Editor's Note: Sessions 4-6 are for advanced learners. We recommend that learners follow these sessions in sequence.
The ion engine
As the discussion in Session 3 illustrates, ion engines were among the very first electric propulsion technologies to be conceptualized by chemical propulsion pioneers such as Robert Goddard and Hermann Oberth. An ion propulsion system converts energy from the spacecraft power system into the kinetic energy of a beam of ions (ionized molecules). This beam propels the spacecraft in the opposite direction as it exits the thruster at high speed. As in the case of any EP technology, the ion thruster propulsion system consists of five major components: a computer for controlling and monitoring system performance; a power source (e.g., solar arrays); a power processing unit (PPU) for converting power from the power source to the proper voltages and current "quality" for the engine; the propellant storage and delivery system; and the thruster itself.
One of the first modern (i.e., gas propellant) "ion-engine-like" devices flown in space was a Hughes Research Laboratories (now Boeing) ion source launched in 1979 on the Air Force Geophysics Laboratory's Spacecraft Charging at High Altitude (SCATHA) satellite. This ion source was not used to propel the vehicle, but to change its electrical charge. This company developed PanAmSat 5, which in 1997 became the first communications satellite to use an ion engine. This ion engine is used solely for stationkeeping. By 2001, Boeing was employing larger ion thrusters both for stationkeeping and orbit raising.
While more than a dozen commercial spacecraft now use ion propulsion for stationkeeping and orbit raising, NASA's Deep Space 1 (DS1), the first in its New Millennium Program of missions, was the first craft to flight-test ion propulsion for deep-space travel. DS1 was launched from Cape Canaveral on October 24, 1998. During a highly successful primary mission, it tested a dozen advanced, high-risk-high-payoff technologies, including the 3,000+ second specific impulse NSTAR ion propulsion system (IPS), new sensors and navigational tools, and advanced concentrator solar arrays.
The total mass of the spacecraft at launch was about 486 kg. The spacecraft had 31 kg of hydrazine for its chemical propulsion thruster and 81.5 kg of xenon for its IPS at launch. The concentrator solar arrays on DS1 generated a maximum of about 2.5 kW of power.
The spacecraft configuration is shown below in an artist's rendition (figure ES1) and in a test firing of the IPS (figure ES2).
NASA |
| Figure ES2: DS1 firing its NSTAR ion propulsion system in a vacuum chamber at the NASA Jet Propulsion Laboratory prior to launch. The IPS was developed jointly by the Jet Propulsion Laboratory, NASA's Glenn Research Center and Boeing. |
NASA |
| Figure ES1: Artist's rendition of Deep Space 1 in flight. |
In an extremely successful extended mission, DS1 encountered comet Borrelly on September 22, 2001, and returned the best images and science data ever captured from a comet. During its fully successful hyperextended mission, DS1 conducted further technology tests. The focus of the hyperextended mission was to challenge the IPS by testing it in various modes that would have been too risky earlier in the mission. The IPS provided a ΔV of 4.2 km/s (9,400 mph) to the spacecraft, while consuming less than 70 kilograms of xenon propellant, for an overall payload mass ratio of over 85%. For comparison, if DS1 had to rely on its hydrazine propulsion system only, the equivalent payload mass ratio would have been less than 14%. The IPS has accumulated more than 640 days of thrust time versus an original goal of 365 days. The spacecraft was retired fully operational on December 18, 2001. Three NSTAR engines identical to the one used on DS1 will be used in 2006 to propel the DAWN spacecraft to the asteroid belt that resides between Mars and Jupiter. DAWN will orbit Vesta and Ceres, two of the largest asteroids in our solar system. We will now use the NSTAR engine to illustrate how an ion thruster works.
Ion engine operation
In the early 1990s, NASA identified electric propulsion as an enabling technology for future deep-space missions. NASA's Glenn Research Center (GRC) partnered with the Jet Propulsion Laboratory (JPL) in NASA's Solar Electric Power Technology Application Readiness (NSTAR) project. The purpose of NSTAR was to develop an ion propulsion system for deep-space missions. In 1996, a prototype NSTAR engine was tested for 8,000 hours in a vacuum chamber to simulate a deep-space mission. This prototype led directly to the construction and integration of the NSTAR flight engine on DS1.
NASA |
| Figure ES3: Operating principles of the DS1 ion thruster. |
An ion thruster such as the NSTAR engine is composed of several subsystems including (figure ES3): the discharge chamber, the discharge cathode assembly (DCA), the grids (also called the ion optics), and the neutralizer. Propellant is injected both through the DCA and through a ring of injectors. The DCA emits electrons that are accelerated by the electric field established between the positively biased discharge chamber walls and the negatively biased DCA. These electrons ionize the propellant by striking the gas atoms, knocking away one or more of the electrons orbiting an atom's nucleus. The "ring-cusp" magnetic field created by the magnets that surround the discharge chamber is used to improve the ionization efficiency of the engine by increasing the electron's "residence time" in the discharge chamber; that is, the longer an electron remains in the discharge chamber, the more opportunity it has to ionize propellant atoms.
The propellant used in DS1's NSTAR ion engine is a chemically inert, colorless, odorless and tasteless gas called xenon. Xenon is the propellant of choice for most electrostatic engines because of its large mass and thus high thrust-to-current ratio (recall Stuhlinger's argument in Session 3) and the relative ease by which it is ionized.
At the rear of the discharge chamber is the ion optics--a pair of Molybdenum grids that are charged to high voltage. The upstream grid, the screen grid, is maintained at DCA potential, which in the case of NSTAR is about 1,090 V above spacecraft ground (reference) at full thruster power. The second grid, the accel grid, is biased some 225 V below ground. The grids are each less than 1 mm thick and are placed about 0.7 mm apart from each other. Ions created in the discharge chamber enter the holes in the screen grid and are accelerated by the roughly 1,300 V drop in potential established by the two grids. Ions emerge from the accel grid at speeds in excess of 88,000 mph (142,000 km/h), fast enough to cover the distance between the Earth and Moon in less than the three hours! The electron-emitting neutralizer that is downstream of the accel grid keeps the spacecraft electrically neutral with respect to its environment by emitting one electron for every positively charged ion that leaves the thruster.
At full throttle, the NSTAR ion engine consumes about 2,300 watts of electric power and puts out 90 mN (0.02 pounds) of thrust. This is comparable to the force exerted by a single sheet of paper or a paper clip resting on the palm of your hand. Typical chemical on-board propulsion systems, on the other hand, produce far greater thrust [450 to 2250 N (100 to 500 pounds)] but for far shorter times. This brings about the story of the tortoise and the hare. While a chemically propelled spacecraft (the hare) gets its big boost and then coasts until the next boost, an ion-propelled spacecraft (tortoise) is under near constant (albeit miniscule) acceleration. For missions that require large ΔVs (above a few kilometers per second), the tortoise will always beat the hare! Moreover, ion propulsion systems are 10 times more "fuel efficient" than chemical propulsion systems and thus use a fraction of the propellant, which in turn results in smaller and lighter spacecraft and lower launch costs.
NASA |
| Figure ES5: The NSTAR ion thruster in operation. |
Image courtesy of Aerojet |
| Figure ES4: An Aerojet 30 cm-diameter, NSTAR class ion engine. Note the electron-beam neutralizer protruding forward from the 11 o'clock position of the thruster. This 2.35 kW thruster produces up to 93 mN of thrust with a specific impulse between 2,540 and 3,580 s. Total impulse of the engine is rated for more than 3 million Ns. |
Ion thrusters have many wonderful attributes. They produce large specific impulses (above 3,000 s) at high efficiencies (above 65%). While the grid erosion from ion impingement is the leading cause of thruster mortality, an NSTAR ion thruster has recently logged more than 22,000 hours (three years!) of operation in a vacuum chamber without a hitch. Because ion thrusters accelerate ions separately from the electrons, these devices have a thrust density limitation (force per unit thruster exit area) that results from so-called space-charge effects. However, to achieve long life, the NSTAR ion thruster (like other ion thrusters) operates nowhere near the space-charge current limit. If we look to the future, the advanced ion thrusters now under development will require lifetimes in excess of five years at specific impulses above 10,000 s. To achieve this long life, the current density through the grids will be reduced to a level even lower than NSTAR's, and more advanced grid materials (such as carbon or titanium, which are tougher than molybdenum) will be employed. This means that the size of the engine per unit of thrust will decrease and hence αt will increase. This limitation may be solved to some extent by another "ion thruster," the Hall thruster.
 |
| There is a friendly debate as to who developed the Hall thruster first and if the "other" side developed it independently. | |
|
The Hall thruster
The Hall thruster is an electrostatic engine that was developed in the 1960s in both the United States and the Soviet Union to alleviate the thrust density limitation of ion thrusters that results from space-charge effects between the grids. Hall thrusters were also attractive from the standpoint that since grids are not required to accelerate ions, they do not suffer from the grid erosion as a life-ending failure mechanism. Interest in the Hall thruster in the United States waned in the early 1970s, however, because of budgetary cuts and because American researchers were never able to demonstrate that these engines could operate at thrust efficiencies near those achieved with ion thrusters. As such, Hall thruster research essentially disappeared in the United States between 1972 and 1985. From 1985 to 1990, Ford Aerospace (now Space Systems/Loral), in conjunction with NASA's Lewis Research Center (now the Glenn Research Center [GRC]), funded a small research effort to determine if Hall thrusters could be used for North-South stationkeeping (NSSK). This program proved to be unsuccessful and was abandoned.
Throughout this period, however, Hall thruster research flourished in the Soviet Union, allegedly because Soviet engineers were never able to develop adequate grids for ion thrusters. Soviet Hall thrusters were first tested in space in 1971 with immediate success. Since then, over 100 Hall thrusters have been used on Soviet and Russian spacecraft, mostly as plasma contactors and for East-West stationkeeping. However, in 1994 Russia launched the first satellite to use Hall thrusters for NSSK. Because of this and numerous experiments that show that Russian Hall thrusters are capable of generating specific impulses of 1500 to 2200 seconds at thrust efficiencies of 50% or more, there has been a great deal of interest in using these engines on American spacecraft for NSSK and for orbit repositioning. For example, the Ballistic Missile Defense Organization (BMDO) in conjunction with NASA GRC and the Naval Research Laboratory (NRL) developed a flight experiment that used a Russian Hall thruster on a US experimental satellite. Space Systems/Loral (SSL) has announced that its latest communication satellite will use Hall thrusters for NSSK, perhaps as early as this year (2002). Clearly this device, which has a performance far superior to that of arcjets and which possesses high efficiency at optimum specific impulses for commercial space applications, is better-suited for Earth orbit missions than gridded ion thrusters--currently the most advanced propulsion system used on American spacecraft. Moreover, Hall thrusters would serve not only as an excellent thruster for orbit stationkeeping and repositioning roles, but potentially could be scaled in power and specific impulse to propel orbit transfer vehicles and future planetary probes.
The closed-drift Hall thruster
There are two types of Hall thrusters that have been studied at great length: the end Hall thruster and the closed-drift Hall thruster (CDT). Both engines, in principle, are capable of producing specific impulses in excess of 1,500 s with xenon at a thrust efficiency of 50% or greater. However, it is the CDT, which has been developed and used in the former Soviet Union over the past 40 years, that is of the most interest to the Western space technology community.
Regents of the University of Michigan |
| Figure ES6: Schematic of a closed-drift Hall thruster. |
The CDT is a coaxial device in which a magnetic field that is produced by an electromagnet is channeled between an inner core (pole piece) and outer ring that are typically made of iron (figure ES6). This configuration results in an essentially radial magnetic field with a peak strength of a few hundred gauss. The magnetic field strength is selected so that only the electrons are strongly influenced by the magnetic field; the much larger ions are not. In addition, an axial electric field is provided by applying a voltage between the anode and the downstream cathode. As the electrons stream upstream from the cathode to the anode, the radial magnetic field coupled with the axial electric field causes them to drift in the azimuthal (ExB) direction, forming a Hall current in the channel. Through collisions, these electrons ionize propellant molecules that are injected through the anode that are subsequently accelerated by the axial electric field.
Image courtesy of Aerojet |
| Figure ES7: Aerojet 4.5 kW-class BPT-4000 Hall thruster in operation. This thruster is capable of producing thrust levels between 170 and 290 mN at specific impulses between 1,750 and 3,000 s. This thruster will eventually replace the MR-510 arcjets for stationkeeping. |
The mixture of electrons and ions in the acceleration zone means that the plasma is electrically neutral, and as such is not space-charge limited in ion current (thrust) density like ion thrusters are. Since the magnetic field suppresses the axial motion of the electrons while exerting essentially no influence on the ions, the plasma can support an axial electric field with a potential difference close to the applied voltage between the electrodes. Thus, the bulk of the ions are accelerated to kinetic energies that are within 85% of the applied discharge voltage. This combination of processes accounts for the CDT's high thrust efficiency.
CDTs come in two variants: the stationary plasma thruster (SPT) (also known as the magnet layer thruster) and the anode layer thruster (TAL). The main difference between these two devices is that the SPT uses a dielectric material that usually contains boron nitride to electrically insulate its acceleration channel walls while the TAL uses a channel made out of metal. Performance characteristics of both engines are virtually identical. Although they vary in size and input power, CDTs that are currently being considered for NSSK (figure ES7) typically operate at discharge voltages of 300 to 500 V, and thruster currents between 4.5 and 20 A, with xenon mass flow rates of 5 to 25 mg/s.
Hall thruster research is now centered on increasing the specific impulse (to over 5,000 s) and efficiency with xenon and metal propellants (such as Bismuth), characterizing life for these high-specific impulse Hall thrusters, scaling up Hall thrusters to input power levels above 50 kW and operating multiple thrusters in a coordinated manner.
