Science desk
< May 20	<< Apr \| May \| Jun >>	Current desk >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

May 21

Question about nuclear reaction

One thing I have always wondered about nuclear reactions: Does fissile material (such as plutonium) undergo a self-sustaining nuclear reaction simply by existing in a big enough mass (see critical mass) or does it need some sort of external stimulus? JIP | Talk 19:18, 21 May 2021 (UTC)[reply]

You don't need any external stimulus. Once you have a critical mass, a few spontaneous fissions or cosmic rays will start the chain reaction. But an external change can make a difference! This is because the critical mass isn't really a constant: it depends on things like temperature, shape and environment. (See Critical_mass#Changing_the_point_of_criticality.) For example, the critical mass for a ball of plutonium on its own is higher than the critical mass for a ball of plutonium surrounded by a beryllium shell (which reflects neutrons). So you can have a ball of plutonium that's below critical mass and more-or-less safe to handle. Then you drop something on it, and suddenly the critical mass of this new configuration is lower, so that a self-sustaining reaction starts. This is a criticality accident. That's what happened two different times with the same demon core in 1945 and 1946, when Harry Daghlian dropped a brick, and Louis Slotin dropped a screwdriver. --Amble (talk) 19:55, 21 May 2021 (UTC)[reply]

So as I understood it, once you have a big enough mass of fissile material, the nuclear reaction practically just starts by itself? JIP | Talk 20:03, 21 May 2021 (UTC)[reply]

Yes, because of spontaneous fission. You could imagine something like supercooling where there's a critical mass but the spontaneous fission rate is so low that the reaction doesn't start for a long time. I don't know if there are any materials that actually have that property, though. Some isotopes have the opposite behavior, where the spontaneous fission rate is too high. The reaction can start before the mass is all the way put together, resulting in a Fizzle_(nuclear_explosion). --Amble (talk) 20:12, 21 May 2021 (UTC)[reply]

This is believed to have occurred spontaneously approximately 1.7 billion years ago in several sites in Oklo, Gabon; see Natural nuclear fission reactor. --Lambiam 13:23, 23 May 2021 (UTC)[reply]

I feel it important to state that criticality =/= explosion. The so-called "demon core" did not explode, it just had a run-away chain reaction that dumped an whole mess of ionizing radiation into the immediate surroundings. Dahlgren and later Slotin weren't blown to bits, they died of radiation sickness. Wikipedia's article on nuclear explosions does a really shitty job of explaining how they work. What leads to the explosion is that the nuclear material is confined or compressed in some way; "out in the open", there's a way to dissipate the heat energy of a criticality event such that you don't get an explosion of nuclear material; you can look at things like the Chernobyl disaster or the Fukushima Daiichi nuclear disaster; in an unconfined environment, the heat from uncontrolled fission leads to a Nuclear meltdown, explosions from such accidents are caused the secondary effects of such heating, such as steam explosions or hydrogen explosions, and are never going to be the same mechanism as an actual nuclear bomb. The technical aspects of making an effective nuclear bomb are very complex, and most of the complexity comes from the means of effectively confining or compressing the nuclear material in such a way as to maximize the explosive effect of it, not in actually causing a criticality event. Making criticality is somewhat trivial (presuming you can get your hands on enough fissile material of sufficient purity); the first such controlled criticality experiment was done under the stands of an abandoned football stadium in Chicago; it was called Chicago Pile-1 because it was literally just a pile of bricks. Most of the work done at Los Alamos was about designing the bomb itself, not about making stuff reach criticality. --Jayron 32 11:59, 24 May 2021 (UTC)[reply]

Our Criticality accident article seems pretty good. File:Godiva-after-scrammed.jpg demonstrates that a run-away nuclear reaction can release a lot of energy (not just radiation to affect humans) and still not asplode. DMacks (talk) 15:04, 24 May 2021 (UTC)[reply]

The most dramatic such accident I can think of from that time period was the SL-1 accident. A supervisor standing on top of the reactor vessel was instructing two technicians as they reconnected a fuel rod during routine maintenance. One of the technicians accidentally moved the rod too far, leading to the reactor going supercritical, partially melting down, and causing a massive steam explosion. The explosion caused one of the rods holding the reactor together to strike the supervisor between the legs; it went through him stem-to-stern and pinned him to the ceiling. The technician who made the mistake was hit by the full brunt of the steam, and killed instantly as he was basically flash-fried. The other technician was standing far enough away to survive the explosion, but he succumbed to his injuries a short while later. Still not a nuclear explosion though. --Jayron 32 18:19, 24 May 2021 (UTC)[reply]

Yes, if the nuclear fuel can sustain criticality on its own. This usually means enriched, weapons-grade fuel. A lot of other nuclides can't support criticiality themselves, but need a moderator to "slow down" released neutrons so they will cause more fission events. Also while self-initiated criticality is possible, I believe most nuclear weapons in practice use neutron sources to kick-start the fission. This helps ensure the detonation goes off as desired and doesn't "fizzle", and ensures more of the fuel fissions and releases energy. --47.155.96.47 (talk) 19:13, 26 May 2021 (UTC)[reply]