Some nuclear power questions for the good (?) Doctor

[Schol-R-LEA]

I think this the first time I've posted here since, what, 2006? Anyway... I had a few ideas fermenting in my skull regarding power plant design, and I wanted to ask Gav if he thought they would work - or rather, why they wouldn't work, as they aren't exactly the most abstruse of ideas one could have and I expect they have been considered and rejected already. I am asking as much to get myself to move on from them as much as anything else, TBH.

• Would it be feasible to use a fusion neutron source (say, a fusor) to pump a sub-critical pile fueled with a low-potential fissile material (e.g., U-233) to criticality, and if so, would the energy return be positive? I am thinking of it as a means of designing a low-reactivity reactor that would be passively failsafe (since cutting power to the neutron source would stop the reaction).
• I have read some of the public material on pebble-bed reactor designs (mainly on Wikipedia, so the accuracy is probably suspect), and what I have seen seemed to indicate that the existing pilot designs used a simple (literal) pile of the fuel pellets in a simple cylindroconical housing, with the spent pellets gravity-fed out the taper. According to the articles, one fault in the test reactors to date is jamming of the pellets in the taper. My question is, would it make sense to use set of channels for the pellets, arranged either as a cylindrical bundle, or in a series of concentric spiraled tubes? This would not only limit the risk of clogging (as the channels wouldn't have enough freedom of motion to get stuck, and even if one did, it wouldn't interfere with the other channels); furthermore, while it would vastly exacerbate the design complexity, it would simplify instrumenting the pile, and if the fuel channels were interspersed with separate channels for coolant, moderators, and controllers (probably as pellets or liquids rather than rods, in keeping with the overall design - perhaps a method could be devised in which absorber and moderator pellets could be interdigitated, with the mix controlling the reaction and an emergency dump for releasing them passively by gravity feed for scramming all or part of the pile), would give a large heat exchange surface area without requiring the coolant to contact the fuel directly.
• Regarding fusion generators, specifically of the magnetic-containment types, assuming that the plasma could be made to flow in a single direction by differential control of the containment fields (I believe this is done in most Tokamak type designs already), has any work or thought been given to using magnetohydrodynamic induction as a continuation ('topping' I think is the term) method - that is, could it be used to provide enough energy to power the containment and pinch fields, separate from any possible thermal or thermal-neutron turbine? I understand that, since MHD induction does not affect the heat of the charged material, it should allow the thermal process to act normally, and that the (very few) gas-jet MHD plants in operation combine the induction process with a conventional steam turbines to maximize operational output. Would this be possible, or would it interfere with the containment fields?

[Wafath]

1) I remember hearing about using ex-warheads in a sub-critical array, with an external neutron source of a linear accelerator being net energy positive. Not quite what you are suggesting, but possible. I also have heard of hybrid designs with fusion, but frankly, if you are using fissile fuel, you might as well just go with a proven design.
2) This sounds like a bubble-gum dispenser problem.
3) I really don't get plasma physics. CFD is complicated enough without adding in another handful of variables.

[Shonkin]

Why bother with uranium anyway? The liquid thorium fluoride breeder reactor is inherently fail-safe. A small amount of U-233 added to prime it produces neutrons. Neutron + Th-232 -->Pa-233 + beta --> U-233 + beta. The reactor only has a small amount of fissile fuel at any given time. If it overheats, the liquid ThF4 expands and is expelled from the reaction zone; this acts as a governor. If the heat is turned off, the ThF4 solidifies and the reaction stops.
The Atomic Energy Commission worked on Th reactors for a time but lost interest because it's not practical to make a nuclear weapon from Th.