The nominal discharge voltage range for FT2 is 23 and 27 Volts, and is highly sensitive to the main and cathode flow rate for each throttle point. Discharge voltage is a critical parameter, as high voltages can lead to increased production of multiply charged ions. FT2 is operated to maintain sufficiently low voltages to limit double ion production, whilst maintaining high propellant utilization efficiency. Figure 6 indicates that TH8 and TH5 segments have the highest discharge voltages, as compared to TH15. The THO segment indicates a significant drop in discharge voltage at 19,000 hours. This was the result of a flow calibration error, and the flow returned to the nominal level at 19,500 hours. The most recent TH15 segment exhibited no change in discharge voltage as

7

compared to the previous full power segments. The current TH5 test segment has also exhibited a relatively constant discharge voltage.

The double-to-single ion current ratio is an important parameter directly related thruster performance and wear. Doubly charged ions are more energetic and hence can sputter erode discharge chamber surfaces. Severe sputter erosion of the discharge chamber surfaces, cathode, and flake formation can lead to thruster failure. Additionally, it takes almost twice the energy to create a doubly charged Xenon ion, reducing power efficiency. Double ion current also reduces overall thrust, further impacting thruster performance. The double-to-single ion current ratio is shown in Fig. 7. During the first test segment, at TH15, the doubles ratio peaked at approximately 2500 hours, and then returned to the level at the beginning of the test segment. During TH8 operation, the double-to-single ion ratio exhibited an increase on the order of one percent. During the second TH15 test segment, the doubles ratio increased; roughly two percent. During the THO test segment, the doubles ratio remained relatively constant. During the last TH15 test segment, the doubles ratio increased less than one percent, the smallest full power change. At the TH5 operating point, the double ratio appears to be increasing at a comparatively high rate, similar to that seen at the beginning of the first TH15 segment, although longer operation time is required to confum the power level trend, as it may just be noise in the data. Comparison of operation at the THO, 5, 8, and 15 operating points, indicates a direct power level dependence on doubles to singles content.

171. , 3

16.5 - : 16 - :

15.5 - : 15 - ;

T H l 5 : m

Vnk Jnk . . ... ... ... .

Neutralizer Performance The neutralizer cathode is the

electron current to prevent spacecraft charging. Similar to the discharge cathode, the neutralizer is a hollow cathode, with neutral xenon gas flowing though it, inside a cylindrical keeper electrode (anode). The keeper power supply maintains the neutralizer current at the fixed level specified in the throttle table. The neutralizer keeper voltage is dependent on the flow rate of xenon through the cathode, the condition of the orifice, and the keeper current. The neutralizer flow rate is set to prevent plume mode operation, whilst minimizing cold flow losses, to improve thruster efficiency. The neutralizer keeper voltage and current are

source of beam neutralizing 0

1 - 1.5 - 1

. 0.5 5000 loo00 15000 2oooO 25000 30000

Runhour Fig. 8 Neutralizer keeper voltage and current

15617 Hours 2 1306 Hours

Fig. 9 Fr2 Neutralizer keeper orifice

shown in Fig. 8. The spikes in the neutralizer keeper voltage seen repeatedly over the 27,290 hours of operation correspond to cathode conditioning events following a cryopump regeneration, or exposure to background pressure in excess of 1E-4 Torr. Cathode conditioning is a process where cathode heater current is applied at various levels to burn off surface contaminants that may deposit/adsorb on the cathode surface during exposure to the higher tank pressure. Following a conditioning the voltage is temporarily higher, but returns to the nominal level with runtime. Visual inspections of the keeper surface reveal minor pitting, but no significant damage or erosion of the keeper surface. Neutralizer keeper performance characteristics have been stable at the

8

TH15 and TH8 test segments. An interesting trend was observed, however, during operation at THO. The keeper voltage began to steadily increase 2000 hours into the test segment,. Photographic inspection revealed the formation of deposits. within the orifice, significantly reducing its effective area. Figure 9 compares the condition of the orifice at the beginning and end of the THO operation. By the end of the THO segment, at 21,306 hours of operation, the measured area of the orifice is 40% of the initial size. This phenomenon is the most probable cause of the increasing neutralizer keeper voltage and the loss in flow rate margin from plume mode operation. Nominal operation of the neutralizer is called spot mode, where the voltage oscillations of the keeper are significantly less than 5 V peak-to-peak. Plume mode occurs when the cathode sheath extends to the anode, resulting in large voltage oscillations, an increase in neutralizer keeper voltage, and the production of energetic ions. These ions have sufficient energy to erode neutralizer surfaces, reducing the lifetime and performance of the neutralizer cathode. Visually, it appears as an expanded, slightly flickering plume from the neutralizer. Operating at sufficient flow rate and keeper current level for the specific throttle point can prevent plume mode operation. As such, flow rate margin from this regime is routinely checked every 2000 hours of operation, and peak to peak keeper voltage is monitored continuously on an oscilloscope.. By the end of the THO segment, the flow rate margin was close less than 0.2 sccm. It is believed that the deposits within the orifice are both responsible for the increase in DC and peak-to-peak increases in keeper voltage. Although the cause is unknown at this time, the phenomenon was only observed at THO, and not at TH8 or TH15. Following operation at TH15 for 5000 hours, the THO keeper and flow rate margin from plume mode returned to nominal levels. This suggests that operation at the higher power levels removed the deposits, and therefore the deposition mechanism may be power level dependent. Additionally, a similar phenomena was observed during the DS1 mission. The flight neutralizer exhibited signs of plume mode operation following extended operation at THO. Several on orbit and ground tests were performed to characterize the effect of varying neutralizer flow and keeper current on energetic ion production from the neutralizer. They revealed that increasing flow or current reduced energetic ion production and increased margin from plume mode. Based on these results the neutralizer keeper current was increased for TH5 operation, to characterize the effect on neutralizer power degradation at the lower power levels.

Thrust Magnitude and Direction Fig. 10 is a comparison of the thrust calculated from electrical thruster

made during FT2 testing. As the beam current and voltage are held constant, the actual thrust is held '

constant, to the first order. The thrust at TH15 is 92 mN and at THO 20 mN. Variations in thrust can then be attributed to increased beam 20 divergence, a higher double ion content in the beam. or oDeratine

loo

parameters and thrust measurements 80

6o I 40 R

Dm Measured :

I Calculated :

21, 25 0 whilst electron backstreaming. The thrust. has remained relativelv 0 5000 10000 15000 20000 25000 30000

constant for each test segment.. The variation between measured and calculated thrust has been less than 3%, with the exception of operation at the end of the TH15 test segment. The thrust appeared to drop off, but it was determined to be an equipment error, and was corrected. The results indicate that the ion thruster has experience no degradation is thrust after 27,290 hours of operation.

Runhour Fig 10 Measured to Calculate Thrust Comparison

The thrust vector is also monitored during the test, as non-axial thrust must be accounted for, and corrected by mission operations. For the first 2000 hours of operation, the thrust vector experienced a variation of half a degree in the horizontal plane. Following this burn in period, the thruster experienced minimal variation in thrust vector location over each individual test segment. Test data also indicates that the thrust vector becomes more

9

off axis, as the power level is increased and as the thruster wears. The variations experienced by FT2, however, are well within the range of a typical spacecraft gimbal system. 18

Electrical Impedance Deaadation Loss of electrical isolation between critical components is another generic failure mode for ion thrusters. At the start of FT2 testing, all impedances were above 1 GQ. However, degradation in the electrical isolation for neutralizer and discharge cathode components was first observed only 447 hours into FT2 testing. Fig. 11 is a plot of neutralizer keeper and neutralizer common to facility ground impedances. The impedance of both fell during the first 2000 hours of operation. The impedance for neutralizer keeper to ground dropped to from 10 GQ to approximately 30-40 kW and has decreased to 26 MQ, over the past 25,000 hours of operation. The impedance loss for neutralizer common to ground, decreased at a lower initial rate, and appears to be power level dependent, with each full power test segment resulting in the highest rate of degradation. Currently the impedance is on the order of 2 MB.

neutralizer keeper to neutralizer common and cathode common to

Fig. 11 also shows degradation of

anode impedance. The neutralizer keeper to common impedance fell most significantly during the first 2000 hours, and leveled off at approximately 2 MQ by 15617 hours. Over the past 12,000 hours of operation the impedance has increased slightly to 2.9 MB. Thruster failure will occur if the discharge cannot be maintained due to high leakage current. The cathode common to anode impedance also fell rapidly during the first 2000 hours, reaching a minimum of 1 MQ at 5000 hours. From 5000 hours on, the impedance has varied widely between 3 and 30 MQ, and is currently 20MQ. Degradation of cathode common-anode impedance can reduce discharge chamber performance, and if severe enough, can cause thruster failure.

45

4o

0 5000 10000 15000 20000 25000 30000 Runhour

Fig. 11 FT2 Neutralizer and Cathode Impedances

CONCLUSIONS

To date, over 27,290 hours of operation and 218 kg of Xenon propellant have been accumulated by the DS1 flight spare ion thruster. The thruster is performing well with no measurable degradation in thrust, specific impulse, or thrust vector variation. Although the accelerator grid downstream wear is severe, preventing operation at the full power point, the engine is still fully operational at the mid to lower end of the throttle envelope. Severe discharge cathode keeper erosion, first observed after 8000 hours of operation, has continued through to the end of the third full power test segment, full exposing the cathode heater and orifice plate. In spite of this, photographic data indicates the heater has not been damaged, and ignition characteristics remain unchanged, after 291 restarts. Cathode life is one of the key life limiting mechanisms for future long duration science missions, utilizing ion thrusters. The IT2 cathode operating time will soon surpass the lifetime of the Space Station Plasma Contactor hollow cathode, establishing a service life capability beyond that previously thought possible for the hollow cathode technology. The neutralizer continues to perform well, with no visible keeper erosion. Previous deposits formed during operation at the minimum power point, appear to have been removed by subsequent operation at the full power point. Neutralizer operation at TH5 continues to be in spot mode after 1500 hours of run time. Degradation in electrical isolation continues to occur, but impedances remain sufficiently high to prevent major leakage paths. The thruster is currently running at the 1KW point, to investigate low power wear mechanisms and gain operating experience for the upcoming Dawn Mission.

10

AKNOWLEDGMENTS

The authors would like to acknowledge the invaluable efforts of the many people at JPL and NASA Glen Research Center who assist in conducting this test. They include, Al Owens, John Anderson, Jay Polk, Vince Rawlin, James Sovey, and Bob Toomath. The Jet Propulsion Laboratory, California Institute of Technology carried out the research described in this paper, under a contract with the National Aeronautics and Space Administration.

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12. 13.

14. 15. 16. 17.

18.

19.

20.

Polk, J. E., et al, “Demonstration of the NSTAR Ion Propulsion System on the Deep Space One Mission,” 27* International Electric Propulsion Conference, October 2001. Polk, J. E., et al., “Validation of NSTAR Propulsion system on the DS1 Mission,” AIAA-99-2246, June 1999. Polk, J. E., et al., “In-Flight Performance of the NSTAR Ion Propulsion System on the Deep Space One Mission,” Z8-0304.PDF, IEEE Aerospace Conference Proceedings, March 2000. Christensen, J., et al., “Design and fabrication of a Flight Model 2.3 kW Ion Thruster for the Deep Space 1 Mission,’’ AIAA-98-3327, July 1998. Anderson, J. R., et al., “Results of an On-going Long Duration Ground Test of the DSl Flight Spare Ion Engine,” AIAA-99-2857, June 1999. Senguptz, A., et al., “Performance Characteristics of the DSI Flight Spare Ion Thruster Long Duration Test, After 21,300 Hours of Operation,” AIAA 2002-3959, July 2002. Polk, J. E., et al., “An Overview of the Results from an 8200 Hour Wear Test of the NSTAR Ion Thruster,” AIAA-99-2446, June 1999. Polk, J. E., et al., ‘The Effect of Engine Wear on Performance in the NSTAR 8000 Hour Ion Engine Endurance Test,” AIAA-97-3387, July 1997. Polk, J., E., et al., “In Situ, Time-Resolved Accelerator Grid Erosion Measurements in the NSTAR 8000 Hour Ion Engine Wear Test,” IEPC-97-047, 1997. Polk, J. E., et al., “A 1000-hour Wear Test of the NASA NSTAR Ion Thruster,” ANA-96-2717, July 1996. Brophy, J.R., et al., “Numerical Simulations of Ion Thruster Accelerator Grid Erosion,” AIAA-2002- 426 1, July 2002. Patterson, M. J., et al., “2.3 kW Ion Thrust er Wear Test,” AM-95-2516, July 1995. Sarver-Verhey,T.R., “Destructive Evaluation of a Xenon Hollow Cathode After a 28,000 Hour Life Test,”AIAA98-3482, , July 1998. Jahn, R. G., The Physics of Electric Propulsion, Mc-Graw Hill, New York, 1968. Rawlin, V. K., et al., “NSTAR Flight Thruster Qualification Testing,” AM-98-3936, July 1998. Garner, C. E., et al., “Methods for Cryopumping Xenon,” AIAA-96-3206, July 1996. Polk, J. E., et al., “Behavior of the Thrust Vector in the NSTAR Ion Thruster,” AM-98-3940, July 1998. Engelbrecht, C. S., “NSTAR Xenon Feed System (XFS) Technical Requirements Document (TRD),” NSTAR Document ND-330, January 1997. Ganapathi, G. B., et. al., “Post-Launch Performance Characterization of the Xenon Feed System on Deep Space One,” AIAA-99-2273, June 1999. Patterson, M., et al., “NASA 30 cm Ion Thruster Development Status,’’ AM-94-2849, June 1994.

11