Can a Fission Fragment rocket be designed without a moderator?

[Al_]

TL;DR Summary

I was interested to come across the concept of a Fission Fragment rocket. Wikipedia has a good article, and there are some studies by NASA. Can there be a design that functions without a heavy moderator using U235?

a dusty plasma design : https://www.nasa.gov/pdf/718391main_Werka_2011_PhI_FFRE.pdf
has nuclear fuel held in place by electromagnetic fields. It uses a very massive moderator around the fuel to slow and reflect neutrons back to the fuel, to enable continued fission. But, there is one kind of fuel that does not need a moderator: U235. Could this be controlled by expanding the plasma density, and prevented from going critical and exploding? Saving 30 tonnes by having no moderator, would make the engine much lighter, and saving the need to cool the moderator with huge radiators.

 

[PeterDonis]

there is one kind of fuel that does not need a moderator: U235.

Huh? Why do you think U235 doesn't need a moderator?

 

[mfb]

Nuclear weapons work without one. But they are also prompt critical.

You won't eliminate the cooling demand completely and cooling the fuel might become even more difficult. Controlling the reaction could become difficult, too.

 

[PeterDonis]

Nuclear weapons work without one. But they are also prompt critical.

I was assuming that this was not a solution since the reaction would be uncontrolled, and we're talking about a spaceship, not a weapon. Also, this statement would apply to any fissionable fuel, not just U235; the OP seems to believe there is something different about U235, and I don't see what it could be.

 

[anorlunda]

I think it would be very difficult to imagine some kind of nuclear propulsion that has not been considered by government and contractors.

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

 

[mfb]

It's not special about U235, it's a high concentration of U235 vs. the normal U235/U238 mixture. The latter needs a moderator to work, the former does not.
I haven't seen a combination of unmoderated and delayed critical, but maybe it is possible, I can't rule it out. Someone will have studied it.

 

[PeterDonis]

a high concentration of U235 vs. the normal U235/U238 mixture. The latter needs a moderator to work, the former does not

I'm not sure I understand. The thermal neutron fission cross section of U235 is almost 600 times higher than the fast neutron fission cross section. [1] So U235 without a moderator to slow down the neutrons fissions much worse than U235 with a moderator. In a bomb this is overcome by having an initiator to provide a huge neutron flux to start the uncontrolled chain reaction.

(For U238 the fast fission cross section is much higher than the thermal fission cross section, so if it were a practical fission fuel, it would fission better without a moderator, but even the fast fission cross section is too low to make it a practical fission fuel.)

[1] https://en.wikipedia.org/wiki/Neutron_cross_section#Typical_cross_sections

 

[mfb]

So U235 without a moderator to slow down the neutrons fissions much worse than U235 with a moderator.

Still good enough to get critical. The initiator is only there to start the chain reaction at a well-defined point, it is irrelevant for a sustained reaction.
Compared to e.g. CANDU reactors weapon-grade uranium has ~100 times more U-235. In addition you get rid of the neutron absorption in water (smaller in CANDU because it's heavy water, but not zero).

Natural uranium without moderator doesn't work.
Highly enriched uranium works without moderator.

 

[PeterDonis]

Still good enough to get critical.

Ok, so basically "doesn't need a moderator" means "can go critical without a moderator"?

Then I still don't see what singles U235 out, since that statement is also true for U233, not to mention all of the known isotopes of Pu, plus isotopes of other elements as well.

 

[mfb]

Ok, so basically "doesn't need a moderator" means "can go critical without a moderator"?

Yes. That's what nuclear weapons do, for example.

Then I still don't see what singles U235 out, since that statement is also true for U233, not to mention all of the known isotopes of Pu, plus isotopes of other elements as well.

It's not singling out U235 among other possible fuels, it's highlighting that there is no (or nearly no) U238 mixed in.

 

[Al_]

Nuclear weapons work without one. But they are also prompt critical.

You won't eliminate the cooling demand completely and cooling the fuel might become even more difficult. Controlling the reaction could become difficult, too.

Thanks for the replies everyone.
Yes, controlling the reaction is basically what I'm asking about. Since the dusty plasma can be controlled by electromagnetic fields which can be rapidly altered, perhaps it could be compressed to the point of criticality, then very quickly expanded again before an explosion happens. Or, perhaps two subcritical masses of U235 can be rapidly passed close to each other to form a just-critical mass, and rapidly away again. (A drive-by criticality!).
Or, maybe if a thin continuous stream of particles was passed through a nearly-critical mass, and shut off as criticality occurred? How fast does criticality take off? Are there any ways to make it stop immediately it starts?

 

[anorlunda]

Yes, controlling the reaction is basically what I'm asking about.

The questions you are asking are the kind that are typically covered in an introductory course in nuclear engineering. Wouldn't a course be better than asking random questions on the Internet. In almost all subjects students are not encouraged to ask questions as an alternative to class lectures and textbooks.

It sounds like you might be trying to describe a nuclear drive for SciFi purposes. Is that the case?

 

[Al_]

No, I'm not writing Sci Fi, I'm just being curious. I'm not an undergraduate student, although I was, years ago. Maybe a course would be a good idea, but I don't have time. I'm sorry if my questions appear random.

 

[anorlunda]

OK Al

Since the dusty plasma can be controlled by electromagnetic fields which can be rapidly altered, perhaps it could be compressed to the point of criticality

Plasma is a poor candidate because its volumetric density is low. You want something with high density to significantly change the reactivity.

How fast does criticality take off?

That is a function of net reactivity.

Are there any ways to make it stop immediately it starts?

Yes, reduce the reactivity.

 

[Al_]

Plasma is a poor candidate because its volumetric density is low.

The NASA study quoted above uses a dusty plasma, and has two designs, one with a critical mass of Pu, and one with U238. In both, the moderator reflects back neutrons to achieve criticality. With no moderator, the mass of fuel would need to be greater, but I don't think it would need to be denser, since the fast neutrons would not decay within the core.
However, a larger core would mean that the fission fragments would be less likely to escape.

 

[Al_]

In that reactivity link I found a mention of submarine reactors that use a high proportion of U235. Wikipedia states:
"many designs use highly enriched uranium but incorporate burnable neutron poison in the fuel rods. This allows the reactor to be constructed with an excess of fissionable material, which is nevertheless made relatively safe early in the reactor's fuel burn cycle by the presence of the neutron-absorbing material"

 

[Al_]

To enable the fission fragments to escape, yet allowing a critical mass, there could be multiple small sub-cores each with its own magnetic coils, which exchange neutrons to achieve criticality as a group. The obvious place for control rods is then between the sub-cores.

 

[anorlunda]

The NASA study quoted above uses a dusty plasma, and has two designs, one with a critical mass of Pu, and one with U238. In both, the moderator reflects back neutrons to achieve criticality.

I stand corrected, I followed the NASA link several layers deep and I found this.

http://www.rbsp.info/rbs/PDF/aiaa05.pdf

By the way, this description says that it is for a homogeneous thermal reactor, not a fast reactor.

 

[PeterDonis]

It's not singling out U235 among other possible fuels, it's highlighting that there is no (or nearly no) U238 mixed in.

I understand that's what you are saying. I'm not sure that's what the OP was saying.

 

[PeterDonis]

The NASA study quoted above uses a dusty plasma, and has two designs, one with a critical mass of Pu, and one with U238.

Where do you see the U-238? The only specific fuel I see mentioned is Pu-239. Which, btw, has a higher fission cross section than U-235 does, so if the designers think they need a moderator for Pu-239, they certainly would think they need a moderator for U-235.

 

[Astronuc]

a dusty plasma design : https://www.nasa.gov/pdf/718391main_Werka_2011_PhI_FFRE.pdf
has nuclear fuel held in place by electromagnetic fields. It uses a very massive moderator around the fuel to slow and reflect neutrons back to the fuel, to enable continued fission. But, there is one kind of fuel that does not need a moderator: U235. Could this be controlled by expanding the plasma density, and prevented from going critical and exploding? Saving 30 tonnes by having no moderator, would make the engine much lighter, and saving the need to cool the moderator with huge radiators.

There's a lot wrong with the concepts mentioned.

The design of a FFRE, instead, allows these same heavy fission products to escape from the reactor, traveling at up to 5% of light speed. Theoretically, heavy fission products traveling at up to 5% of light speed produce thrust at a specific impulse of one million seconds (over 200 times better than electric engines). The efficiency of a FFRE, as measured by the quantity of fission fragments that escape as a beam rather than remain inside the reactor and produce waste heat, in this study was about 11%.

I don't believe that performance claim.

The sub-micron sized dust, composed of Uranium Dioxide, melts at over3000 Kelvinsand enables operating the FFRE at apower of approximately 1000 MW thermal. Fission fragments that travel forward, rather than aft, are reflected by the superconducting mirror magnet and pass twice through the core on their way to escape. This "double jeopardy" reduces the fraction that escapes and reduces the average exhaust velocity to about 1.7 percent. This FFRE configuration was estimated to produce almost 10 lbf of thrust at a delivered specific impulse of 527,000 seconds.

How sub-micron? UO₂!? U-hydride would be better. I don't believe the 5% or 1.7% of speed of light. Not all fissile atoms will fission simultaneously, but rather a fraction will fission, so fission products will necessarily collide with other atoms (fuel and other fission products) and slow down. Range of fission products is on the order of microns in solid UO₂. Moving a powder of sub-micron particles can be challenging.

One would have to determine the atomic density and multiply by the flux, or actually integrate the product of atomic density of fissile atoms and energy dependent cross-section as a function of energy-dependent flux over the energy spectrum. Based on the claims, I doubt the authors did some serious computation on this system.

I've seen various gas core concepts, and a lot of wacky ideas.

The authors mention Pu-239 early on, then U-235. Pu-239 is better for small cores. One can do a fast, thermal or mixed spectrum reactor. Fast reactors usually require highly enriched fuel in order to offset the low cross-sections at fast energies. Fast neutrons have a long mean free path, so that's a motivation to moderate. A moderated reactor, which could be epithermal, requires lower enrichment.

For a gas core, one would want to basically do a coupled core with a solid core driving the gas fuel.

Another challenge with highly enriched material is storing the fuel in a subcritical configuration.