Colloid Thrusters for space propulsion

[rklaehn]

I think given the spectacular advances in solar cells in the last years, the most efficient way to get around the inner solar system will be solar electric.

But one thing that is a big disadvantage of solar electric over other means of propulsion is that most common forms of electric propulsion require rare and exotic fuels such as Xenon. While it is possible to do ISRU with chemical propulsion (LOX/LH2 or LOX/Methane), this will never be possible with Xenon because it is extremely rare. So the ideal electric propulsion method for opening up space would be an electric propulsion method that can use common chemicals like water with decent Isp and efficiency. If you can combine electric propulsion with ISRU the entire inner solar system opens up!


It seems that in the sixties there was a large interest in colloid thrusters. A colloid thruster is a thruster that creates microscopic charged droplets and accelerates them via an electric field. Basically an ion engine, but with each charged particle being a droplet consisting of millions of molecules of a conductive liquid instead of a single atom or molecule.

A big advantage of a colloid thruster over an ion engine is that the ionization energy per mass is much lower, so most of the energy goes into acceleration of the droplet. Disadvantages that caused the interest in colloid thrusters to wane are the following:

- you only get a miniscule amount of thrust from each droplet emitter, so you need thousands or even millions of them to get to macroscopic (Newton) thrust levels. This of course was a large problem given the manufacturing techniques of the 60s and 70s.

- since the charge per mass is very low compared to ions, you need very large voltages to get decent velocities and thus Isp. Efficiently creating high voltage from low voltage DC such as what you get from solar cells was very difficult in the 70s.


But both of these issues should be solvable with modern technology. Creating a large number of droplet emitters should be relatively easy using semiconductor manufacturing technologies. And transforming DC voltages has also become relatively easy due to advances in semiconductor technology. High voltage DC is now routinely used for power transmission with miniscule losses.

So is anybody aware of studies or experiments for using colloid thrusters for main spacecraft propulsion? Or is there a fundamental reason this kind of propulsion can not be scaled up to Newton levels of thrust by using high voltages and large numbers of etched emitter nozzles?

From what I read, you should be able to freely choose the Isp of your engine in a wide range by just regulating the acceleration voltage (just like ion engines).

There seems to be some recent interest in using them for stationkeeping and formation flying applications, such as this:
infoscience.epfl.ch/record/.../IEPC-2009-189.pdf‎

Here are some lecture notes from the wikipedia article that describe the math behind colloid thrusters in great detail:
http://ocw.mit.edu/courses/aeronautics-and-astronautics/16-522-space-propulsion-spring-2004/lecture-notes/lecture23_25.pdf

[JasonAW3]

Why not use a Heliostat system to superheat a colloidal material to provide thrust in a similar but more direct way than Solar Electric?

     There's been talk of using Sterling Engines to produce power for space stations in the past, so why not look backwards to go forwards?

     The simpler a system is, the less likely it is to fail.

Jason

[rklaehn]

Why not use a Heliostat system to superheat a colloidal material to provide thrust in a similar but more direct way than Solar Electric?

There is no way to get good Isp with anything thermal. The only material that gets halfway decent Isp with solar thermal is hydrogen because of the lowest possible molecular weight. And even that will get you only 800 to 900s. Anything with higher molecular weight will get an Isp in or below the chemical range (<300s).

There's been talk of using Sterling Engines to produce power for space stations in the past, so why not look backwards to go forwards?

Current space solar cells have efficiencies above 30%, and that with no moving parts. A stirling generator would be very complex in comparison. There might be a case for stirling generators once you have to build the entire propulsion system from ISRU resources. But that is multiple decades off. For now, building the power supply on earth and using ISRU just for propellant would be a huge step forward.

The simpler a system is, the less likely it is to fail.

I agree. But for energy generation, nothing is simpler than a solar cell with no moving parts. And for the actual engine: there must be some way to create an electric engine that does not need exotic materials like xenon. Maybe you can just create charged dust from regolith and accelerate that?

[A_M_Swallow]

Hall thrusters can use Argon instead of Xenon.  Much cheaper.

As for using water as a propellant the CAT plasma thruster will be having a second go at raising funds in a few weeks time.
http://www.kickstarter.com/projects/597141632/cat-a-thruster-for-interplanetary-cubesats

[rklaehn]

Hall thrusters can use Argon instead of Xenon.  Much cheaper

But that reduces efficiency since ionization energy (which is lost) per mass increases. Still good though.

As for using water as a propellant the CAT plasma thruster will be having a second go at raising funds in a few weeks time.
http://www.kickstarter.com/projects/597141632/cat-a-thruster-for-interplanetary-cubesats

Interesting. I might chime in. That is still a thermal approach, so the velocity distribution of the exhaust will be non-uniform. And you will not get the ionization energy back.

With a colloid thruster there is negible ionization losses because the propellant is not ionized plasma, but just slightly charged macroscopic droplets. And if you can get the droplet size or at least charge per mass uniform, you should be able to get to really high efficiencies (approaching 1).

But as long as the efficiency is not ridiculously low it might not matter. With the power per weight of solar cells increasing at a rapid rate, things like power per weight for the engine might become just as important as efficiency.

[Solman]

 For power generation consider Spectrolab solar concentrator type PV which has been demonstrated to achieve efficiency over 40% at over 900 Suns. A large concentrator may be used but a radiator is required. However STR has higher overall efficiency in terms of sunlight into enthalpy in exhaust and can achieve over 1200 seconds with hydrogen.

[Jim]

     The simpler a system is, the less likely it is to fail.

And solar cells are the simpler system

[cordwainer]

What would the Isp for sodium chloride be, seems that would be cheap enough and possible exist in some quantity for ISRU?

[jongoff]

So is anybody aware of studies or experiments for using colloid thrusters for main spacecraft propulsion? Or is there a fundamental reason this kind of propulsion can not be scaled up to Newton levels of thrust by using high voltages and large numbers of etched emitter nozzles?

I know Busek, JPL, and MIT are all actively engaged in "Electrospray thrusters" (another name for Colloid thrusters). Busek was involved in the DARPA Phoenix program during Phase 1, but I don't yet know if they made the cut for Phase 2. They've developed some cubesat-scale electrospray systems under previous contracts, and I think are working on some larger scale systems. I'm personally pretty interested in them, because of their very good thrust/W and thrust/power ratios at decent Isp's.

That said, I don't know how good they are for ISRU. The ionic fluids they use for electrospray thrusters are some rather unusual molten salts. Great materials, but I'm not sure how feasible they'd be to produce from local resources.

~Jon

[JasonAW3]

Considering the usual make-up of comets, it shouldn't be to hard to mix something usable up.

Jason

[neviden]

SEP using water would be very good combination. Water is space storable, non-toxic, cheap, can also serve as radiation shield and is easy to get via ISRU.

How far advanced and how scalable is this thruster?

Alternative thrusters that can use water are Microwave Electro-thermal Thruster with 800 - 1000 s isp and maybe even VASIMR with even higher isp.

[cpooley]

For Microlaunchers (can google that) we plan for later versions either colloid, using glycerin, giving about 1500 Isp, or indium for Isp about 6000.  Spacecraft mass to be about 200 grams, so thrust will be small, and power level a few watts.

[rklaehn]

For power generation consider Spectrolab solar concentrator type PV which has been demonstrated to achieve efficiency over 40% at over 900 Suns. A large concentrator may be used but a radiator is required. However STR has higher overall efficiency in terms of sunlight into enthalpy in exhaust and can achieve over 1200 seconds with hydrogen.

Solar thermal requires very precise pointing of the mirror. And it only gives decent Isp with hydrogen, which is probably the most difficult to handle propellant. It might work at very large scale, but technologies that also work on a very small scale will evolve more quickly.

Solar electric works with dense, non-cryogenic liquids and requires only rough pointing. It also scales down to cubesat size. Also, an electric power system will be very useful once you reach your destination. So I think solar electric is more practical.

[jongoff]

Considering the usual make-up of comets, it shouldn't be to hard to mix something usable up.

There's a lot more to chemical engineering than just starting with the right mix of atoms...

The chemicals used for most colloid thrusters are very complex, and IIRC the path from the kind of raw constituents you could find on a comet to useable colloidal thruster propellant is likely really long, complex, and requiring lots of hardware.

Colloidal thrusters are neat for a lot of reasons--ISRU probably isn't one of them.

~Jon

[rklaehn]

So is anybody aware of studies or experiments for using colloid thrusters for main spacecraft propulsion? Or is there a fundamental reason this kind of propulsion can not be scaled up to Newton levels of thrust by using high voltages and large numbers of etched emitter nozzles?

I know Busek, JPL, and MIT are all actively engaged in "Electrospray thrusters" (another name for Colloid thrusters). Busek was involved in the DARPA Phoenix program during Phase 1, but I don't yet know if they made the cut for Phase 2. They've developed some cubesat-scale electrospray systems under previous contracts, and I think are working on some larger scale systems. I'm personally pretty interested in them, because of their very good thrust/W and thrust/power ratios at decent Isp's.

Great.

That said, I don't know how good they are for ISRU. The ionic fluids they use for electrospray thrusters are some rather unusual molten salts. Great materials, but I'm not sure how feasible they'd be to produce from local resources.

I think they use ionic liquids because they have low vapor pressure combined with good conductivity. This is so you can get very small droplets with high charge per mass (e/m) so you don't need a high voltage to accelerate them to an useful velocity. High voltage equipment does not scale down very well to cubesat size. But the basic mechanism that produces droplets of uniform size and charge works for any conductive liquid.

(colloid thrusters have the exact opposite problem of normal ion thrusters: in ion thrusters you try to minimize e/m by using high atomic mass atoms or molecules to minimize ionization loss per mass. In colloid thrusters you want droplets with as high an e/m as possible so you don't need too much voltage)

But for a larger system where ISRU makes sense, needing high voltages will not be as much of a problem. High power/high efficiency DC/DC converters to high voltage are commercial off the shelf for DC power transmission. So I think for a very large "charged droplet" thruster it would be acceptable to have bigger droplets with smaller e/m, so any slightly conductive liquid like salty water would do just fine. You will have to increase the length of the "droplet accelerator" to avoid arcing, but there is no lack of space in space :-)

By the way: there is a huge number of technologies that are conceptually simple, but require conditions that are very expensive to reproduce on earth. We will only start making progress in these areas once we can cheaply do experiments in space. Not ISS style experiments that are planned a decade in advance, but experiments more like "I wonder what happens if I do ...".

[rklaehn]

For Microlaunchers (can google that) we plan for later versions either colloid, using glycerin, giving about 1500 Isp, or indium for Isp about 6000.  Spacecraft mass to be about 200 grams, so thrust will be small, and power level a few watts.

That's even smaller than the currently commercially available cubesat propulsion systems. Do you plan to develop those yourself?