Tritium

I don't think we should continue this discussion in public. I have no wish to make nuclear weapon proliferation any easier than it is already.

I love the disclaimer that availability may be limited by country. Thermo took over the company that I used to work for a long time ago.

All the stuff I know is available in books from Amazon, or online. The neutron generators are for sale for commercial applications. A PC now has the compute power that was Cray turf... and has the throughput of a billion slide rules.

Even some of our products are export controlled. We have to get plausible end-use statements from purchasers, and there are places we can't ship at all.

The rules on pulse generators were recently changed. It used to be that anything that would put 5 volts into 50 ohms in under 1 ns was export controlled. Nowadays, all sorts of 30 cent CMOS chips will do that. The new rule is something like 6 volts in 750 ps, still easy to do with commodity parts.

John

The Uranium 'gun' type bomb can be detonated without compression, by bringing two subcritical pieces of U235 together so quickly that they don't fly apart before criticality has a chance to run for a while. This design is so simple that they didn't even bother to test it, and it was used in the first atomic weapon, Little Boy, used on Hiroshima.

Plutonium 'gun' would in principle be also possible, but the natural neutron activity of practical Pu is so high that it would pre-detonate. This, plus the fact that compression requires much smaller amount of material for a weapon, and, as you say, the resulting weapon is secure because to set it off requires very specific firing sequence for the conventional explosives.

Not necessarily. Modern designs can be initiated from only two points, and if you are willing to sacrifice performance a single point detonation linear compression can be used eg flying plate

This type of weapon is so dangerous and so inefficient that to my knowledge, only one was ever built. I suppose a rogue nation wanting to experiment early in their weapons program might also build one like that. But, it takes a HUGE amount of 95%+ enriched Uranium to make one that works, and that Uranium could be used to breed a whole bunch of Plutonium in a reactor. The implosion-type devices allow MUCH less fissile material to be used, at standard density the mass is something like half of a prefect spherical critical mass. Also, the inertia of the implosion maintains the criticality until the fissile material is used up in the chain reaction. So, the reaction doesn't stop due to the mass being flung apart by the rebound, it stops because the fissile material is undergoing fission and just dissappearing. That process is even more rapid than the rebound.

The one unit at Hiroshima gave a poor yield due to pre-detonation, as they knew was likely. The real danger is this design can detonate when dropped and NOT armed, or in a plane crash. In the early US nuclear days, they had a NUMBER of accidents, some including cooking off the high explosives or triggering the self-destruct mode of the permissive action link. If any of those had been gun-type devices, a REAL disaster could have happened.

I didn't mention these as Jan's original message specifically mentioned US weapons.

Jon

These are generally shunned, as some tests with the early Swan devices showed that they were capable of some nuclear yield when only a single initiator was fired. This worried the safety people and they made changes to the design such that this wouldn't happen.

Jon

I suppose a rogue nation wanting

No. Only a fraction of the plutonium in an implosion bomb is fissioned. The inertia is inconsequential compared to the thermal stuff going on. Fat Man fissioned about 13% of its plutonium, admittedly more than gun-type bombs.

So, the reaction

No. The fission core disassembles in tens of nanoseconds.

John

Are we looking at some bead windows here?

They are also limited in fission yield. Not sure what they were originally designed for eg low yield "backpacks" or fusion triggers