Alternative propellants for Nuclear Thermal Rockets

[redliox]

Although I doubt NTRs will materialize quickly, if or when they do they are almost a simpler setup than chemical rockets, excluding of course the shielding from both radiation and heat.  One reason I say they are simpler is that there is no combustion involved; the propellant going in is the same chemical going out, just very hot and accompanied by the sporadic neutron and loose uranium atom.  A thought had occurred to me...

While hydrogen is ideal because it is light, storing it rarely is nor is manufacturing it (mostly because of the former reason).  However, why not use another raw material in its place?  I would exclude methane because it isn't naturally abundant and still requires manufacturing.  Water and carbon dioxide come to mind since they are abundant on Mars, Venus, and the poles of Mercury and Luna.  Perhaps nitrogen too, although its sources would be cheifly the Earth and Titan (although extracting it from Venus might be more viable than Mars).

I have started reading into Zubrin's salt-water NTR which seems to be up this alley, although I don't know enough on it.

How can we calculate how much an alternative to hydrogen might work in place of hydrogen?  I presume the following propellants:
Cardon Dioxide (CO2)
Water (H2O)
Nitrogen (N2)

[KelvinZero]

I think oxygen is often suggested because it can be produced on the moon.. more to augment the thrust at the expense of lower ISP http://www.projectrho.com/public_html/rocket/enginelist.php#lantr

As well as purely NTR rockets, perhaps there is something technically more like an ion rocket but more direct than a nuclear engine creating electricity that powers an ion drive.. or a nuclear pumped laser.. because the heat from that could be concentrated to a point hotter than the fissioning materials, perhaps. This might escape the temperature limits that make hydrogen the most useful fuel.. because you can't go hotter than what will melt your mechanical bits, especially whatever contains the fissioning materials.

[MATTBLAK]

In Zubrin's 'Case For Mars' book he talks about a nuclear-thermal propulsion system named 'NIMF' for Nuclear Indigenous Martian Fuel. The reactor powers pumps that suck in and compress CO2 for storage in propellant tanks. Then use the liquid CO2 as a monopropellant. I don't remember the specific impulse but I Googled it and found a link to the paper on it:

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19920001880.pdf

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19910012833.pdf

280 seconds Isp? Not great - but still a cool idea and the specific impulse could probably be tweaked. I wonder how NIMF would work with nitrogen or ammonia? If NIMF used carbon monoxide; the engine could possibly use liquid oxygen injected like an 'afterburner'

[redliox]

I think oxygen is often suggested because it can be produced on the moon.. more to augment the thrust at the expense of lower ISP http://www.projectrho.com/public_html/rocket/enginelist.php#lantr

I had been told oxygen, if ionized, can be very nasty to deal with.  Would a NTR heat the oxygen enough to worry about ionization?  Otherwise I'd agree it'd be a nice viable option.

[redliox]

In Zubrin's 'Case For Mars' book he talks about a nuclear-thermal propulsion system named 'NIMF' for Nuclear Indigenous Martian Fuel. The reactor powers pumps that suck in and compress CO2 for storage in propellant tanks. Then use the liquid CO2 as a monopropellant. I don't remember the specific impulse but I Googled it and found a link to the paper on it:

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19920001880.pdf

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19910012833.pdf

280 seconds Isp? Not great - but still a cool idea and the specific impulse could probably be tweaked. I wonder how NIMF would work with nitrogen or ammonia? If NIMF used carbon monoxide; the engine could possibly use liquid oxygen injected like an 'afterburner'

Thanks for finding this! 

It looks like water is the best option for something purely indigenous while methane (if you're willing to manufacture of course) is very useful at 670 second.

[hkultala]

The whole idea of NTR is based on idea of the nuclear reactor allowing propellant with low molecular mass. Without such, it will have WORSE isp than chemical rockets, due lower temperatures.

isp is square-root relative to molecular mass. As light molecule as possible is desired. That means also light atomic mass.


Lets look at pure elements:

We can achieve about ~1000 seconds with  H2, atomic mass of 1, molecular mass 2.

Going from hydrogen to Helium, which is the next element, has atomic and molecular mass 4, which already divides our isp by sqrt(4/2) so get only about 700 seconds. This is the only one with decent isp

The next ones are Lithium, beryllium, borom and carbon, which are solid in reasonable storage temperatures, not viable.

Then we have nitrogen. Atomic mass 14, molecular mass 28. Which means sqrt(28/2) = 3.7 times worse isp than hydrogen. We are already at worse isp than chemical rockets using easily storable fuels.

Oxygen is even worse than nitrogen. Atomic mass 16, molecular mass 32. 4 times worse isp than hydrogen. AND hot hydrogen is very corrosive and makes everything burn easily.

Getting so bad already that no point of looking towards heavier elements.


So what about molecules consisting of multiple elements?

Methane has molecular mass of 16. sqrt( 8 ) = 2.8 times worse isp than hydrogen, about 360s. Same isp can be achieved by burning that methane chemically, achieving much safer and cheaper system with much better T/W.

Water has molecular mass of 18. sqrt(9) =3 times worse isp than hydrogen, so about 330. Better can be achieved by storable elements with chemical rockets.

[hkultala]

280 seconds Isp? Not great - but still a cool idea and the specific impulse could probably be tweaked.

No, it could not be improved considerably.

It's consequence of the high molecular mass of the exhaust. There is no "tweaking" than can get you over the physics that are really against these high-molecular-mass NTRs.

[ncb1397]

The next ones are Lithium, beryllium, borom and carbon, which are solid in reasonable storage temperatures, not viable.

Lithium's melting point is 454 K, which is 181 K above STP. Hydrogen's boiling point is 20 K, which is 253 K below STP. I guess liquid lithium's advantage would be 7x density, perhaps easier to store in the (far) inner solar system with a penalty to isp of 53% of H2 because dilithium is an uncommon species of lithium vapor. If the base isp for hydrogen is 1100, that translates to around 550 seconds. Density is what tends to kill NTP. For instance, say you take a BFS and fill all the tanks with hydrogen and replace the raptors with NERVA. Say those mods didn't affect the dry mass. Your delta-v with 0 payload based on the rocket equation goes down. Works well as an upper stage because the BFS would weigh so much less though. Do the same thought experiment with lithium and your delta-v with zero payload goes up a few km/s.

[MATTBLAK]

280 seconds Isp? Not great - but still a cool idea and the specific impulse could probably be tweaked.

No, it could not be improved considerably.

It's consequence of the high molecular mass of the exhaust. There is no "tweaking" than can get you over the physics that are really against these high-molecular-mass NTRs.

Yeah: having higher temperature and pressures would help - but you're right, not a lot of improvement to be had. The only attraction of NIMF is the local readily available quantities of CO2. The Isp would be poorer than hypergolics, but at least you wouldn't need separate propellant tanks.

Using a NIMF system on Titan would get you 98% percent nitrogen as your reaction mass. As pointed out above: low Isp, but the low gravity of Titan plus the plentiful, thick nitrogen atmosphere would allow the use of wings on a NIMF rocket plane! Or would that be 'NITF'?! Actually, 'NIFTy' would make a better nickname - Nuclear Indigenous Titan Fuel... Someone write an engineering paper, quick! I'm not qualified...

[mmeijeri]

So what about molecules consisting of multiple elements?

Methane has molecular mass of 16. sqrt( 8 ) = 2.8 times worse isp than hydrogen, about 360s. Same isp can be achieved by burning that methane chemically, achieving much safer and cheaper system with much better T/W.

Water has molecular mass of 18. sqrt(9) =3 times worse isp than hydrogen, so about 330. Better can be achieved by storable elements with chemical rockets.

Hydrazine and especially ammonia could work.

[speedevil]

Hydrazine and especially ammonia could work.

Hydrazines ISP when run through a nuclear engine comes notably close to the ISP you get if you leave off the nuclear engine.

[KelvinZero]

The whole idea of NTR is based on idea of the nuclear reactor allowing propellant with low molecular mass. Without such, it will have WORSE isp than chemical rockets, due lower temperatures.

My basic understanding is that this becomes an issue only because we can assume there is some fixed maximum temperature of the rocket hardware before it melts etc. If we had unobtainium we could just run our nuclear thermal rockets hotter, and we would not worry about the relationship between ISP and temperature anymore.

That got me wondering about whether there was any way to get the propellant hotter than the fissionable materials. I had a few half formed ideas, but in any case I came across this which probably comes down to the same principle.

https://en.wikipedia.org/wiki/Pulsed_nuclear_thermal_rocket

Anyway, they are claiming to be able to get propellant hotter than the fuel, and also absurd ISPs even greater than Fission Fragment rockets. I don't think you WANT an ISP that high, since absurdly high ISP is often tied to absurdly low thrust, but they could be on to something that in a practical sense breaks the unbreakable relationship between molar mass and ISP in a thermal rocket (a limit which non thermal rockets already do not have to deal with)

.. I mean, it breaks it by allowing the temperature part of the equation to go off the scale for a tiny instant without breaking your rocket. That was my guess from a quick reading.

[mmeijeri]

Hydrazines ISP when run through a nuclear engine comes notably close to the ISP you get if you leave off the nuclear engine.

That's not true. Hydrazine decomposes into nitrogen, ammonia and hydrogen, you shouldn't look at the molecular weight of hydrazine itself. I can't find the correct numbers right now, but the performance was >500s Isp, which is way higher than ordinary chemical propulsion, let alone hydrazine monopropellant.

[Symmetry]

There are certainly ways to get a higher temperature out of a nuclear rocket.  There's no reason, in theory, you can't just let the fuel melt or vaporize.  Just as we run modern jets at tempreatures higher than the melting point of the jet's turbine blade through clever use of airflow people have proposed doing the same thing with nuclear rockets and a transparent quartz casing.  Of course, modern Jet engines represent a stupendous engineering investment and they have the advantage that the fuel is also part of the propellant, making heat transfer much less of a problem.  I don't know that this is a practical solution on any reasonable timeframe, but it is one that doesn't require unobtainium.

[Cherokee43v6]

I would say that your fuel choice depends entirely on what you are using the NTR for.

From a layman's perspective and in automotive terms, ISP = Miles per Gallon whereas Thrust = Engine Torque

Usually, a highly fuel efficient car does not have high torque, so you wouldn't use it to pull stumps out of the ground on your farm, whereas a high torque vehicle would not be a good car for a road trip.

So, that said, are you using your NTR to leave a gravity well or are you using it for low-grade constant thrust to shorten travel times and give a sense of gravity?  Or is it both and you're switching fuels between modes?

[Asteroza]

The basic issue is classic solid fuel nuclear thermal rockets can't operate above the fuel melting point normally. Which is why NASA went after cermet fuel assemblies to boost the temperature a bit, thus boosting the ISP a bit. Your alternatives to a classical NTR is much harder, partly because of the residence times of the heat of propellant invading your structure. Sure, there are various gas core type nuclear rockets, but you usually try to not let the gas core escape, and the propellant is still LH2.

The most direct NTR descendant would be ESA's design called NTER, which keeps propellant cooler when running through the solid fuel matrix to improve safety margin, then because of the tremendous cold sink that LH2 is, runs a inductive heating device called a turboinductor (basically a block of tungsten heated via induction) to raise propellant temperature to near tungsten melting temperatures after being "preheated" by the nuclear fuel matrix. ESA seems to think running a final turboinductor stage without the tungsten heat exchanger block may be possible, as the hot hydrogen is near plasma in composition, allowing direct inductive heating without a tungsten intermediary.

While ESA envision the rest of NTER to be a conventional chamber/throat/nozzle, it doesn't necessarily have to be so. Assuming you had available power, the plasma output of the turboinductor can be fed into a otherwise hydrogen plasma compatible electric thruster. NTER effectively replaces the helicon ionizer in a VASIMR, so any MHD style accelerator and magnetic nozzle should work. The problem there is where to source the power to run that apparatus, as NTER was designed conceptually as a direct drive inductive heater with zero power electronics. While there may be remaining extractable power in the reactor assembly, you need an auxiliary power cycle, with attendant generator, power conditioning/conversion, and a radiator to cool that aux cycle (likely a high output helium brayton cycle turbogenerator as usually expected with bimodal NTR's/NEP's). There's also the performance tradeoffs between balance of plant to drive a high throughput helicon directly attached to the reactor exit, compared to the weight of the tungsten block in the turboinductor.

This is largely in the context of fairly conventional gas/liquid propellants, notable LH2. We haven't discussed the nuclear equivalent a solid rocket, either where you have a solid propellant block that gets vaporized by the reactor from the end (fed like a wire essentially, or reactor travels up the propellant block). Conventional solids aren't that attractive from an ISP standpoint if doing things on a pure thermal basis with no post-vaporization electric acceleration.


Now unconventional propellants are a different story. There are far-out propellants like metastable hydrogen but nobody has made bulk quantities so that's out. Boron nitride fullerenes might be a possibility, if you can figure out a way to crack them open safely and controllably (say pump the fulborenes through an empty reflector core where the reactors peak neutron flux zone is to crack them).  I vaguely remember carbon dust propellants being proposed somewhere for fusion plasma rockets, that might be applicable though more in the plasma thruster variant design space though.

[speedevil]

If you are starting out by assuming a large reactor, it may be worth also looking at if it's a better bet to use that reactor in-situ for power at the place you are getting your novel propellant to process it into more conventional propellant.

For example, a hose into Titans ocean while sitting on a ice mountain to get either hydrogen/oxygen, or methane/oxygen (which would have less problems with drag), and providing lots of power for a surface station with greenhouses that only need to import trace elements.

[KelvinZero]

There are certainly ways to get a higher temperature out of a nuclear rocket.  There's no reason, in theory, you can't just let the fuel melt or vaporize.  Just as we run modern jets at tempreatures higher than the melting point of the jet's turbine blade through clever use of airflow people have proposed doing the same thing with nuclear rockets and a transparent quartz casing.  Of course, modern Jet engines represent a stupendous engineering investment and they have the advantage that the fuel is also part of the propellant, making heat transfer much less of a problem.  I don't know that this is a practical solution on any reasonable timeframe, but it is one that doesn't require unobtainium.

Those are all options, but I think hydrogen is still by far the best propellant to run through these designs?

What I was really getting at is that there are possible ways to escape that constraint that makes hydrogen so superior. Non-thermal rockets escape this requirement. For example you could imagine a mass driver propulsion with magnetic buckets throwing sand. Molar mass would be totally irrelevant. Laser heating also lets you escape this requirement, because the laser can run at moderate temperatures while producing millions of degrees for tiny fraction of a second on a microscopic layer of propellant. You can even initiate fusion. That is an example of how absurdly hot lasers can make a target while not being near that temperature themselves.

That "nuclear lightbulb" for example. If that was outputting a beam of laser light I think you could create absolutely arbitrary ISP, at the usual trade off of lower thrust. You would have totally escaped the physics that makes hydrogen so superior and your propellant could probably be a lump of iron.

[speedevil]

What I was really getting at is that there are possible ways to escape that constraint that makes hydrogen so superior. Non-thermal rockets escape this requirement. For example you could imagine a mass driver propulsion with magnetic buckets throwing sand. Molar mass would be totally irrelevant. Laser heating also lets you escape this requirement, because the laser can run at moderate temperatures while producing millions of degrees for tiny fraction of a second on a microscopic layer of propellant. You can even initiate fusion. That is an example of how absurdly hot lasers can make a target while not being near that temperature themselves.

Dusty fission rockets are also a possibility if you're going this route.
https://en.wikipedia.org/wiki/Fission-fragment_rocket#Dusty_Plasma

Radioactive decay of nanoparticles of fuel make them strongly charged. This lets you capture them electrostatically in principle, and then deflect the decay products out the nozzle magnetically at a large fraction of the speed of light.

Power conversion losses might be avoided, which other schemes have problems with.

[gin455res]

Here is an interesting web-site that argues that for dv < 3km/s water is the ideal propellant.

http://neofuel.com/staiff1999/index.html

I wonder though what the maximum isp of monopropellant HTP is in a vacuum and whether it would be better to keep the reactor on the moon and use it to supply a fleet of ships with HTP. Or perhaps use a two stage system with the first stage using nuclear produced HTP, and a water NTR second stage.

[Propylox]

edit: I don't know what happened to my post. It looked good, then vanished so I'll restate.

  If two propellants operate at the same Temperature, yes the lighter will have higher velocity and thus isp and Thrust. Temperature = (molecular mass * velocity^2)/2 ; so if the mass is (double) that of Hydrogen the velocity is 1/sqrt(double), or around 70%.
  However, Pressure and Temperature are interchangeable in ideal gasses. Pressure * vol = Temperature ; so if the core Pressure is (double) for a mass it will have the same velocity, isp and Thrust as Hydrogen while keeping the same operating Temperature.
  I'd be remiss if I didn't point out Helium has a critical pressure 1/6 that of Hydrogen, easily becoming a supercritical fluid. In this state, and at core pressure (double) that of Hydrogen, the Helium will conduct heat better and can therefor use a smaller core to achieve the same performance. The critical pressure decline from Halogen to Noble gasses is also true, but around 1/2 ; so Neon would need to operate at around 30atm and Krypton 55atm to be in a supercritical state and compare to the performance of Hydrogen.

[speedevil]

Here is an interesting web-site that argues that for dv < 3km/s water is the ideal propellant.

http://neofuel.com/staiff1999/index.html

I wonder though what the maximum isp of monopropellant HTP is in a vacuum and whether it would be better to keep the reactor on the moon and use it to supply a fleet of ships with HTP. Or perhaps use a two stage system with the first stage using nuclear produced HTP, and a water NTR second stage.

The cost of rocket fuel and other mass in space can drop by orders of magnitude because Earth can launch "pumps." A pump delivers orders of magnitude more mass than its own mass.  Like a pump, each ton of hardware sent to space to operate a steam rocket architecture can return between 10,000 and 100,000 times its mass in rocket fuel at Low Earth Orbit. This is completely equivalent to orders of magnitude drop in costs.  A chemical architecture would only return 100's of times its mass.

This does rely on the cost deltas being similar.
If the reactor costs 'only' $500M for 20 tons of very high power reactor, and you can launch 2500 tons of chemical stuff for the same price, suddenly you're back in the same ballpark.

I would really love a political climate in which launching uncritical reactors made from low enriched uranium was not an intractably expensive thing to do, and did not have political risks to missions, I don't see any realistic prospect of this any time in the near future.

[ArbitraryConstant]

I would really love a political climate in which launching uncritical reactors made from low enriched uranium was not an intractably expensive thing to do, and did not have political risks to missions, I don't see any realistic prospect of this any time in the near future.

FWIW I think any Earth launched reactor is going to be highly enriched. The design considerations are similar to naval reactors except even more so. And I don't think it's politics except inasmuch as they have to keep cutting things to fund shuttle derived launchers.

[edzieba]

All the current work on the notional SLS Nuclear Upper stage engine is on adapting existing NTR designs to operate on low-enriched fuels.

[speedevil]

I would really love a political climate in which launching uncritical reactors made from low enriched uranium was not an intractably expensive thing to do, and did not have political risks to missions, I don't see any realistic prospect of this any time in the near future.

FWIW I think any Earth launched reactor is going to be highly enriched. The design considerations are similar to naval reactors except even more so. And I don't think it's politics except inasmuch as they have to keep cutting things to fund shuttle derived launchers.

High enriched is significantly worse than low enriched, as it adds proliferation as a concern, and makes sourcing fuel much harder.
The political aspect arises as you absolutely can't do such a mission privately and have an assurance you will be allowed to launch, or even to proceed to fueling.

This also means that if it's a government mission, with a high power nuclear reactor, it's not just going to be ~$1B or so for 20 tons, but much much more, if the usual set of prime contractors gets their hands on it.