Clean lithium fission rocket

[sevenperforce]

How much fuel (of every kind) would be needed to get to Mars or Jupiter in a decent short time? And so, how big would be the spaceships with this engine? Considering the possibilities talked about in this thread, that is, high ISP and decent thrust, it would still be feasible to launch from a few Falcon Heavies. Never land it again. Every time it comes back to Earth, keep it in orbit and refuel it.

Getting an estimate of actual performance would depend on how small a high-flux, low-heat reactor can reasonably be made. Unfortunately I really have no idea what that would be, or how long it would last. The reactor will eventually run low on fissile material to produce the desired neutron flux and need to have its fuel rods replaced.

This isn't quite torchship level; I doubt it could manage a 1g brachistochrone to Mars (though it could probably do a 1g brachistochrone to the moon without much difficulty). It could potentially manage a partial brachistochrone, or an 0.3g brachistochrone to Mars. That's kind of the sweet spot because it reduces transit time dramatically and also prepare the crew for Martian gravity.

The need for fuel rods replacement means that it would be really nice to be able to return to Earth under its own power, flood, and remotely service.

To reduce tritium release, what if an element was included (either in the propellant or in the salt) which preferentially binds to tritium and would keep it out of the water supply long enough for it to decay? Chlorine might work. My nuclear chemistry is somewhat lacking though.

[Impaler]

I think Tritium is just going to be a showstopper for Earth surface launch unless the rockets production of tritium is really really tiny and the amount released would not significantly raise background levels that exist naturally and are released from conventional nuclear industry activities.  Can we try to estimate the quantity that might be released per unit of rocket impulse.

I'm thinking that the main challenge of the engine is going to be keeping a high enough neutron flux in the combustion chamber to get an acceptable burn-up percentage of the lithium while the lithium is removing neutrons from the chain-reaction that is producing these neutrons.  If you take too many neutrons and the chain reaction is quenched and the propellant will stop burning and likely can't be lit again until the lithium flow is halted so the neutron flux can recover.  If the reactor is over producing neutrons then you have a potential super-critical state and explosion particularly when their is no propellant flow in the reactor.  You want to have a way of shutting down the reactor and dropping the neutron flux, perhaps you really do want to propellant flow to contain most of the neutron moderators as their absence will naturally quench the reactor as fast neutrons don't absorb well and don't sustain a chain reaction as easily.  Also the proposed coolant loop could contain Helium which is an efficient moderator, losing the coolant either in a breach or by intentionally changing it to a different gas such as one with boron would rapidly quench any chain-reaction.

[sevenperforce]

I think Tritium is just going to be a showstopper for Earth surface launch unless the rockets production of tritium is really really tiny and the amount released would not significantly raise background levels that exist naturally and are released from conventional nuclear industry activities.  Can we try to estimate the quantity that might be released per unit of rocket impulse.

I'm thinking that the main challenge of the engine is going to be keeping a high enough neutron flux in the combustion chamber to get an acceptable burn-up percentage of the lithium while the lithium is removing neutrons from the chain-reaction that is producing these neutrons.  If you take too many neutrons and the chain reaction is quenched and the propellant will stop burning and likely can't be lit again until the lithium flow is halted so the neutron flux can recover.  If the reactor is over producing neutrons then you have a potential super-critical state and explosion particularly when their is no propellant flow in the reactor.  You want to have a way of shutting down the reactor and dropping the neutron flux, perhaps you really do want to propellant flow to contain most of the neutron moderators as their absence will naturally quench the reactor as fast neutrons don't absorb well and don't sustain a chain reaction as easily.  Also the proposed coolant loop could contain Helium which is an efficient moderator, losing the coolant either in a breach or by intentionally changing it to a different gas such as one with boron would rapidly quench any chain-reaction.

One atom of tritium would be released for every 4.8 MeV lithium-6 fission event. So for every liter of enriched saltwater which fully fissions we are looking at roughly 55 grams of the stuff.

The combination of very high thrust and very high impulse makes it hard not to want to pursue it anyway though. There are numerous mechanisms for scrubbing tritium from exhaust vents or enriching it, so I am sure there is some way of dealing with it. If the expelled boundary-layer coolant contained fluorine or chlorine or sulfur, I'm pretty sure all three of those bond preferentially to tritium over protium, so perhaps they would keep it in a biologically inactive molecule long enough for it to decay.

Using the fuel as a carrier and propellant and coolant and moderator definitely adds the highest safety margin on the surface. The problem of lithium essentially acting as a neutron poison is worth analysis. However, the configuration shown in the diagram is not a limited one; you can have the fuel flowing around the outside of the reactor instead of through it.

[aceshigh]

This isn't quite torchship level

is torchship only a sci-fi term or is there a more serious design that could be called a "torchship"?

I doubt it could manage a 1g brachistochrone to Mars

thanks , I learned a new word.

[sevenperforce]

This isn't quite torchship level

is torchship only a sci-fi term or is there a more serious design that could be called a "torchship"?

The term originates in scifi, but it's not anything that couldn't exist in principle. It's basically any ship with a drive capable of accelerating at high levels of thrust (1g or greater) for a long, long time, long enough to take off vertically and not worry about pesky problems like gravity drag. No existing engine is that powerful, of course, but we can easily figure out what it would take. Antimatter-catalyzed continuous fusion drives would have that kind of performance.

Because a torchship uses a real engine rather than something purely scifi like "gravity displacement drive" or similar nonsense, we can demonstrate trivially that the exhaust of such a drive would need to be a furiously roaring fountain of relativistic death. Which is problematic. You kind of want an engine which mixes air in as reaction mass while in the atmosphere to cool things down a bit.

[Hanelyp]

To summarize the engine I see under discussion:

- A mass of nuclear fission fuel, subcritical absent the propellant.
- Introduce the propellant through coolant passages and down the reactor core, neutrons are moderated sufficiently to shift the reactor critical.
- Some of the neutrons produced react to fission lithium in the propellant, producing more heat.

Observations / questions:
- heating the propellant will reduce density, reducing neutron moderation effect.  Get the design right and the reactor will be largely self regulating.
- How much heat would be produced by the fission fuel producing the neutrons relative to heat from lithium fission?  The original post implicitly assumed a majority of heat being from the lithium fission, not the neutron source.

[sevenperforce]

To summarize the engine I see under discussion:

- A mass of nuclear fission fuel, subcritical absent the propellant.
- Introduce the propellant through coolant passages and down the reactor core, neutrons are moderated sufficiently to shift the reactor critical.
- Some of the neutrons produced react to fission lithium in the propellant, producing more heat.

Observations / questions:
- heating the propellant will reduce density, reducing neutron moderation effect.  Get the design right and the reactor will be largely self regulating.
- How much heat would be produced by the fission fuel producing the neutrons relative to heat from lithium fission?  The original post implicitly assumed a majority of heat being from the lithium fission, not the neutron source.

Yeah, you've got it about right. Note that there may be separate coolant/moderator channels so that the neutron moderation level can be adjusted independent of the lithium-salted fuel flow. Note also that the lithium saltwater may run around the reactor rather than through it, with a pure moderator/coolant channel in the center, so that the lithium doesn't poison the criticality. These are all things that will need to be specifically studied to get the ideal configuration.

Another thing that makes the whole arrangement self-regulating is that the moderator is being expelled as it is heated, so the higher the neutron flux rises, the faster the moderator will leave. A pump failure will thus automatically scram the reactor.

The majority of the energy will come from lithium fission. This is by necessity and by design. A reactor can be designed to produce a high level of neutron flux for very low levels of heat (at least compared to the typical reactor). You want the fissile neutron source to last as long as possible because it is hard to handle/refuel. The lithium saltwater propellant is very easy to refuel.

[gospacex]

You guys must be missing something about Lithium being such a nice fission fuel, since none of the existing fission reactors or bombs use lithium in any way. If it would be usable, surely it would be used by military instead of, or as an additive to the heavy and more expensive Uranium.

[Crispy]

You guys must be missing something about Lithium being such a nice fission fuel, since none of the existing fission reactors or bombs use lithium in any way. If it would be usable, surely it would be used by military instead of, or as an additive to the heavy and more expensive Uranium.

The idea is to induce fission in the propellant by bombarding it with neutrons from a secondary source.

[sevenperforce]

You guys must be missing something about Lithium being such a nice fission fuel, since none of the existing fission reactors or bombs use lithium in any way. If it would be usable, surely it would be used by military instead of, or as an additive to the heavy and more expensive Uranium.

Hey, good questions!

Lithium-6 can't be used in nuclear reactors, nor as the fissile primary for a nuclear weapon, because its decay produces no neutrons. No neutrons means that it cannot participate in a chain reaction; rather, the presence of lithium-6 will rapidly terminate a chain reaction by sucking up all the loose neutrons. Thus, it is exactly the opposite of what you would want in a nuclear reactor for sustained power generation. Fissile material like enriched uranium or plutonium will automatically begin a chain reaction once enough accumulates in one place because their naturally ongoing radioactive decay produces the initial neutrons needed to start the chain reaction, but lithium-6 is completely stable and requires an external neutron source to fission at all.

Adding lithium to existing reactors to increase yield wouldn't work well either. Lithium-6 fission has a much higher energy yield than an equivalent mass of enriched uranium or plutonium, but those fuels are so much denser that they pack more energy into a smaller space, which is what you need for a reactor. Plus, it's not good to use a neutron poison as part of your reaction fuel; as it is used up, it will slowly increase criticality and could lead to a runaway chain reaction and meltdown. Adding lithium to try and increase the yield would require MORE fissile uranium or plutonium.

However, lithium's fissionable properties ARE used in weapons...in ever thermonuclear weapon after Ivy Mike, in fact. Fusion weapons are packed with lithium-6 deuteride so that the burst of x-rays from the smaller fission bomb will compress it, followed by the burst of neutrons which fissions the lithium. This extraordinarily high energy release in a very small area is enough to trigger fusion between the deuterium and the tritium produced in the decay of the lithium-6. Pretty powerful stuff.

For a rocket engine, you don't really want a chain reaction in your propellant. That's why lithium triggered with external neutrons would work so well.

[gospacex]

You guys must be missing something about Lithium being such a nice fission fuel, since none of the existing fission reactors or bombs use lithium in any way. If it would be usable, surely it would be used by military instead of, or as an additive to the heavy and more expensive Uranium.

Hey, good questions!

...Plus, it's not good to use a neutron poison as part of your reaction fuel; as it is used up, it will slowly increase criticality and could lead to a runaway chain reaction and meltdown. Adding lithium to try and increase the yield would require MORE fissile uranium or plutonium.

Actually, this is not true that adding nuclear poisons to the fuel is a bad idea. For years, nuclear power industry is working towards making fuel campaigns longer, since each reload is a 20-30 day long reactor shutdown. 12 month power campaigns are generally replaced by 18-month ones now, and 24-month ones are in works.

To that end, more enriched fuels were necessary, and by now the 5% enriched fuel is standard. But reactors were not designed to use such reactive fuel, and nuclear poisons are deliberately added (I heard about Europium), to make fuel reactivity-versus-time curve flatter, and lower.

If Lithium is a nuclear poison which also exotermically fissions (unlike Europium), why is it not added to the fuel? Such fuel would be more energetic.

[gospacex]

I think I see the flaw with the proposal. To fission Lithium, you need a neutron source to generate *one neutron for every Lithium atom*:

Li-6 + n -> T + He-4 + 4.7829 MeV

Then you need to avoid thermalization of fission fragments and you need to collimate them into a directional beam. How far away form 100% efficient that will be?

You can save yourself a lot of trouble if you just use your neutron source as a rocket engine. Must be about the same ballpark wrt trust....

Which hints that the showstopper here is that it's hard to generate that many neutrons. What do you propose to be your neutron source? What's its luminosity?

[sevenperforce]

...it's not good to use a neutron poison as part of your reaction fuel; as it is used up, it will slowly increase criticality and could lead to a runaway chain reaction and meltdown. Adding lithium to try and increase the yield would require MORE fissile uranium or plutonium.

Actually, this is not true that adding nuclear poisons to the fuel is a bad idea. For years, nuclear power industry is working towards making fuel campaigns longer, since each reload is a 20-30 day long reactor shutdown. 12 month power campaigns are generally replaced by 18-month ones now, and 24-month ones are in works.

To that end, more enriched fuels were necessary, and by now the 5% enriched fuel is standard. But reactors were not designed to use such reactive fuel, and nuclear poisons are deliberately added (I heard about Europium), to make fuel reactivity-versus-time curve flatter, and lower.

If Lithium is a nuclear poison which also exotermically fissions (unlike Europium), why is it not added to the fuel? Such fuel would be more energetic.

Lithium-6 isn't a typical neutron poison. Europium can accept quite a few neutrons; thus, it doesn't really get "used up" right away. Lithium-6 is only a neutron poison in that it has a large neutron cross-section; once it is struck by a neutron, though, it instantly fissions. Adding lithium-6 to an active reactor would cause its criticality to increase as a function of time and thus lead to a meltdown.

Reactor design isn't about increasing energy as much as it is about controlling energy. You want to increase the stable power output period of your reactor, not add in sudden energy spikes followed by near-meltdown.

I think I see the flaw with the proposal. To fission Lithium, you need a neutron source to generate *one neutron for every Lithium atom*:

Li-6 + n -> T + He-4 + 4.7829 MeV

Then you need to avoid thermalization of fission fragments and you need to collimate them into a directional beam. How far away form 100% efficient that will be?

You can save yourself a lot of trouble if you just use your neutron source as a rocket engine. Must be about the same ballpark wrt trust....

Which hints that the showstopper here is that it's hard to generate that many neutrons. What do you propose to be your neutron source? What's its luminosity?

The two problems you cite—thermalizing fission products and collimating the fission beam—are in fact each other's solution. I WANT to thermalize the fission products. I DON'T want to try and collimate relativistic ions; it is far far easier to dissolve the lithium salt in water and let the water serve both as a thermalization medium and as the pressurized reaction mass.

The fuel is dissolved in water; the fuel is used to heat the water; the water and the fuel's reaction products are expelled together. Pretty simple.

And if you think that equivalent thrust could be produced by using a neutron source directly...no. Thrust is momentum exchange per unit time; momentum is mass times velocity. How do you imagine that a thermal neutron traveling at 2,200 m/s will produce greater thrust than that same neutron inside a helium atom traveling alongside a tritium atom (7 times more total mass) traveling at 4% of the speed of light?

Of course even that amount of thrust is dwarved by the amount of thrust produced when that 4.8 MeV is thermalize across ten or twenty or fifty times more mass in a heavy-water solution. The usual rule in rocket science is to reduce the reaction mass for a given amount of energy so that your specific impulse is as high as possible, but when we are dealing with these energy levels we are going to want to cram as much reaction mass through our engine as we possibly can.

As I have said above, the neutron since would need to be a high flux low heat nuclear reactor.

[Impaler]

This seems to be the kind of core you would want

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

As we speculated earlier a large jacket of neutron reflector (beryllium) surrounds the core which has a 5 inch across hollow center to allow samples to be exposed to the high thermal neutron flux which is 2.5 x 10^15 neutrons per cm^2/sec, given this value it should be possible to calculate the fission rate of lithium within the reactor based on the known cross section of lithium-6.

[sevenperforce]

This seems to be the kind of core you would want

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

As we speculated earlier a large jacket of neutron reflector (beryllium) surrounds the core which has a 5 inch across hollow center to allow samples to be exposed to the high thermal neutron flux which is 2.5 x 10^15 neutrons per cm^2/sec, given this value it should be possible to calculate the fission rate of lithium within the reactor based on the known cross section of lithium-6.

The HFIR core is 0.61 meters high. We don't know the flow rate of our lithium saltwater, but that's okay; we can use the size of the flux chamber to get a generalized upper bound on what kind of power output and thrust we would be able to get from a reactor of this type and size.

The inner surface area is 1217 square centimeters so the total internal neutron flux is 3e18 neutrons per second. This offers a potential maximum fission energy of 2.31 MJ per second, burning just 0.11 milligrams of lithium hydroxide, or about 1.1 milligrams of our saturated saltwater solution.

However, the HFIR uses a bunch of different reflectors and neutron poisons to try and make the neutron flux as constant with respect to location as possible. Using the wild guess that a maximum-flux version would have a total flux ten times greater, and assuming arbitrarily that we use a 1:1 ratio of saltwater fuel to freshwater coolant by mass, this comes to a power output of 23.1 MW and a mass flux of 22 milligrams per second. That is 33 N of thrust with a specific impulse of 1,449 km/s.

[Vultur]

Zubrin throws out a tank mass fraction of just 4% with no supporting evidence at all and which I find laughably small

However, the specific impulse is so good that it hardly matters.

At Isp 7000, even a mass ratio of 2 would have nearly 50 km/s delta-v, far more than any chemical rocket or NTR could have.

[Impaler]

The math looks to be a lot more complex then that sevenperforce, these links look like they provide the means to calculate it.  See if they help, I'll dive into them myself tomorrow and see if I can produce an independent calculation to see if we agree.



Lesson1

Lesson2

[sevenperforce]

The math looks to be a lot more complex then that sevenperforce, these links look like they provide the means to calculate it.  See if they help, I'll dive into them myself tomorrow and see if I can produce an independent calculation to see if we agree.

Yeah, the maths will be significantly more complicated than that. I was just doing a first order approximation to get an idea of what power levels the HFIR could generate.

If flux isn't high enough for a straight path through the reaction chamber, it would be possible to design the flow path with a helical arrangement of some other similar geometry to increase neutron exposure. If all else fails, it could be an augmentation to a traditional NTR to reduce shielding requirements and increase propellant energy above the melting point of the reactor core.

Or we could look into the possibility of using a D-T fusion primary with a deuterium ion accelerator, designed to induce fission in a radioactive secondary layer and thus exponentially multiply neutron count. This could allow very high neutron flux even if total power wasn't much more than the full reactor design.

[sevenperforce]

The math looks to be a lot more complex then that sevenperforce, these links look like they provide the means to calculate it.  See if they help, I'll dive into them myself tomorrow and see if I can produce an independent calculation to see if we agree.

It looks like the best fluxes attainable from continuous high flux reactors are on the order of 10¹⁶-10¹⁷ neutrons per square centimeter per second. Molten salt or pebble bed reactors might be able to get higher...perhaps as high as 10¹⁹ n/cm²*s, but that's pushing it.

Higher rates might be attainable using a synchroton proton beam directed at a mercury target to produce spallation; spallation can create pulses higher than 10¹⁷ n/cm²*s without too much trouble. With a depleted uranium, tantalum, or beryllium neutron multiplier, this could go very high. Attached an exemplary layout.

[Impaler]

That would look more like a pulse system which is a much more complex system that would require a massive capacitor and power source.  I'd suspect that a pellet system and a spherical combustion chamber similar to some of the fusion implosion setups.