Clean lithium fission rocket

[sevenperforce]

Lithium-6 decays exothermically into helium and tritium when irradiated with sufficiently energetic neutrons.

A layer of lithium-6 hydride deposited onto a beryllium bell nozzle and irradiated with high-energy neutrons from a standard neutron source ought to undergo fission and release hydrogen and helium at impressively high exhaust velocities. You'd need a bit of mass to give the neutron source a high enough energy, but wouldn't the clean fission give you a tremendous gain?

[aceshigh]

Interesting. Is the idea yours?

[sevenperforce]

Yeah, my idea.

I can crunch a few numbers to get an idea of the specific impulse it would offer, as well as thrust-specific power requirements. It should be a lot better in terms of thrust/power than ion thrusters because the power supply need only sustain the neutron source, not actually provide the energy to accelerate the propellant.

Higher thrust could be obtained by mixing in lithium-7 hydride, which has an endothermic fission reaction but spits out a neutron of its own, doubling the reaction mass for a given neutron flux but reducing exhaust velocity.

A huge advantage is that your propellant remains solid right up until the reaction. Having monatomic hydrogen as your reaction mass isn't as much of an advantage as you might think, though, because half of it is tritium rather than protium. Each fission event releases a tritium atom, a helium atom, and a protium atom. I suppose you could get more energy if the tritium and protium burned to form diatomic hydrogen, though this would result in the propellant all being basically helium by molecular weight.

[acsawdey]

Are you thinking of something like a focus fusion device as the neutron source?

[sevenperforce]

Are you thinking of something like a focus fusion device as the neutron source?

Yeah, this would be controlled fusion-initiated fission (as opposed to the uncontrolled fission-initiated fusion in a nuclear weapon). The neutron source could potentially be an inertial electrostatic confinement design or, more preferably, an accelerator-based design. An accelerator design could be beam-to-target (a beam of deuterium ions fired at a static tritium-enriched target) or beam-to-beam (a beam of deuterium ions fired at a beam of tritium ions).

If I recall correctly, the energy fraction in fission events is divided among the fission products roughly in proportion to mass so that momentum is conserved. I don't think lithium-6 fission produces gamma rays. And the chemical binding energy of lithium hydride is something like a million times lower than lithium-6's fission energy so that's negligible. You'll end up with a helium-4 atom, a tritium atom, and the protium atom previously attached to the molecule, with 4.78 MeV of energy to divide among them. If my recollection is right, the protium atom will get an eighth of this (0.60 MeV), the tritium atom will get 3/8 of this (1.79 MeV), and the helium-4 atom will get half of this (2.39 MeV). Exhaust velocity is 10,734 km/s. Not m/s; km/s.

[acsawdey]

Exhaust velocity is 10,734 km/s. Not m/s; km/s.

That is the lovely thing about nuclear reactions in this case. Chemical bond energy is generally negligible in comparison.

With an Isp of ~1.1e6 seconds, you wouldn't need much reactant mass. Given that the nuclear fragments are almost certainly going to be ionized, could you use a magnetic field for the nozzle?

[sevenperforce]

Exhaust velocity is 10,734 km/s. Not m/s; km/s.

That is the lovely thing about nuclear reactions in this case. Chemical bond energy is generally negligible in comparison.

With an Isp of ~1.1e6 seconds, you wouldn't need much reactant mass. Given that the nuclear fragments are almost certainly going to be ionized, could you use a magnetic field for the nozzle?

My thought was to cast/deposit the lithium-6 hydride in the shape of an exhaust bell and place the neutron source at roughly the focal point. But with particle energies that high you might need some sort of confinement....

The challenge would be getting neutron fluxes high enough for useful thrust. Back-of-the-envelope estimates suggest 10¹⁹ neutrons per second for a single Newton of thrust.

[acsawdey]

The challenge would be getting neutron fluxes high enough for useful thrust. Back-of-the-envelope estimates suggest 10¹⁹ neutrons per second for a single Newton of thrust.

So, compute the power output of a 1N thruster with 1.1e6 Isp. By comparison to ion thrusters, 1N is a huge thrust.

[sevenperforce]

The challenge would be getting neutron fluxes high enough for useful thrust. Back-of-the-envelope estimates suggest 10¹⁹ neutrons per second for a single Newton of thrust.

So, compute the power output of a 1N thruster with 1.1e6 Isp. By comparison to ion thrusters, 1N is a huge thrust.

Well, the power output at 1N is simply 5.5e6 watts. But that is the output, not the input. The input depends on how you are generating your neutron flux.

To match the thrust of a single SuperDraco you would need a neutron source capable of producing 5.91e23 neutrons per second.

[acsawdey]

Well, the power output at 1N is simply 5.5e6 watts. But that is the output, not the input. The input depends on how you are generating your neutron flux.

The quote that comes to mind is

a reaction drive's efficiency as a weapon is in direct proportion to its efficiency as a drive -- Larry Niven

So, you've got a drive that produces 5.5 megawatts of power expressed as a beam of relativistic ions. At close range this is a weapon. I don't think you need anything like the thrust of a SuperDraco because you'll not want to use this on a planet because you'll damage things. Also the effect in atmosphere could be interesting.

Probably the limiting factor is the neutron source. But even if you can only make micro-newtons of thrust, at 1.1e6 Isp you can thrust for a long time on one block of fuel. I wonder if you could build something like this into a cubesat?

[sevenperforce]

Well, the power output at 1N is simply 5.5e6 watts. But that is the output, not the input. The input depends on how you are generating your neutron flux.

The quote that comes to mind is

a reaction drive's efficiency as a weapon is in direct proportion to its efficiency as a drive -- Larry Niven

So, you've got a drive that produces 5.5 megawatts of power expressed as a beam of relativistic ions. At close range this is a weapon. I don't think you need anything like the thrust of a SuperDraco because you'll not want to use this on a planet because you'll damage things. Also the effect in atmosphere could be interesting.

Probably the limiting factor is the neutron source. But even if you can only make micro-newtons of thrust, at 1.1e6 Isp you can thrust for a long time on one block of fuel. I wonder if you could build something like this into a cubesat?

Damn, that would be quite a weapon.

You can mix in lower enrichments of lithium to bring up the thrust and bring down the velocity. The limit is around 30℅ enrichment though.

[Mark K]

To match the thrust of a single SuperDraco you would need a neutron source capable of producing 5.91e23 neutrons per second.

What kind of neutron source can generate a gram/sec of neutrons?What kind of power input would you need for that?

Even milligrams/sec is awesome. Micrograms/sec ...

[RotoSequence]

How would you avoid, or at least mitigate, chemical ignition of the lithium fuel? Lithium will chemically react with both of the fission products.

[Impaler]

Coating the fissile material on the inside of a nozzle dose not seem practical, the neutron bombardment will penetrate into the coating even if it is fairly thick and cause fission in the whole mass of fissile material.  The resultant fissile byproducts will be generated inside a solid fuel mass and will collide with neighboring atoms and cause standard thermal interactions as they slow down, the whole fuel mass will simply heat up rather then emit a surface sputtering that I think your looking for.

Some form of tiny-pellet feed system inside a magnetic nozzle seems to be the only way to get a well controlled reaction that would produce manageable thrust.  The pellets can also contain non-fissile materials preferably highly transparent to neutrons which will be vaporized and contribute the bulk of the reaction mass, this will lower the ISP down to usable ranges, different pellet mixtures could produce different thrust and ISP ranges for the engine allowing a kind of throttling which has been shown to improve overall trajectory planning.  The neutron source would need to be pulse based and would only need to produce a single standard pulse both of which should make it simpler.

[KristianAndresen]

Lithium-6 decays exothermically into helium and tritium when irradiated with sufficiently energetic neutrons.
[...]
You'd need a bit of mass to give the neutron source a high enough energy

According to what I could look up, this decay can happen at any neutron energy... and don't you need quite slow neutrons in order to have a high interaction cross-section and therefore a low penetration depth, considering what Impaler said?

[sevenperforce]

What kind of neutron source can generate a gram/sec of neutrons?What kind of power input would you need for that?

Even milligrams/sec is awesome. Micrograms/sec ...

It might work to use a "dirty" fissile primary as your neutron source, like a subcritical plutonium sphere surrounded by neutron reflectors. A little unsafe, but at least you don't have to worry about carrying a power source, and fissile material isn't released into the environment. I don't know what kind of neutron count is generated in controlled criticality...

How would you avoid, or at least mitigate, chemical ignition of the lithium fuel? Lithium will chemically react with both of the fission products.

Well, I had proposed using lithium hydride rather than pure lithium metal, as this increases the particle count and also prevents the lithium from reacting (since it has already reacted). Of course the helium won't react with anything. The monatomic hydrogen might react with itself but that's not necessarily a bad thing. Specific impulse is less important here so an increase in molecular weight is not really a bad thing.

See below for more discussion of storage.

Coating the fissile material on the inside of a nozzle dose not seem practical, the neutron bombardment will penetrate into the coating even if it is fairly thick and cause fission in the whole mass of fissile material.  The resultant fissile byproducts will be generated inside a solid fuel mass and will collide with neighboring atoms and cause standard thermal interactions as they slow down, the whole fuel mass will simply heat up rather then emit a surface sputtering that I think your looking for.

Some form of tiny-pellet feed system inside a magnetic nozzle seems to be the only way to get a well controlled reaction that would produce manageable thrust.  The pellets can also contain non-fissile materials preferably highly transparent to neutrons which will be vaporized and contribute the bulk of the reaction mass, this will lower the ISP down to usable ranges, different pellet mixtures could produce different thrust and ISP ranges for the engine allowing a kind of throttling which has been shown to improve overall trajectory planning.  The neutron source would need to be pulse based and would only need to produce a single standard pulse both of which should make it simpler.

I was actually thinking of a very thin coating with a beryllium neutron reflector on the other side to produce more of a sputtering reaction. You could imagine stacking a series of thin "fuel layers" alternating between lithium-6 hydride and beryllium, with each pair of layers being discarded once most of the lithium hydride had been used up.

A magnetic nozzle is a possibility; lithium hydride is itself highly diamagnetic, but it's also an ionic conductor, so there are quite a few ways that could be used.

Since reaction mass is a much bigger concern than specific impulse (seriously, weakly relativistic ion spray is NOT nice), what about using something like enriched lithium-6 hydroxide saltwater? Lithium hydroxide dissolves in water at 129 g/L; a single liter of such saltwater therefore has 5.4 moles of lithium-6, with a total potential energy of 2.5e3 gigajoules. For reference, that's about 4% the yield of the Little Boy bomb dropped on Hiroshima...absolutely insane. Lithium-6 has the highest specific energy of fission of any fissile material.

If you used heavy water rather than light water, then you could use (cheaper) natural uranium as the neutron source. You'd want to cast natural uranium into a cylinder with a converging end and wrap it in a beryllium neutron reflector; put a tungsten diverging nozzle on other side. Give the inside some sort of neutron-transparent but very heat-resistant coating (because, for the love of all that is good, you don't want fissioning uranium spraying out with your exhaust stream); not sure what material might work for that. When you pump the enriched lithium-6 hydroxide heavy saltwater into the exhaust stream, the heavy water will slow the fast neutrons from natural decay down into thermal neutrons, causing the natural uranium to go critical and release a spray of neutrons. These, in turn, collide with lithium-6 nuclei and trigger the lithium fission.

A liter of enriched lithium-6 hydroxide heavy saltwater masses 1.229 kg, so the lower bound for exhaust velocity is going to be on the order of 2e6 m/s. Of course, you can decrease this and increase thrust by simply decreasing the enrichment of the lithium hydroxide or decreasing the amount of lithium hydroxide dissolved in the heavy water, though doing the latter tends to decrease the density of your propellant which is less ideal. Another issue is that not all the lithium-6 will fission; I don't really know how to estimate the percentage though.

The only power requirement is the turbopump to push the enriched heavy saltwater into the "combustion" chamber, and you can probably use a fuel coolant loop for that. I'd need to dig pretty deep into nuclear reactor design and criticality to see what the minimum possible size would be. Thrust to weight ratio is going to be respectable; if you were using full enrichment and full fission, matching the thrust of a SuperDraco would require a mass flow of just 36 grams per second.

If this design isn't clear let me know and I will do a graphic mockup.

Does anyone know of a material with very high heat resistance that is neutron-transparent?

[sevenperforce]

Did a little digging and it looks like someone else already had a similar idea:

Nuclear Salt Water Rockets Revisited

This proposal suggests the use of Lithium-6 deuteride, with the expectation that tritium-deuterium fusion will take place via Jetter Cycle. Such a cycle is highly unlikely in a rocket engine; the conditions for any fusion are extremely demanding. However, the use of a shielded high flux reactor is definitely a better plan than using a coated cylinder of raw uranium. Using lithium fission as the power source rather than imagining D-T fusion as the power source also allows for a much more manageable/realistic impulse and thrust. Finally, using dissolved lithium hydroxide rather than a "lithium hydride suspension" makes the propellant much more stable, inexpensive to manufacture, and easier to handle.

[aceshigh]

I had posted a thread about nuclear salt water rocket a few months ago.

quite interesting that the proposal of your link is basically the NSWR without the hiper radioactive exhaust!

[sevenperforce]

I had posted a thread about nuclear salt water rocket a few months ago.

quite interesting that the proposal of your link is basically the NSWR without the hiper radioactive exhaust!

The proposal at the link is a Thermonuclear Saltwater Rocket (TNSWR), which simply won't work. But dialing it back to a Lithium Fission NSWR would be simpler, more realizeable, and a lot cheaper to build/refuel/operate.

In the simplest form, you have something that looks like the attached image. You have natural uranium surrounded by a tungsten neutron reflector, but it cannot sustain a chain reaction without a neutron moderator. Once the heavy saltwater is pumped in, however, the neutrons are immediately moderated and a chain reaction is triggered, irradiating the lithium-6 and triggering fission, which expels everything out the bottom end at tremendous speed.

[Impaler]

This is a significant improvement over the traditional NSWR concept which would have relied on a fuel so enriched that it will spontaneously detonate in the absence of a neutron absorbing materials.  That would require a large and heavy storage tank to prevent the entire fuel mass from becoming the worlds largest bomb.  Needless to say this would have made fuel handling monstrously expensive and difficult, the only advantage is that the engine was devoid of moving parts and was just a nozzle without neutron absorbent.

But by using a stable non explosive fuel like lithium the fuel handling process is massively simpler and in fact you can have a JIT mixing system (presumably inside the turbo-pump itself) to combine water with a lithium rich 'syrup' (much like how you get Soda from a fountain) just before the flow enters the engine.  This would again allow for throttling the engine in thrust and ISP (which isn't possible with traditional NSWR), as well as smoothly shutting down or starting up the engine by transitioning from and too a pure water flow thus avoiding the 'hard start' issue with most rocket engines.

Lastly and perhaps most significantly it would allow such a vessel to carry a surplus 'syrup' which can presumably be sourced only on Earth, while refueling with more easily obtained water from in-situ sources in space (though probably not the desirable heavy water).  This would allow nearly unlimited range and proportionally higher payload fractions.

I'm going to term this the 'Nuclear Soda Fountain Rocket' concept.