Electric Propulsion

I n December 2002, when France’s Stentorisatellite was all set to use electricpropulsion for stationkeeping, ESA’s

SMART-1 was just completing its first end-to-end spacecraft test. Then Stentor was lost inthe Ariane-5 launch failure, making SMART-1the first and only technology demonstrationmission with Hall-effect plasma propulsion. Asa result, there was a great deal of interest inthe electric propulsion community in SMART-1’sflight.

IntroductionSMART-1 was the first of ESA’s SmallMissions for Advanced Research inTechnology. The objective was to test newenabling technologies for the forthcomingESA Cornerstone science missions, thenext being the BepiColombo Mercuryorbiters. SMART-1 was launched fromKourou, French Guiana, on 27September 2003, as an auxiliarypassenger on an Ariane-5; its highlysuccessful mission concluded with adeliberate impact on the Moon on3 September 2006.

The critical technology demonstratedby SMART-1 was the main solar-electric propulsion using a PPS-1350GHall-effect plasma thruster, developed

Denis Estublier, Giorgio Saccoccia & Jose Gonzalez del AmoPropulsion and Aerothermodynamics Division,Directorate of Technical and QualityManagement, ESTEC, Noordwijk,The Netherlands

esa bulletin 129 - february 2007 41

Electric Propulsionon SMART-1

A Technology Milestone

Estublier 2/8/07 3:15 PM Page 40

Electric Propulsion

I n December 2002, when France’s Stentorisatellite was all set to use electricpropulsion for stationkeeping, ESA’s

SMART-1 was just completing its first end-to-end spacecraft test. Then Stentor was lost inthe Ariane-5 launch failure, making SMART-1the first and only technology demonstrationmission with Hall-effect plasma propulsion. Asa result, there was a great deal of interest inthe electric propulsion community in SMART-1’sflight.

IntroductionSMART-1 was the first of ESA’s SmallMissions for Advanced Research inTechnology. The objective was to test newenabling technologies for the forthcomingESA Cornerstone science missions, thenext being the BepiColombo Mercuryorbiters. SMART-1 was launched fromKourou, French Guiana, on 27September 2003, as an auxiliarypassenger on an Ariane-5; its highlysuccessful mission concluded with adeliberate impact on the Moon on3 September 2006.

The critical technology demonstratedby SMART-1 was the main solar-electric propulsion using a PPS-1350GHall-effect plasma thruster, developed

Denis Estublier, Giorgio Saccoccia & Jose Gonzalez del AmoPropulsion and Aerothermodynamics Division,Directorate of Technical and QualityManagement, ESTEC, Noordwijk,The Netherlands

esa bulletin 129 - february 2007 41

Electric Propulsionon SMART-1

A Technology Milestone

Estublier 2/8/07 3:15 PM Page 40

and qualified by SNECMA (F). Thisdemonstration created an impressivenumber of ‘firsts’ and broke a number ofESA, European and world records.

A Number of FirstsThe SMART-1 electric propulsionsystem (EPS) was directly managed andprocured by the Electric PropulsionSection of ESA’s Directorate of Techni-cal and Quality Management, anddelivered as ‘customer-furnished equip-ment’ to the SMART-1 Project in ESA’sScience Directorate and then to theindustrial prime contractor, the SwedishSpace Corporation. This was the firsttime that EPS was used for something ascritical as primary propulsion. It wasalso the first time that:

– electric propulsion was used to escapeEarth from a geostationary transferorbit;

– electric propulsion was in conjunctionwith gravity-assist manoeuvres;

– electric propulsion achieved captureusing the weak-stability boundariesand a descent around a celestial body;

– ESA flew a lunar orbiter;– a lunar orbiter had a propellant

propellant-demanding scientific missionsor long-duration commercial missions.

With its mission, SMART-1 has madea very valuable contribution to spread-ing the use of EP.

EP Mission PhasesThe 3-day Low Earth Orbit Phasestarting after separation from Ariane,commissioned all the critical subsystemsand rapidly raise the altitude in order tolimit radiation exposure of the solarcells and sensitive electronics. The VanAllen Belts Escape Phase used near-continuous thrust to minimise thepassage time through the radiation belts,although battery capacity was notenough for thrusting during eclipses.Also, thrusting below 3600 km altitudewas not possible because blurring of thestartracker images prevented finepointing. During this phase, the thrustwas directed along the velocity vector inorder to raise the perigee height quicklyabove 13 600 km, well beyond theradiation belts.

During the Earth Escape CruisePhase, EP was active for only half ofeach orbit because thrusting wasperformed only around the perigee,where it used the propellant moreefficiently.

As SMART-1’s orbit spiralled outbeyond 200 000 km it began to feelsignificant tugs from the Moon’s gravity,until the critical capture manoeuvrewith 4.5 days of EP thrust. During950 hours of cumulative thrust, EP thenlowered SMART-1 into a polar orbit of2000 x 4600 km, where the LunarScience Phase began in February 2005.

After 6 months of science operationsin ‘free drift’, a new EP phase of340 hours of thrusting extended the life of the lunar observation orbit by1 year.

The SMART-1 PlatformUsing EP required a 3-axis stabilisedspacecraft with two solar wings. Thecentral structure was designed around a49-litre cylindrical tank containing82 kg of xenon at launch. Since thethrust vector and solar wings had to be

fraction as low as 22% (only 82 kg ofxenon propellant).

Finally, from a technology point ofview, it was the first time that:

– a Hall-effect plasma thruster was usedfor primary propulsion;

– a Hall-effect propulsion system wasused under variable power conditions;

– a plasma thruster had operatedcontinuously for more than 240 hoursin space;

– a plasma thruster had accumulatednearly 5000 hours of operations inspace.A demonstration mission was key for

full acceptance of the maturity ofelectric propulsion (EP), and to showthat it could benefit from the latestadvances in space power generation.The mission not only showed that thepropulsion system performed asexpected, but also that the systemreliability (designed mostly as ‘singlestring’) was very good, despite thechallenging environmental conditionsthroughout the mission (repeatedcrossings of radiation belts, solar flares,lunar albedo).

optimally pointed simultaneously, thewings were fitted with a drivemechanism. For maximum power, theyused integrated triple-junction gallium-arsenide solar cells of 23.7% efficiency(beginning of life), reduced to 18–19%in the 40–50ºC operating temperature

range, with cover glass and at fixed busvoltage. They were sized to deliver1850 W at the beginning of the mission.

Power was delivered to the EPS via a50 V fully regulated bus. State-of-the-artlithium-ion batteries provided powerduring eclipses, for a maximum 2.1 hourswithout thrusting.

The thruster was mounted on anorientation mechanism (developed byContraves Space AG) to point through acentre of mass that was moving owingto propellant depletion and thermaleffects. SMART-1 also demonstratedthe use of a gimballed electric thrusterfor unloading the reaction wheels,thereby saving precious kilograms ofhydrazine.

In addition, an understanding offlight operations with EP was funda-mental. EP was often viewed asinconvenient owing to the long thrustperiods needed to compensate for thelow thrust. Compared to the muchshorter thrust periods required bychemical propulsion, EP must be pairedwith relatively high onboard autonomyand accurate pointing of the thrustvector and solar array over long periods.

This certainly requires a new way ofcontrolling missions from the ground.SMART-1 has paved the way whilecollecting important experience, such asin the sensitivity to the protonenvironment, the need to interruptthrust during eclipses and open-looponboard power regulation.

From a mission point of view, thelong low-thrust manoeuvres have thegreat advantage that, because changesoccur slowly, errors can be easilycorrected at negligible propellant cost.Failures can be more easily recoveredfrom than with traditional manoeuvresbecause there is much more timeavailable for a second attempt or todecide on a fall-back solution. Finally,EP is very often a mission ‘enabler’ for

Technical & Quality Management

esa bulletin 129 - february 2007esa bulletin 129 - february 2007 www.esa.intwww.esa.int 4342

Electric Propulsion

SMART-1 ready for encapsulation on its launcher (ESA/CSG)

412th dayMoon injecMoon injec tiontion3700 hours of thrust289 on/off operations332 orbits around Earth84 millions Km57 kg of Xenon used

1st day (27/9/2003)LaunchLaunch

1072nd day (3/9/2006)Moon impacMoon impac tt

3rd dayFF irst Engine Ignitionirst Engine Ignition

721st dayLast Engine OLast Engine O perapera tiontion4958.3 hours of thrust844 on/off operations82 kg of Xenon used(only 250 g left)

532nd dayReached MoonReached MoonOO perapera tional Otional O rbitrbit4627 hours of thrust526 on/off operations72.5 kg of Xenon used

The SMART-1 Hall-effect thruster flightmodel (above right).Thruster tests (left)

Estublier 2/8/07 3:15 PM Page 42

and qualified by SNECMA (F). Thisdemonstration created an impressivenumber of ‘firsts’ and broke a number ofESA, European and world records.

A Number of FirstsThe SMART-1 electric propulsionsystem (EPS) was directly managed andprocured by the Electric PropulsionSection of ESA’s Directorate of Techni-cal and Quality Management, anddelivered as ‘customer-furnished equip-ment’ to the SMART-1 Project in ESA’sScience Directorate and then to theindustrial prime contractor, the SwedishSpace Corporation. This was the firsttime that EPS was used for something ascritical as primary propulsion. It wasalso the first time that:

– electric propulsion was used to escapeEarth from a geostationary transferorbit;

– electric propulsion was in conjunctionwith gravity-assist manoeuvres;

– electric propulsion achieved captureusing the weak-stability boundariesand a descent around a celestial body;

– ESA flew a lunar orbiter;– a lunar orbiter had a propellant

propellant-demanding scientific missionsor long-duration commercial missions.

With its mission, SMART-1 has madea very valuable contribution to spread-ing the use of EP.

EP Mission PhasesThe 3-day Low Earth Orbit Phasestarting after separation from Ariane,commissioned all the critical subsystemsand rapidly raise the altitude in order tolimit radiation exposure of the solarcells and sensitive electronics. The VanAllen Belts Escape Phase used near-continuous thrust to minimise thepassage time through the radiation belts,although battery capacity was notenough for thrusting during eclipses.Also, thrusting below 3600 km altitudewas not possible because blurring of thestartracker images prevented finepointing. During this phase, the thrustwas directed along the velocity vector inorder to raise the perigee height quicklyabove 13 600 km, well beyond theradiation belts.

During the Earth Escape CruisePhase, EP was active for only half ofeach orbit because thrusting wasperformed only around the perigee,where it used the propellant moreefficiently.

As SMART-1’s orbit spiralled outbeyond 200 000 km it began to feelsignificant tugs from the Moon’s gravity,until the critical capture manoeuvrewith 4.5 days of EP thrust. During950 hours of cumulative thrust, EP thenlowered SMART-1 into a polar orbit of2000 x 4600 km, where the LunarScience Phase began in February 2005.

After 6 months of science operationsin ‘free drift’, a new EP phase of340 hours of thrusting extended the life of the lunar observation orbit by1 year.

The SMART-1 PlatformUsing EP required a 3-axis stabilisedspacecraft with two solar wings. Thecentral structure was designed around a49-litre cylindrical tank containing82 kg of xenon at launch. Since thethrust vector and solar wings had to be

fraction as low as 22% (only 82 kg ofxenon propellant).

Finally, from a technology point ofview, it was the first time that:

– a Hall-effect plasma thruster was usedfor primary propulsion;

– a Hall-effect propulsion system wasused under variable power conditions;

– a plasma thruster had operatedcontinuously for more than 240 hoursin space;

– a plasma thruster had accumulatednearly 5000 hours of operations inspace.A demonstration mission was key for

full acceptance of the maturity ofelectric propulsion (EP), and to showthat it could benefit from the latestadvances in space power generation.The mission not only showed that thepropulsion system performed asexpected, but also that the systemreliability (designed mostly as ‘singlestring’) was very good, despite thechallenging environmental conditionsthroughout the mission (repeatedcrossings of radiation belts, solar flares,lunar albedo).

optimally pointed simultaneously, thewings were fitted with a drivemechanism. For maximum power, theyused integrated triple-junction gallium-arsenide solar cells of 23.7% efficiency(beginning of life), reduced to 18–19%in the 40–50ºC operating temperature

range, with cover glass and at fixed busvoltage. They were sized to deliver1850 W at the beginning of the mission.

Power was delivered to the EPS via a50 V fully regulated bus. State-of-the-artlithium-ion batteries provided powerduring eclipses, for a maximum 2.1 hourswithout thrusting.

The thruster was mounted on anorientation mechanism (developed byContraves Space AG) to point through acentre of mass that was moving owingto propellant depletion and thermaleffects. SMART-1 also demonstratedthe use of a gimballed electric thrusterfor unloading the reaction wheels,thereby saving precious kilograms ofhydrazine.

In addition, an understanding offlight operations with EP was funda-mental. EP was often viewed asinconvenient owing to the long thrustperiods needed to compensate for thelow thrust. Compared to the muchshorter thrust periods required bychemical propulsion, EP must be pairedwith relatively high onboard autonomyand accurate pointing of the thrustvector and solar array over long periods.

This certainly requires a new way ofcontrolling missions from the ground.SMART-1 has paved the way whilecollecting important experience, such asin the sensitivity to the protonenvironment, the need to interruptthrust during eclipses and open-looponboard power regulation.

From a mission point of view, thelong low-thrust manoeuvres have thegreat advantage that, because changesoccur slowly, errors can be easilycorrected at negligible propellant cost.Failures can be more easily recoveredfrom than with traditional manoeuvresbecause there is much more timeavailable for a second attempt or todecide on a fall-back solution. Finally,EP is very often a mission ‘enabler’ for

Technical & Quality Management

esa bulletin 129 - february 2007esa bulletin 129 - february 2007 www.esa.intwww.esa.int 4342

Electric Propulsion

SMART-1 ready for encapsulation on its launcher (ESA/CSG)

412th dayMoon injecMoon injec tiontion3700 hours of thrust289 on/off operations332 orbits around Earth84 millions Km57 kg of Xenon used

1st day (27/9/2003)LaunchLaunch

1072nd day (3/9/2006)Moon impacMoon impac tt

3rd dayFF irst Engine Ignitionirst Engine Ignition

721st dayLast Engine OLast Engine O perapera tiontion4958.3 hours of thrust844 on/off operations82 kg of Xenon used(only 250 g left)

532nd dayReached MoonReached MoonOO perapera tional Otional O rbitrbit4627 hours of thrust526 on/off operations72.5 kg of Xenon used

The SMART-1 Hall-effect thruster flightmodel (above right).Thruster tests (left)

Estublier 2/8/07 3:15 PM Page 42

A key milestone in the project was thefull-scale integrated test of the propul-sion system. During the end-to-end testin December 2002, SMART-1 (sanssolar array) was installed in the HBF-3vacuum chamber in ESTEC for a 1-hourtrial of the thruster and its orientationmechanism, thereby validating all theinterfaces and control hardware.

The SMART-1 Electric Propulsion SystemThe SNECMA PPS-1350G Hall-effectthruster is the European version of theRussian Fakel SPT-100, but usingEuropean technologies and qualitystandards required by western telecomm-unications satellites. The thruster’sground qualification totalled 9200 hours

Controller (XFC). A simple and robustcontrol loop algorithm in the regulationelectronics (PRE Card) controlled thepressure delivered by the regulator byopening or closing two solenoid valvesin series following a predefined sequence.Then the XFC delivered, in a controlloop with the discharge current, thexenon flow to the thruster anode andcathode.

The thruster was controlled andpowered by the Power Processing Unit(PPU), developed and qualified byAlcatel ETCA (B). All telemetry andtelecommands were interfaced throughthe PRE Card. An electrical Filter Unit,produced by EREMS (F), reduced thethruster discharge oscillations to anacceptable level and protected the PPUelectronics. Both the PRE Card andPPU carried software with ‘automaticmode’ subroutines that minimised thenumber of commands to be suppliedroutinely to the EPS through spacecrafttelemetry.

To cope with decreasing and varyingpower from the solar array as themission progressed, the EPS wasdesigned to be throttled easily over awide range of input power. The PPUallowed 117 power levels to be specified,ranging from 462 W up to 1190 W. Italso performed the automatic ignitionsequence. After the ‘auto exec’command was given, the cathode washeated, the xenon flow began and thethruster was turned on by the ignitionpulse.

The EPS was a single-string system,but with some internal redundancy. Thepressure regulator valves were duplica-ted, as were the thruster’s hollowcathodes and their XFC connections.

In routine operation, EPS wasrelatively simple to manage. It wascommanded to thrust by a pre-builtsequence of 35 commands, and switchedoff by one unique command. It was thena simple matter of supplying the desiredthrust level and ignition time whenever athrust arc was required. Based on thedesired orbital change, there were onlytwo events and one parameter for eachthrust arc: EP-ON time, thrust level, and

of cumulative operation and 7200 on/offcycles for a total impulse of 2.9 MNs.This covers the Alphabus requirement,for which the PPS-1350G(4) engine hasbeen chosen, to perform 15 years of dailystationkeeping in geostationary orbit.

Hall-effect, or plasma, thrusters oper-ate with a noble gas such as xenon, well-known for good storability and user-friendliness during ground operations.This is a major advantage over chemicalpropulsion, which requires costly safetyprocedures during propellant handling.

The xenon is ionised and acceleratedto 60 000 km/h by an electric field,thereby producing thrust. Ionisationoccurs when an electron current isgenerated by the external cathode. This

EP-OFF time. This information wasthen processed by the control systemand automatically incorporated into theSMART-1 command sequence and themission planning system as part of theroutine command uplinks.

SMART-1 Plasma Diagnostic InstrumentsThe potentially undesirable effects fromEPS included surfaces contaminated withsputtered material and eroded by ionimpingement. The Electric PropulsionDiagnostic Package (EPDP) wasdeveloped by Alcatel Alenia Space (I) tomeasure the effects and prove whetherthey affected spacecraft operations.Surfaces could have charged up, creatingelectrical problems. Measurements invacuum chambers were not veryrepresentative and the real spacecraftconfiguration could not be reproduced;flight experience was essential.

In order to measure these effects,EPDP included:

Retarding Potential Analyser (RPA):measured the ion energy and currentdensity distribution.

Langmuir Probe (LP): measured theplasma potential, electron density and

electron temperature. The LP and theRPA were on the same panel with thethruster at the centre; RPA’s axis wasoriented towards the thruster.

Solar Cell (SC): positioned away fromthe solar array.

Quartz-Crystal Microbalance (QCM):used to ‘weigh’ deposited material,providing real contamination data. SCand QCM were installed on thespacecraft’s outer –X panel.

EPS Flight PerformanceSMART-1 successfully completed all ofits EP operations, including someadvanced low-pressure thrusting for the1-year extended mission up to theimpact in September 2006. Orbitalchanges were monitored by ESA’s Euro-pean Space Operations Centre (ESOC)in Darmstadt (D), and the craft’sacceleration measured. From these data,the actual thrust levels were accuratelydetermined – valuable feedback fromSMART-1 to compare with ground dataand to confirm the good behaviour ofthe thruster and EPS overall.

During ground tests, the thrust wasmeasured with an accuracy of ±2.5%and the variable thrust-versus-powerrelationship was mapped using an olderQualification Model thruster. Thesecorrelated data were used to predict thethruster’s flight performance.

current is constrained by a radialmagnetic field. The Hall effect meansthat the ions and electrons swerve inopposite directions in the magnetic field,creating an electric field. This expels thexenon ions as a propulsive jet. Electronsfrom an external cathode neutralise thebeam to prevent the spacecraft frombecoming electrically charged.

On SMART-1, the xenon was storedin the central tank under supercriticalconditions, at 1.7 times the density ofwater and high pressure (up to 150 bars).A pressure regulator, designed bySNECMA Moteurs and Iberespacio(E), reduced the xenon pressure to anominal 2 bars. The low-pressure gaswas then fed into the Xenon Flow

Technical & Quality Management

esa bulletin 129 - february 2007esa bulletin 129 - february 2007 www.esa.intwww.esa.int 4544

Electric Propulsion

Electric Propulsion Performance

SMART-1 launch mass 370 kgCumulative EPS duration 4958 hoursCumulative regulated thrust 4913 hoursTotal impulse 1.2 MNsVelocity increment 3.7 km/sLongest burn duration 240 hoursNumber ON/OFF cycles 844Number valve activations 1 256 505Xenon at launch 82 kgXenon remaining at impact 0.28 kgDischarge power range 462–1190 WAverage discharge power 1140 WAverage projected thrust 65.7–9.1 mNAverage measured thrust 67 mNAverage effective mass flow 4.44 mg/sAverage effective specific impulse 1540 s

1 SIR IR Spectrometer

2 Sun sensors

3 SPEDE potential &plasma environment

4 Camera

5 D-CIXS Imaging X-raySpectrometer

9

Fuel tank for attitude control

10

Star tracker

11

Motor to turn solar array

12

Communication transponders

13

Ion engine control electronics

14

Attitude control thrusters8

7 EPDP sensors

6 Communication antenna

Ion engine with orientationmechanism (to maintain thruster pointing as fuel tanks drain)

12

45

2

3

6

7

78

9 10

11

12

13

13

13

1314

7

The PRE Card controlled the xenon flow

The Xenon Flow Controller delivered the propellant to thethruster

Estublier 2/8/07 3:15 PM Page 44

A key milestone in the project was thefull-scale integrated test of the propul-sion system. During the end-to-end testin December 2002, SMART-1 (sanssolar array) was installed in the HBF-3vacuum chamber in ESTEC for a 1-hourtrial of the thruster and its orientationmechanism, thereby validating all theinterfaces and control hardware.

The SMART-1 Electric Propulsion SystemThe SNECMA PPS-1350G Hall-effectthruster is the European version of theRussian Fakel SPT-100, but usingEuropean technologies and qualitystandards required by western telecomm-unications satellites. The thruster’sground qualification totalled 9200 hours

Controller (XFC). A simple and robustcontrol loop algorithm in the regulationelectronics (PRE Card) controlled thepressure delivered by the regulator byopening or closing two solenoid valvesin series following a predefined sequence.Then the XFC delivered, in a controlloop with the discharge current, thexenon flow to the thruster anode andcathode.

The thruster was controlled andpowered by the Power Processing Unit(PPU), developed and qualified byAlcatel ETCA (B). All telemetry andtelecommands were interfaced throughthe PRE Card. An electrical Filter Unit,produced by EREMS (F), reduced thethruster discharge oscillations to anacceptable level and protected the PPUelectronics. Both the PRE Card andPPU carried software with ‘automaticmode’ subroutines that minimised thenumber of commands to be suppliedroutinely to the EPS through spacecrafttelemetry.

To cope with decreasing and varyingpower from the solar array as themission progressed, the EPS wasdesigned to be throttled easily over awide range of input power. The PPUallowed 117 power levels to be specified,ranging from 462 W up to 1190 W. Italso performed the automatic ignitionsequence. After the ‘auto exec’command was given, the cathode washeated, the xenon flow began and thethruster was turned on by the ignitionpulse.

The EPS was a single-string system,but with some internal redundancy. Thepressure regulator valves were duplica-ted, as were the thruster’s hollowcathodes and their XFC connections.

In routine operation, EPS wasrelatively simple to manage. It wascommanded to thrust by a pre-builtsequence of 35 commands, and switchedoff by one unique command. It was thena simple matter of supplying the desiredthrust level and ignition time whenever athrust arc was required. Based on thedesired orbital change, there were onlytwo events and one parameter for eachthrust arc: EP-ON time, thrust level, and

of cumulative operation and 7200 on/offcycles for a total impulse of 2.9 MNs.This covers the Alphabus requirement,for which the PPS-1350G(4) engine hasbeen chosen, to perform 15 years of dailystationkeeping in geostationary orbit.

Hall-effect, or plasma, thrusters oper-ate with a noble gas such as xenon, well-known for good storability and user-friendliness during ground operations.This is a major advantage over chemicalpropulsion, which requires costly safetyprocedures during propellant handling.

The xenon is ionised and acceleratedto 60 000 km/h by an electric field,thereby producing thrust. Ionisationoccurs when an electron current isgenerated by the external cathode. This

EP-OFF time. This information wasthen processed by the control systemand automatically incorporated into theSMART-1 command sequence and themission planning system as part of theroutine command uplinks.

SMART-1 Plasma Diagnostic InstrumentsThe potentially undesirable effects fromEPS included surfaces contaminated withsputtered material and eroded by ionimpingement. The Electric PropulsionDiagnostic Package (EPDP) wasdeveloped by Alcatel Alenia Space (I) tomeasure the effects and prove whetherthey affected spacecraft operations.Surfaces could have charged up, creatingelectrical problems. Measurements invacuum chambers were not veryrepresentative and the real spacecraftconfiguration could not be reproduced;flight experience was essential.

In order to measure these effects,EPDP included:

Retarding Potential Analyser (RPA):measured the ion energy and currentdensity distribution.

Langmuir Probe (LP): measured theplasma potential, electron density and

electron temperature. The LP and theRPA were on the same panel with thethruster at the centre; RPA’s axis wasoriented towards the thruster.

Solar Cell (SC): positioned away fromthe solar array.

Quartz-Crystal Microbalance (QCM):used to ‘weigh’ deposited material,providing real contamination data. SCand QCM were installed on thespacecraft’s outer –X panel.

EPS Flight PerformanceSMART-1 successfully completed all ofits EP operations, including someadvanced low-pressure thrusting for the1-year extended mission up to theimpact in September 2006. Orbitalchanges were monitored by ESA’s Euro-pean Space Operations Centre (ESOC)in Darmstadt (D), and the craft’sacceleration measured. From these data,the actual thrust levels were accuratelydetermined – valuable feedback fromSMART-1 to compare with ground dataand to confirm the good behaviour ofthe thruster and EPS overall.

During ground tests, the thrust wasmeasured with an accuracy of ±2.5%and the variable thrust-versus-powerrelationship was mapped using an olderQualification Model thruster. Thesecorrelated data were used to predict thethruster’s flight performance.

current is constrained by a radialmagnetic field. The Hall effect meansthat the ions and electrons swerve inopposite directions in the magnetic field,creating an electric field. This expels thexenon ions as a propulsive jet. Electronsfrom an external cathode neutralise thebeam to prevent the spacecraft frombecoming electrically charged.

On SMART-1, the xenon was storedin the central tank under supercriticalconditions, at 1.7 times the density ofwater and high pressure (up to 150 bars).A pressure regulator, designed bySNECMA Moteurs and Iberespacio(E), reduced the xenon pressure to anominal 2 bars. The low-pressure gaswas then fed into the Xenon Flow

Technical & Quality Management

esa bulletin 129 - february 2007esa bulletin 129 - february 2007 www.esa.intwww.esa.int 4544

Electric Propulsion

Electric Propulsion Performance

SMART-1 launch mass 370 kgCumulative EPS duration 4958 hoursCumulative regulated thrust 4913 hoursTotal impulse 1.2 MNsVelocity increment 3.7 km/sLongest burn duration 240 hoursNumber ON/OFF cycles 844Number valve activations 1 256 505Xenon at launch 82 kgXenon remaining at impact 0.28 kgDischarge power range 462–1190 WAverage discharge power 1140 WAverage projected thrust 65.7–9.1 mNAverage measured thrust 67 mNAverage effective mass flow 4.44 mg/sAverage effective specific impulse 1540 s

1 SIR IR Spectrometer

2 Sun sensors

3 SPEDE potential &plasma environment

4 Camera

5 D-CIXS Imaging X-raySpectrometer

9

Fuel tank for attitude control

10

Star tracker

11

Motor to turn solar array

12

Communication transponders

13

Ion engine control electronics

14

Attitude control thrusters8

7 EPDP sensors

6 Communication antenna

Ion engine with orientationmechanism (to maintain thruster pointing as fuel tanks drain)

12

45

2

3

6

7

78

9 10

11

12

13

13

13

1314

7

The PRE Card controlled the xenon flow

The Xenon Flow Controller delivered the propellant to thethruster

Estublier 2/8/07 3:15 PM Page 44

The EPS was used in the variablepower mode, especially at start-up, butthe thruster power used remained closeto its maximum (1140 W vs. 1190 W).The average measured thrust overalmost 5000 hours was well within thepredicted average thrust range (67 mNvs. 65.7–9.1 mN), even though theprojection was done using measure-ments from a different, older thruster.SMART-1 therefore demonstrated thevery good reproducibility and stabilityin time of thruster performance.

Operation History of the EPSUp to 1.3 million valve activations werecounted, with no signs of propellantleakage even though the qualificationlimit of 1 million cycles was wellexceeded. The procedures developedduring flight allowed the residual xenonmass in the tank to be reduced from1.8 kg to 0.28 kg.

Thanks to the plasma diagnosticinstruments, correlated data from thelocal plasma potential and from theenergy of the backflow particles showedthe very good operation of the cathode.A cathode is usually considered to be asensitive component and is thereforemade with redundancy. The mainparameter for controlling its goodhealth is the working reference potential.Because of the periodic variations of thespacecraft reference potential, a secondindependent reference potential wasrequired. However, the potential of thenominal cathode was shown to be highlystable throughout the mission. In

addition, the thruster always (844 times)started at the first ignition pulse on thecathode. Thus SMART-1 proved thistype of hollow cathode developed bySNECMA.

Furthermore, none of the EPSperformance parameters was affected bythe artificial plasma surrounding thespacecraft or by the periodic oscillationof the spacecraft potential.

Interaction with the EnvironmentIn the initial stage of the mission, themain effect on SMART-1 from itsenvironment was on the solar cells,damaged by the protons in the radiationbelts. Taking into account the Earth-Sundistance changing with season, the solararray power fell by around 8%, whichincluded the effect of the very intensesolar flare of October 2003. Thedegradation ended after only 2 months,when the perigee altitude rose above5900 km. The solar array was sized tocover a 12.3% drop in power at 40ºC,and 17.7% at the maximum celltemperature of 100ºC. Most of the timeduring EPS operation, the cell tempera-ture was around 60ºC.

From the analysed EPDP flight data,the artificially generated plasma fromthe propulsion system had no criticaleffect on SMART-1. No indication ofincreased erosion or contamination wasfound; no surface-charging effects weredetected. On the contrary, the denseartificial plasma generated by the EPShad a neutralising and stabilising effecton the spacecraft potential, which could

otherwise have had a large positivepotential, in particular when crossingthe radiation belts.

The spacecraft potential was extremelyneutral, stable and fully independent ofthe surrounding natural plasma, inclu-ding that of the proton and electronsbelts. This would have certainly not beenthe case for a spacecraft without EP andunprotected by a dense, surroundingartificial plasma. Furthermore, thesmall variations in spacecraft potentialare fully understood and exactlycorrelated to specific thruster operatingconditions and solar array attitude withrespect to the thruster plume.

ConclusionsBy exceeding its mission objectives,SMART-1 has made a very significantcontribution to the promotion ofelectric propulsion. The impressiveresults from Europe’s first scientificlunar experiments have also helped EPenormously in gaining a larger audiencearound the world.

In addition, thanks to the EPSperformance, the limited degradation ofthe solar arrays and the optimisedtransfer strategy, SMART-1 completeda 1-year mission extension, therebytripling the scientific observation period.On 11 November 2004, SMART-1became the first ever mission to escapeEarth using electric propulsion, and thefirst to use it for capture by thegravitational field of another celestialbody.

All the operational elements of thishighly successful spacecraft weredemonstrated, paving the way for futuremissions. The performance of a singlethruster, in various operating andenvironmental conditions, is clearevidence of the robustness of theSMART-1 EPS design, and of thecapability of this type of thruster toaccomplish a wide range of spacemissions, including scientific andcommercial. e

Technical & Quality Management

esa bulletin 129 - february 2007 www.esa.int46

Detailed information on SMART-1 and its mission canbe found at www.esa.int/smart1

Estublier 2/8/07 3:15 PM Page 46