Fission fireball, cold neutrons

G R L Cowan · Veteran Member

Fission fireball, cold neutrons

10kBq Jaro · Senior Member

Hi Graham,

If I understand your question correctly, if all its neutrons are coming from cold hydrogen, a critical ball of 235-U vapour could be extremely hot, and function in steady-state mode no problem.

That's because uranium nuclei are far more massive than neutrons, and as such are very ineffective as neutron moderators -- as you know, the best moderators are nuclei with mass close to that of the neutron, such as hydrogen (m=1) and deuterium (m=2). Beryllium is also pretty good (m=9), but carbon is much less effective (m=12), requiring much more bulky reactors or high enrichment levels, and anything heavier than carbon that is pretty useless.

I should add that moderators can work both ways -- to cool down neutrons, as well as to heat them up. Likewise, non-moderators, such as uranium, will not be effective at heating up cold neutrons -- especially not if the core is a low-density vapour. If instead of vapour the core was a very hot solid (metal), the effect would likely be to see a slightly increased fission rate close to its surface, where the coldest neutrons come in, before they pick up any heat.

But if all the neutrons are coming from cold hydrogen, that implies that there will be a significant heat load to manage to maintain that hydrogen cold. This will be both due to the fast neutrons coming out of the vapour core, as well as gamma rays (and assuming the thermal insulation between the core & moderator is perfect, to avoid additional heat loads due to conduction or infrared radiation). We have exactly that problem in CANDU reactors, where the fuel channels are cooled by pressurized water at ~300 C, but the heavy water moderator is maintained only ~80 C. To minimize heat transfer between the two, the fuel channel pressure tubes are separated from the (concentric) calandria tubes, with a small annular gas space. The fast neutrons and gammas dump quite a few megawatts of heat into the moderator, which must be constantly removed with heat exchangers -- being low grade heat, its simply discharged as a loss to the system.

From a practical point of view - for space propulsion applications - hot uranium fuel is not particularly desirable as propellant, because of its low specific impulse, and because you want to hang on to it, to use as your thermal power source. What you want is exactly the opposite : a low-mass material, similar to, but not necessarily the same as your reactor moderator, to use as propellant, at the highest possible temperature.

For space electric power generation, there is a much wider range of possible designs. The vapour core design you describe actually sounds fairly similar to the VCR concept developed by INSPI, in which a partly-ionized uranium vapour core is maintained in a critical state by a large amount of cooled beryllium oxide moderator around the core. It makes sense in this case, because the uranium fuel is recirculated back into the reactor, following its passage through MHD generators, to produce power....

Hope this answers your question.

-- Edited by 10kBq Jaro at 01:08, 2005-11-14

G R L Cowan · Veteran Member

10kBq Jaro included:

Hi Graham,
If I understand your question correctly, if all its neutrons are coming from cold hydrogen, a critical ball of 235-U vapour could be extremely hot, and function in steady-state mode no problem.

... if all the neutrons are coming from cold hydrogen, that implies that there will be a significant heat load to manage to maintain that hydrogen cold. This will be both due to the fast neutrons coming out of the vapour core, as well as gamma rays (and assuming the thermal insulation between the core & moderator is perfect, to avoid additional heat loads due to conduction or infrared radiation)...

The second Figure 3 in http://www.lascruces.com/~mrpbar/Space%20Policy%2002.pdf, the one nearest the end, shows the sort of thing I was asking about. The idea is not to insulate the core from the moderator but strongly connect them with thermal radiation, and that radiation won't be so much in the infrared as in the far UV.

Where that light is being absorbed, the moderator is hot and so must be neutrons in equilibrium with it. But maybe, if hydrogen is dark enough at those wavelengths, the thermal photons can have much less range in it than neutrons do. Then the hot layer that is about to go out the nozzle can be neutronically thin and the neutrons the core is getting can bring in coolness from a greater thickness, farther out, of cold hydrogen that is continually being replaced.

(I don't see a thick cool layer in that drawing but earlier, cruder ones showed the U(g) in an actual sphere, centred in a spherical chamber of about 4/3 the radius, the extra third perhaps allowing some cool thickness. If U(g) is a bitch, maybe we can just tell it, Sit! Stay!, and get good results.)

The effective-temperature deal is simple enough if the neutrons' temperature is zero. For thermal collisions between hot fuel nuclei and neutrons at 0 K to occur with the same slowness as in a 1500-K equilibrium fuel/neutron mixture, where the neutrons go at an average of 5 km/s, it must be the nuclei that go at that speed, and are at 1500 K times the ratio of molecular weights, which is to say, 1500 K times ~234. That should be enough.

--- Graham Cowan, former hydrogen fan
boron as energy carrier: real-car range, nuclear cachet

GoogleNaut · Guru

I'm really not sure how it could be possible to reduce the neutron's temperture to nearly zero, since the neutrons originate within the fissioning fuel--presumably within the 'fireball' at the center of a vapor core reactor. It would seem to me that such a reactor would almost certainly have to operate in a 'fast mode' because it may be technically too difficult to place that much 'cold' hydrogen close enough to act as a moderator. Also there is the neutron reflector which must be placed outside this 'hydrogen layer.'

BeO is mentioned often enough, because Be is a fair moderator, and BeO has a high thermal conductivity (at lower temperatures!) This could be a problem if radiative coupling to the inner wall is severe enough.

I would think that at the supposed core operating temperature of a full on uranium gas core plasma operating at nearly 260,000 K, uranium ionization due to thermal scattering of electrons may induce Bremstrulung radiation (radiation emitted by charged particles moving within a magnetic field) to be so severe that it radiatively quenches the plasma in the reactor. In other words, the Uranium can lose enough electrons in its outer shells, to become a fantastic emitter of EM radiation--which of course because of the temperature of the "Black Body" would have a peak at the extreme-UV/soft X-Ray part of the spectrum. This doesn't sound like a big deal, but when I was taking physics in college, just for fun I tried plugging in the numbers for 'typical' operating parameters for density, power levels, and operating temperautes into some of the black bodie equations and came to the conclusion that operating at nearly 260,000K would result in a plasma that would radiate almost 10 times the power that the reactor produced. The reactor would quickly cool to a lower equilibrium operating temperature, assuming of course the thermal pulse impinging upon the inner wall of the reactor would not in fact, vaporize it! I did not even try to estimate the thermal load from the fast neutron flux, I was only aware that it would simply add more heat to the wall, which made me quickly realize that it probably could not operate at 260,000 K.

The VCR or vapor core reactor has the advantage of using uranium in the form of a gas already, UF4 I think, which significantly simplifies the startup. Operating temperatures are high, but far, far lower than "straight" uranium. Reactivity levels can be throttled fairly easily by controlling the pressure of the UF4, in the core, and it can be cooled by KF which will partially vaporize which will add to the vapor content of the gasses exiting the core. Additional reactivity control could be obtained by using moveable 'windows' within the reflector surrounding the core.

The trouble with VCR's is that it is very difficult to operate in a thermal rocket mode with sufficient thrust to do any good because of the need for several heat exchangers (fairly heavy and bulky items,) but it is far more efficient as a power source for NEP (nuclear electric propulsion) options, which makes it a possibly ideal power source manned interplanetary or translunar space craft.

-- Edited by GoogleNaut at 16:44, 2005-11-16

10kBq Jaro · Senior Member

G R L Cowan wrote:

The second Figure 3 in http://www.lascruces.com/~mrpbar/Space%20Policy%2002.pdf , the one nearest the end, shows the sort of thing I was asking about.

OK, that helps.

This reactor type is, from the neutronics point of view, similar to INSPI's Vapour Core Reactor : The moderator material is the dark stuff around the vortex chamber (probably intended to be either graphite or BeO, depending on the size -- BeO is expensive !!). Also shown are cooling fins intended to cool the moderator radiatively, as well as two cylinders inside the moderator, presumably representing reactivity control mechanisms (drums with neutron absorber).

In this scheme, the hydrogen pumped into the vortex chamber does not act as a moderator, because there's too little of it, and its at low density. The hydrogen is simply used as propellant, similar to the hot coolant water in CANDU reactors, which is kept separate (and in relatively small total quantity) from the cold moderator. Fast neutrons from the fissioning uranium vapour will easily pass through to the solid moderator, then slow down and diffuse back into the core without picking up much heat along the way (due to the low density and total mass of the hot propellant hydrogen).

When the director of INSPI briefly discussed this concept at a seminar I attended a few years ago, he expressed considerable pessimism about it -- which I would definitely agree with.