Tuning Up Ion Propulsion

by Paul Gilster on August 5, 2009

A story on MIT’s Technology Review site looks at ion propulsion, and specifically at improvements made in the technology at Glenn Research Center. Comparing the recent work to the engines used in the Deep Space 1 and Dawn missions, the story quotes GRC’s Michael Patterson as saying, “We made it physically bigger, but lighter, reduced the system’s complexity to extend its lifetime, and, overall, improved its efficiency.”


That’s good news, of course, and Patterson presented it to the AIAA’s Joint Propulsion Conference & Exhibit this week in Denver. With sessions on everything from Electric Propulsion Thruster Wear and Life Assessment to Advanced Propulsion Concepts, Denver was clearly the place to be for propulsion mavens. An entire session was devoted to the new ion thrust work, which goes under the name NASA’s Evolutionary Xenon Thruster (NEXT).

Quoting from an abstract of one of the talks:

The NASA NEXT thruster is engineered to be extremely flexible in terms of input power and specific impulse, while maintaining acceptable efficiency, and embodies a number of technological advances over previous ion engine systems.

Image: Follow the blue light. The xenon ion engine in this photograph was being tested in a vacuum chamber at the Jet Propulsion Laboratory. Credit: NASA/JPL.

Deep Space 1, launched in 1998, used an ion propulsion system that proved workable even if such engines produce only a tiny thrust. A continuous small push adds up, so one proof of concept was to keep Deep Space 1’s engine burning, which it did for 677 days (a spare of the same engine would later run for almost five years in NASA tests). The system used xenon gas released into a thrust chamber surrounded by magnets. The gas was ionized and expelled from the back of the chamber by electrically charged grids, demonstrating a method that, compared to chemical engines, yields ten times the thrust for the same amount of fuel.

So an ion engine is efficient, but the thrust is small. Used properly, however, the advantages of these systems are evident. Here’s an excerpt from a NASA description of the basic ion system, not yet amended to include the most recent NEXT work:

Modern ion thrusters are capable of propelling a spacecraft up to 90,000 meters per second (over 200,000 miles per hour (mph). To put that into perspective, the space shuttle is capable of a top speed of around 18,000 mph. The tradeoff for this high top speed is low thrust (or low acceleration). Thrust is the force that the thruster applies to the spacecraft. Modern ion thrusters can deliver up to 0.5 Newtons (0.1 pounds) of thrust, which is equivalent to the force you would feel by holding nine U.S. quarters in your hand. To compensate for low thrust, the ion thruster must be operated for a long time for the spacecraft to reach its top speed.

Deep Space 1 was only one of many ion thrust missions; in fact, testing on the basic concepts was begun in the 1970s. Since then, we’ve seen missions like Artemis (ESA), Hayabusa (an asteroid rendezvous mission by JAXA, the Japanese space agency), and Smart 1 (an ESA moon mission). The Dawn mission uses ion thrusters in its mission to the asteroids, while several other ion missions are now in progress.

Like Deep Space 1’s basic design, the improved NEXT engine also works with xenon gas. Here’s Brittany Sauser’s description from the Technology Review article:

The new ion engine builds upon the electric propulsion systems used by both DS1 and Dawn… It uses the same method to achieve thrust: xenon gas flows into a reaction chamber inside the engine and is ionized by electrons; electromagnets positioned around this chamber enhance the efficiency of ionization. Electrodes positioned near the engine’s thrusters (known as ion optics) are then used to accelerate the ions electrostatically and shoot them out of the exhaust to push the spacecraft forward.

Compared to earlier iterations, however, NEXT offers a larger throttling dynamic range, meaning it can operate for longer periods of time by adjusting power levels along the way. But the choice of destination says much about the problems these technologies face. Solar energy can power up an ion engine in the inner system, but it will take nuclear sources to keep one operational at the distance of the outer planets. The efficiencies involved in ion engines compared to the chemical alternative make resolving nuclear safety concerns an imperative.


Tulse August 5, 2009 at 10:36

How do these ion engines compare to VASIMR in terms of thrust, efficiency, and top speed? My impression was that VASIMR solves a lot of the problems of ion engines (such as grid erosion), and has the benefit of using readily-available hydrogen for propellant, thus making it in principle more easily refueled from non-terrestrial sources.

Administrator August 5, 2009 at 12:57

I don’t have a straight-up comparison available on NEXT vs VASIMR, but we still have to wait with the latter to see if the optimistic projections of the VASIMR team work out in practice. That may take some time, I think.

philw1776 August 5, 2009 at 14:03

Ion engines offer a couple orders of magnitude less thrust than VASIMIR. Wimpy. Ion offers higher ISP and has the advantage of actually having flown in space. My money is on the higher thrust, variable thrust, variable ISP VASIMIR in the long run (decades) once its proven, hopefully starting with an ISS deployment around 2012.

James M. Essig August 5, 2009 at 14:57

Hi Paul and Tulse;

The beautiful thing about ion engines is that the effective rest mass specific impulse of the ionic reaction mass is limited only by the exhaust velocity of the ions, or more specifically, the relativistic kinetic energy and momentum of the ions.

Some sort of matter antimatter reaction powered ion rocket might be ideal if the pair annihilation energy to thrust stream energy conversion efficiency could be made close to 100 percent.

Ion rockets could enable space craft, including manned starships, to reach extreme gamma factors, especially if some sort of zero point field energy extraction technology could be utilized. As we learn more about the structure of the vacuum, and the zero point field dynamics, perhaps such energy extraction can be made to occur.

Even if it is a very long time before any form of warp drive, wormhole transport, or teleportation technologies could be developed, perhaps we can find power sources with appropriately high mass specific power densities that could drive our star ships to gamma factors limited only by the ion rocket power.

By the way, Paul, I really enjoy the photo of the blue exhaust from the ion engine. It reminds me of the exhaust streams of the Star Destroyer ships of the Star Wars movie “The Empire Strikes Back”.

Enzo August 5, 2009 at 17:45

I remember reading that the now defunct JIMO mission would have used up nearly all the known reserves of xenon gas.
This seems like one more reason to use more ordinary substances like VASIMIR does. I also like the fact that it doesn’t have the grid erosion problem : it would be great to have a design where a probe/starship can return, be refueled and reused.
I understand that VASIMIR efficiency is not yet at the same level as standard ionic propellers but that might improve.

VASIMIR powered by nuclear reactor seems to be one of the best hopes for the (sort of) near future and exit the chemical-rocket-only logic .
Otherwise solar powered ionic propulsion as efficient as it might be, it is excruciatingly slow.

Wedge August 5, 2009 at 19:38

I should point out that SMART-1 actually used a Hall effect thruster, which uses a magnetic field (rather than an electrostatic grid) to accelerate its ionized exhaust. Hall thrusters have several advantages over ion engines, not the least of which is that there are less downsides to using alternative propellants (such as Mercury or Bismuth). In ion engines, non-noble-gas propellant increases grid erosion.

VASIMR is a good concept but still has much to prove operationally.

I suspect that the field of propulsion technologies which will become the backbone of future interplanetary transportation will probably be surprisingly different from what we might predict today. I suspect “sail” technologies such as solar sails, mag sails, and m2p2 will be heavily used to move less time sensitive payloads (cargo, unmanned research instruments, etc.) And I wonder if the most heavily used propulsion technology for moving humans around the Solar System will be chosen based on constraints we don’t fully appreciate now. It’s hard to say exactly what the economics of scale will be for propulsion and propellant technologies in the context of an industrial base that spans across planets, asteroids, moons, etc. Perhaps it will be fusion rockets, perhaps it will be VASIMR, perhaps it will be fission (gas-core NTR, ORION, NSWR), or perhaps it will be something nobody has yet invented yet. Hall thrusters were invented within the last 50 years, VASIMR 30 years, NSWR 20 years, who knows what else will be invented in the next 20 or 50 years, especially once off-Earth populations and industry start to grow.

Adam August 5, 2009 at 23:08

Hi All

Ion-drives are limited in their ultimate Isp by breakdown voltages in the acceleration grids. In otherwords at too high power and voltage they start short-circuiting. But that’s gridded ion-engines. Ungridded systems may not be so limited. According to a study by Geoff Landis externally powered ion-drive star-probes can be very efficient even at high speeds (I misread the paper when I said here previously it was more efficient than laser-sails only below 0.2 c. In fact the two can work together synergistically even at high speeds.

Personally I think VASIMR is better for variable Isp missions, but ion drives have proven reliability. VASIMR potentially evolves smoothly into true fusion rockets as will be required for Outer Planet missions with travel times of less than ~1 year. Neither system has to be in competition with the other at our current level of engineering ignorance.

Paul Titze August 5, 2009 at 23:18

Hi Paul,

Noticed there was a Breakthrough Propulsion Physics part of the AIAA conference chaired by Marc, anything interesting to report from this?

Cheers, Paul.

Administrator August 6, 2009 at 9:13

Paul, Marc wasn’t at the Denver conference after all. He tells me, though, that the breakthrough propulsion session didn’t have any really new material, but was basically updates on existing ideas. I’ll hope to have more later.

ljk August 6, 2009 at 10:23

Just remember the most important thing about designing a starship and
its means of propulsion: It has to look cool. :^)

Tulse August 6, 2009 at 10:54

Agreed, ljk, which is another point in favour of these kind of drives — a ghostly blue glow at the exhaust end of an engine seems pretty cool to me. Heck, it’s practically the standard for movie/TV science fiction starship engines (Star Trek being a notable exception).

george scaglione August 6, 2009 at 14:02

ljk,tulse, i am glad that you have come to your senses! what could be more important than a starship looking cool!? or,lol,are our current comments somewhat,no pun intended,”warped”! and,paul (!) “only” updates of existing ideas??!! i sure hope you will fill us in on THOSE soon i can’t wait to hear! talk to you all soon,i hope! all the very best! george

Adam August 7, 2009 at 0:08

Since all current starships are “spaceships of the mind” then cool is definitely a factor.

Brian Koester August 16, 2009 at 16:34

@James Essig

Centauri Dreams Readers – Here is your moment of ZEN…

So how visionary was George Lucas in all this? I mean you all realize that TIE fighter stands fro “Twin Ion Engine” …….I know! ….. ;-)

george scaglione August 17, 2009 at 19:52

jim,yes but… unfortunately (and so help me i am not knocking you) sf does indeed tend to have “all” the answers.while we here must just await breakthroughs etc lol not that i have not had more than my fair share of admiring sf and wondering – gosh! i sure wish we could do THAT! everybody that knows me realizes it i am sure,no doubt! anyway i wish we would spend alot more time developing new and better means of propulsion.now,THAT ,could be a real help! i sure wish that nasa’s breakthrough propulsion physics pro gram could be re instated ! at present i am mainly pushing among my friends for a workable fusion engine.now that would be able to get us out into the solar system bigtime the true starting point to journeys to the stars! sincerely wonder if our good friend marc millis would be interested in trying that hat on for size once again.by the way i was very impressed the other day on the history channel i saw a pro gram about the physics of star trek in which he spoke.did not recall seeing him on tv before .it was really cool . but anyway all thanks alot your friend george

g.h.geary September 3, 2009 at 19:24

I am just a layperson, so this may sound stupid, but can such a system actually lift off the planet or is it for interspace travel only?

Administrator September 3, 2009 at 21:39

g.h. leary, it’s not a stupid question at all. The answer is that ion propulsion doesn’t produce high thrust and thus isn’t suitable for liftoff from a planetary surface. But because it can produce a continuous (tiny) thrust over long periods, it can get a spacecraft up to high velocities over time.

g.h.geary September 7, 2009 at 15:55

Thank you. So is there anything in the works, even at a theoretical level, that can lift astronauts off Earth without strapping a bomb between their knees? Or do we just not know enough about gravity yet?

Administrator September 7, 2009 at 16:04

Work continues on various options, but the problem of getting astronauts into low-Earth orbit is complicated by the need for high thrust. A nuclear design like NERVA could use nuclear power to heat a propellant and do the job far more efficiently than a chemical rocket. We don’t yet have a good way to come up with high thrust, though, without some pretty lively rocketry. Some kind of breakthrough in understanding gravity that could help us counter its effects would certainly be welcome, but we’re a long way from that.

buoy December 11, 2009 at 14:55

g.h.geary, there are a few ways people are tossing around at conferences etc in regards to getting things into various orbits from low earth to moon. the ones that strike me the most are the space fountain and launch loop concepts. if you google those terms you’ll find useful information within the first few links surely.

Rohan November 21, 2010 at 8:11

why is it that these engines do not use hydrogen as the propellant?
the exhaust velocity is dependent only on the charge to mass ratio (for a given potential drop) and hydrogen is the lightest element. therefore, using hydrogen would increase the Isp.
Also hydrogen has a comparable 1st ionization energy (13.5eV) as compared to xenon’s ( 12.1eV).
also, why is there a need for a decelerator grid in such engines, if the exhaust velocities being produced are too high, we could always operate the engine intermittently.

Comments on this entry are closed.