ELECTRIC ROCKET ENGINE BASICS: What the different types of electric rocket engines look like and how they work.

(Return to the main page for more medical topics or to browse 80 other subjects: everything from kaleidoscopes and metal detectors to the strange world of lucid dreaming.)


Chemical rocket engines, like those on the space shuttle, work by burning two gases to create heat, which causes the gases to expand and exit the engine through a nozzle. In so doing they create the thrust that lifts the shuttle into orbit. Smaller chemical engines are used to change orbits or to keep satellites in a particular orbit.

For getting to very distant parts of the solar system chemical engines have the drawback in that it takes an enormous amount of fuel to deliver the payload. Consider the Saturn V rocket that put men on the moon: 5,000,000 pounds of it's total take off weight of 6,000,000 pounds was fuel.

Electric rocket engines use less fuel than chemical engines and therefore hold the potential for accomplishing missions that are impossible for chemical systems. To understand how, we have to understand a number called specific impulse.

Take one pound of the hydrogen/oxygen propellants used by the shuttle's main engines and burn it in one second and you'd generate 370 pounds of thrust. That "370" is a measure for the combined efficiency of the engines and the propellants burned in them. If we could discover a different chemical combination that produced twice as much thrust for the same amount of fuel, it would have a specific impulse of 740. The units of specific impulse are "seconds." The advantage of having a higher specific impulse engine is that you need considerably less fuel to accomplish the mission. Take the Saturn V rocket, if it had a specific impulse of 700 seconds and could still produce the same level of thrust it would only need 2,500,000 pounds of propellant. Take that much weight off it and the structure could be made lighter, which would mean that it would need even less fuel, and so on. Reducing the weight of the fuel by increasing the specific impulse is one of the most powerful ways of reducing the overall weight of a spaceship.

The problem is that all the energy for chemical engines comes from the energy stored in the propellants. There is a limit to this and we've already pretty much reached it with the shuttle engines. The way around this problem is simple: we cheat.

Instead of relying only on the energy stored in the propellants, we can add energy using electricity. In the simplest concept an electric heater is used to increase the temperature of the propellant above what it could get through combustion. The higher the temperature the greater the expansion and the more thrust per pound of propellant is obtained. Because the energy no longer comes from the propellants, we also now have a much wider range of propellant options since they no longer have to be able to combust.

Sounds great but there are three problems. First, you need a power supply to provide the electricity and this adds weight to the rocket ship. (If the specific impulse is high enough the propellant needed is reduced so much that even with the power supply weight added in the total ship weight is reduced.) Second, because space power systems only put out small amounts of power compared to chemical engines, the amount of thrust electric rocket engines produce is very small: on the order of pounds or even fractions of a pound instead of the tens of thousands to millions of pounds of thrust chemical engines produce. (Electric engines make up for this by running for months or years instead of minutes like chemical engines. In so doing the total impulse (amount of thrust multiplied by the time the engine burns) can actually be greater.) Third, electric rocket engines only work in the vacuum of space because atmospheric pressures hinder the physical processes by which they create thrust.

In spite of these problems, millions of dollars are being invested every year to develop electric rocket engines because in the long run they are destined to play a major roll in humankind's conquest of space. In fact, dozens of small electric thruster are already in space performing many low-thrust missions much better than their chemical counterparts.

As the specific impulse provided by an engine increases, the mass of propellant needed decreases for a specific mission. Going hand in hand with this is the fact that as specific impulse increases so does the amount of electrical power needed to run the engines (at the same thrust) so the mass of the power supply increases, offsetting the savings realized from reducing the propellant. Plotting the combined propellant and power supply mass results in a graph that looks like a "U," where the bottom-most part of the "U" is the optimal point where total spacecraft weight is at a minimum. Spacecraft designers use this graph to determine the optimum specific impulse and power for the mission in question. While the number varies with thruster efficiency, power supply density and so on, as a general rule of thumb you want specific impulses on the order of 300-1,000 sec for orbit maintenance missions, 1,000-2,000 sec for orbit transfers, and 2,000-6,000 sec for interplanetary missions.

So, what do these electric rocket engines look like and how do they work? The following simple diagrams and explanations will help introduce you to this interesting field:

Electrothermal Rocket Engines:

This class of electric rocket engine works by heating a propellant.


A resistojet simply uses electricity passing through a resistive conductor, something like the wires in your toaster, to heat a gas as it passes over the conductor. As the conductor heats up the gas is heated, expands, exits through a nozzle and creates thrust.

diagram of a resistojet

In real resistojets the conductor is a coiled tube through which the propellant flows. This is done to get maximum heat transfer from the conductor to the propellant.

Almost any gas and even some liquids can be used as fuel, the most common being hydrazine (N2H4). Hydrogen, nitrogen, ammonia and many other fuels have also been used. For a given engine and power level, the lighter the propellant the higher the specific impulse and the lower the thrust. Hydrogen produces very high specific impulses (as high as 400 sec.) This may not sound very high but resistojets are designed for small-thrust missions like orbital station keeping and the best chemical engines in this range only have specific impulses of 200 sec or less. They use so little power, 350 watts or less and then only intermittently, that they can operate using residual electrical power already available on the satellite. Because of their increased specific impulse they need hundreds of pounds of fuel less than the next best chemical engine. That's weight that can be used for more propellant, so the satellite can remain on orbit years longer, or for extra payload. Resistojets produce thrusts on the order of small fractions of a pound.

a space-qualified hydrazine resistojet

As with any electrical device, resistojets are not perfectly efficient. Typically they convert 50 percent of the electric energy passed through them into thrust energy. (Don't sneer at this... the engine in your car is at best only 25 percent efficient.)

Resistojets can be scaled down to very small sizes, anywhere from ones that fit in shoe boxes to others as small as a thimble, making it easy to tailor them to any low thrust mission. Dozens of them are currently in orbit helping satellites maintain their orbits. Next time you're watching the Superbowl you can probably than a resistojet for keeping the satellite transmitting the signal to be where it should be for you to pick it up.

The problem with resistojets is that the physical limitations of the conductor means that the maximum temperature they can achieve is 1800 degrees C. Run them hotter than this and they start to melt. (Mission directors hate it when that happens.) Fortunately, there's a solution: the arcjet.


An arcjet is simply a resistojet where instead of passing the gas through a heating coil it's passed through an electric arc.

diagram of an arcjet

Because arcs can achieve temperatures of 15,00 degrees C. this means the propellant gets heated to much higher temperatures (typically 3,000 degrees C.) than in resistojets and in so doing achieve higher specific impulses, anywhere from 800 sec for ammonia to 2,000 seconds for hydrogen. Arcjets tend to be higher power devices, typically 1 to 2 kilowatts, and used for higher thrust applications, like station keeping of large satellites. Several are currently in orbit.

a small space-qualified arcjet


hydrogen arcjet firing


ammonium arcjet firing

The largest arcjet used in space was a 26 kilowatt engine operating on ammonia with a specific impulse of 800 sec. It was part of the USAF's ESEX space experiment program.

Arcjets can run at up to 35 percent efficiency.

Two problems hounds arcjets: the electrodes run glowing hot causing erosion and this heat can get conducted to the spacecraft heating it to unacceptable levels. For station keeping missions they aren't on long enough for the heating to be a serious problem. But it could be for large engines designed to operate for long periods of time. Arcjets don't scale down as easily as resistojets and the smallest are small shoe-box affairs.

Electrodeless Electrothermal Engines:

There are several variants of the resistojet or arcjet engine that use microwaves or some sort of inductive coupling to heat the propellant. They have some advantage in that power can be coupled directly to the propellant without having to heat part of the engine. They suffer the disadvantage that this power coupling may be inefficient and the the microwave or inductive power itself may be inefficient to produce. There are working models of these in laboratories but none have been used to support and actual mission.

Electrostatic Rocket Engines:

Ion engines:

Rub a balloon against your hair or shirt and then hold it near your arm, the hairs on your arm will feel tingly and be attracted to the balloon. Bring the balloon near the carpet and bits of lint will be pulled to it. What's happening is that electrons have been deposited onto or removed from the balloon depending on what it was rubbed against, giving it an electrostatic charge, which creates an electrostatic field. A similar field can be used to produce thrust in a rocket engine called an ion thruster.

ion engine diagram

As propellant enters the ionization chamber (the small ns on the left), electrons (small -s in the middle) emitted from the central hot cathode and attracted to the outer anode collide with them knocking an electron off and causing the atoms of the propellant to become ionized (+s on the right). This means that they have an electric field around them like the balloon. As these ions drift between two screens at the right hand side of the ionization chamber, the strong electric field of the "+" side repels them and the "-" side attracts them, accelerating them to very high velocities. The ions leave the engine and since the engine pushes on them to accelerate them, they in turn push back against the engine creating thrust. Ion thrusters typically use Xenon (A very heavy, inert gas) for propellant, have specific impulses in the 3,000 to 6,000 range and efficiencies up to 60 percent. An average thruster is one to two feet in diameter, produces thrust on the order of small fractions of a pound and weighs some tens of pounds.

a typical ion engine

Downstream of the exhaust is a hot cathode emitter that injects electrons into the exhaust stream. Without this, the exiting ions would slowly cause a charge to build up in the spacecraft that could interfere with its operation and create a pull on the ions that would reduce the thrust. You can see the small electron emitter in the upper right corner of the picture above.

an ion engine firing

Ion thrusters are well developed and have been used on a few space missions, such as a comet encounter. With their high specific impulses they are well suited to deep space types of missions.

Colloid Thruster:

A colloid is a microdroplet like inkjet printers use to spray their ink on paper. Given an electrical charge, these microdroplets, or colloids, can be accelerated in a thruster similar to an ion thruster. The advantage of a colloid thruster is that because the individual particles being accelerated are so much larger and heavier than the atoms in a regular ion engine, the specific impulse can be lowered and thrust increased to make a better fit for a particular mission. Also, the variety of propellants that can be used is much greater. Although colloid thrusters have been around almost as long as ion engines they have not been developed to flight status. In the laboratory they typically have specific impulses around 1,000 sec.

diagram of a single colloid thruster emitter
Many of these would be tied together to create a single large thruster.

Electromagnetic Rocket Engines:

This is by far the largest group of electric thrusters with many different techniques used to create thrust. As widely divergent as these thrusters may seem they all use the same principle: the Lorentz force.

If you have an electric current flowing perpendicular to a magnetic field, the magnetic field will push against the current. If the current is flowing through a solid conductor or even a gas the gas will be pushed out as well. This is the Lorentz force. Mount such a device on a space ship and you have an electric rocket engine.

Rail guns:

Drive enough current down one side of two parallel conducting rails, across a conductor that can slide down the rails while maintaining contact, and up the other side and you have a rail gun. The current flowing in the rails create a strong magnetic field between them. The same current flowing across the sliding conductor is the current the magnetic field wants to push away.

Pump enough power into the thing and the sliding conductor will be accelerated to thousands of feet per second. Mount the beast on a tank and you you have an electric cannon. Put it on a spacecraft and you have a rocket engine. You'll need to added something like a machine gun feeder to supply it with a constant source of sliding projectiles (which we should now call "propellant" since we're using it as a rocket engine) but that's a simple mechanical engineering problem.

a weapon-type rail gun firing

While such a propulsion system would work in principle, the length of the rails makes a practical application of this concept difficult to imagine. Also, as the sliding conductor accelerates down the rails contact friction and erosion from arcing between the contacts erodes the rails. For an engine such a device would have to fire repeatedly over a long period of time and this erosion could be a problem.

Magnetoplasmadynamic (MPD) Thrusters:

Crank the power up on the rail gun high enough and the sliding conductor would vaporize. The engine would still work because the plasma would continue conducting current and be blown out the end of the gun. Flatten and bend one of the rails around in a tube surrounding the other rail and you have an MPD thruster. (For further details about MPD thrusters please check out the MY ELECTRIC ROCKET ENGINE page on this site.)

MPD thruster diagram

MPD thrusters are unique among the electric rocket engine fraternity because they are capable of producing thrusts as high as 50 pounds in an engine small enough to fix in a large shoe box. The problem with them is the electrodes wear out from handling all the current and they eat up enormous amounts of power: on the order of megawatts. There is currently no space power system that comes even close to this level. Typical performances numbers are 30 percent efficiency at 2,500 sec specific impulse. In laboratories they usually run on argon, but anything that can be pumped into them can be used. (When I was working in the lab I always wanted to run one of vaporized sodium metal.) Using hydrogen would push the specific impulse into the 15,000 sec or higher range.

MPD thruster firing

Because of their compact size and potential for high thrust MPD thrusters are one of the few viable options for primary propulsion on high-mass, deep space missions. I like to think of them as being the progenitors of the impulse engines of Star Trek fame.


Hall Thrusters:

These engines are popular with the Russians and over 100 have been used on their space missions.

diagram of a Hall thruster

Electromagnets around the outside cylinder and inside core create a magnetic field pointing radially inward. The interplay of this magnetic field and the electric field between the anode propellant injectors and the electron cloud created outside of the thruster causes a current (called the Hall current) to be induced to flow azimuthally around the open annulus in the thruster. The magnetic field pushes on the current and accelerates it, and the gas it's traveling through, out of the thruster to create thrust.

Picture of a Hall thruster.
Note the electron gun mounted on top of the thruster.

Typical thrusters in US laboratories are a foot or so across, use 1 to 5 kilowatts of power, operate at 2,200 sec specific impulse, produce less than one pound of thrust, and are 50 to 60 percent efficient. They are noted for their durability.

Hall thruster firing

New designs using vaporized bismuth can have efficiencies as high as 70 percent making them the efficiency rulers in the electric propulsion world.

Hall thrusters come in two main variants: the Stationary Plasma Thruster (SPT) and the Thruster with Anode Layer (TAL). The SPT has insulating walls on the acceleration chamber and is longer, the TAL has conducting material lining the walls and is shorter.


Pulsed Inductive Thrusters:

Imagine an electromagnet sitting on its end on a table. Now place a metal ring on top of it. Pulse current through the coil and a second current will be induced to flow through the ring. This induced current will flow around the ring in the direction opposite to the coil. We now have another case of a current (flowing in the ring) moving perpendicular to a magnetic field (created by the coil and directed along the coil's axis) so the Lorentz law tells us there will be a force on the ring wanting to push it away. If enough current is forced through the coil the ring will shoot straight up into the air. If the current in the coil is high enough and increases fast enough, the ring will be vaporized and ionized. Even in the gaseous state it'll still conduct the current and be accelerated away from the coil. That's what a pulsed inductive thruster is.

Pulsed Inductive Thruster

The only difference between actual PIT thrusters and the coil analogy is that a special valve and nozzle unit mounted in the center of the coil directs a short pulse of gas down to cover the face of the coil. The current pulse through the coil is synchronized with the gas pulse so that the gas is ionized and accelerated away, creating thrust, before it can dissipate into the vacuum of space.

Pulsed inductive thrusters, are big, beautiful, sexy looking thrusters up to a meter in diameter. They operate in a pulsed mode at up to 1,000 pulses per second with specific impulses between 2,000 and 5,000 sec and thrusts of fractions of a pound to tens of pounds. Efficiencies can be as high as 50 percent.

Although the inductive coupling between the engine and plasma should imply that there is no erosion as there is in the MPD thruster, at high pulse rates the large surface area of the engine will be exposed to the thermal loading of having a virtually constant plasma mere centimeters from it. The coils could heat to the point where increased electrical resistance could cause problems or even melting. These engines are also power gluttons, eating up tens of kilowatts to megawatts of electricity.


Pulsed Plasma Thrusters:

These small electric thrusters have been around for decades and have flown on many space missions performing station keeping functions.

Pulsed Plasma Thruster diagram

In the pulsed plasma thruster, a bar of solid propellant (could be anything but Teflon is the usual fuel of choice) is spring loaded against two stops near the exit of the thruster. When it's desired to fire it, a energy storage unit discharges an arc across the face of the propellant, ablating a small amount of the Teflon bar. Just like the rail gun and magnetoplasmadynamic thruster, the current flowing through the vaporized propellant ionizes it and reacts with the magnetic field created by the current to accelerate the propellant out of the engine, creating thrust. As the propellant bar is eroded, the spring pushes it forward for the next pulse.

Pulsed Plasma Thruster


Pulsed Plasma thruster firing


These engines are extremely simple, reliable, and robust. They have to be operated in the pulsed mode but can be pulsed rapidly to provide almost continuous thrust. They typically use 30 watts or less power, have efficiencies around 30 percent, specific impulses of 1,000 seconds, and thrust levels measured in micropounds to millipounds.


Field Emission Electric Propulsion (FEEP) Thrusters:

These are extremely small thrusters that operate somewhat like a colloid thruster in that they have a sharp propellant emitters. The difference is that in the FEEP the emitter is so small that individual ions are pulled from the emitter instead of droplets. Also, ionization occurs as a side effect of the emission process so an ionization chamber isn't required.

Field Emission Electric Propulsion thruster

Because the emitter hole or slit is so small, only 0.001 millimeters across, capillary action both draws the liquid propellant into it and prevents it from exhausting into space, therefore a valve is not required.

FEEPs typically have specific impulses from 6,000 to 12,000 seconds and use melted indium as a propellant.


Mass Drivers:

A mass driver is similar to a rail gun in that it accelerates a solid projectile down a long runway. The difference is that a mass driver uses pulsed electromagnets lined up down the length of the runway to pull on the projectile, accelerating it to high speeds and thereby generating thrust.

Mass drivers are attractive because they can be extremely efficient (as high as 95 percent) run cool and can be designed so there is no guide rail wear. The problem is that they are so long, hundreds to thousands of feet, and heavy that it's doubtful they could ever be used for propulsion. But, if someone could develop a lightweight, high temperature superconductor they may yet have a place in the electric propulsion pantheon.


Electric Propulsion Research Focal Points:

I recommend that anyone wanting more detailed and up-to-date information on electric propulsion contact someone at one or more of the following organizations that are active in electric propulsion research:

University of Michigan (http://www.engin.umich.edu/dept/aero/spacelab)

University of Colorado

Princeton University: the School of Mechanical and Aeronautical Engineering

TRW, Los Angeles, CA

Air Force Rocket Propulsion Laboratory, Edwards AFB, CA (Electric Propulsion Laboratory)

JPL Pasadena, CA

Marshall Space Flight Center

Northrop Grumman Space Technology, Redondo Beach, CA

University of Dayton, OH

Besides the United States and Russia; Japan, Germany and Italy are active in electric propulsion research.


In Closing:

Hardly a year goes by in the electric propulsion world without someone thinking up a new concept. These are always variations of the thrusters outlined in this page and attempt to get around one problem or another through an innovative geometry, ionization scheme, or other concept. It would be impossible to chronicle all of them but I hope the thrusters that have been represented on this page provide a basic understanding of the world of electric propulsion.

The explanations of the thrusters on this page are oversimplifications of what are in fact extremely complex devices. It takes a PhD and many years of working with these devices to understand them at the current state of the art.

The bible for electric propulsion is the text: The Physics of Electric Propulsion by Dr. Robert Jahn.

Electric propulsion research is extremely expensive. While many of the thrusters can be manufactured for a few thousands of dollars, the enormous vacuum chamber required to test one of them can easily top $1,000,000 to build and hundreds of thousands of dollars a year to operate.


(Return to the main page for more medical topics or to browse 80 other subjects: everything from kaleidoscopes and metal detectors to the strange world of lucid dreaming.)