[dale_glass]

Despite it being so famous, and the memes, I still don't understand what Slotin was doing.

So I get it, it was a demonstration of how to perform an experiment. But I can't understand how the screwdriver makes any sense at all. What's being measured? What does success and failure look like? What does the experiment produce, what data in what format?

Because in my head, a proper experiment has data collection and precise measurements. Somebody's working on a data table that says "At position X, we measured value Y". But randomly wiggling stuff around with a screwdriver, I can't see how one can do anything of the sort. And I figure at this level, "more coverage = more radiation" is kind of a trivial point that doesn't really need to be demonstrated.

[lm28469]

It's all in the wiki: https://en.wikipedia.org/wiki/Demon_core

> It required the operator to place two half-spheres of beryllium (a neutron reflector) around the core to be tested and manually lower the top reflector over the core using a thumb hole at the polar point. As the reflectors were manually moved closer and farther away from each other, neutron detectors indicated the core's neutron multiplication rate. The experimenter needed to maintain a slight separation between the reflector halves to allow enough neutrons to escape from the core in order to stay below criticality. The standard protocol was to use shims between the halves, as allowing them to close completely could result in the instantaneous formation of a critical mass and a lethal power excursion.

> Because in my head, a proper experiment has data collection and precise measurements.

In your head yes, in early nuclear science it seems protocols weren't that important as long as it went boom in the end. As with many industries, regulations are written in blood

[atomicnumber3]

I've read about this in a few different mediums before and no it's not just that protocols weren't that important.

The guy doing this experiment was *notorious* for it and multiple other manhattan project people had already told him he was going to die if he kept doing it. But he had the kind of bravado and personality that he kept doing it.

So to be clear: all of the other people whose risk tolerance levels already had them handling weapons-grade plutonium as a career ALSO thought this guy was insane for doing this.

[vundercind]

He took a dip in a pool with a functioning nuclear pile some time before that just to avoid having to shut the thing down before doing some maintenance, taking a pretty big dose. He was a daredevil and had the kind of bravado of someone on a work site who scoffs at PPE and rolls their eyes when you tell them they need to wear a damn helmet. Those types usually end up having a bad time eventually.

[cpeterso]

> He took a dip in a pool with a functioning nuclear pile

xkcd published a What If? video about the consequences of swimming in a nuclear fuel pool: https://youtu.be/EFRUL7vKdU8

[immibis]

A pool containing nuclear fuel is different from a pool containing a running nuclear reactor with no other shielding, though.

[kqr]

A "pile" is not just a description of the thing, it is an actual formal term (old one!) meaning "reactor". So not just a bunch of radioactive matetial.

[itishappy]

Here it is in written form. The punchline is worth the read/watch.

https://what-if.xkcd.com/29/

[gopher_space]

> Those types usually end up having a bad time eventually.

It's entirely possible to build up skills allowing you to avoid using PPE, but every kid who sees you is being put at risk just so you can swing your dick around.

[MeetingsBrowser]

What skill do you develop to avoid the need for a helmet? Is it like a spidey sense, or do you hit yourself in the head so frequently your skull thickens enough to protect your brain from falling objects?

[lapetitejort]

I would like to learn the skill to dodge harmful prolonged sound waves. A technique similar to the safety squint, but with your ears?

[doubled112]

Don't mind me, boss. Just hitting my head off of this pipe to build up a resistance to physical trauma.

[gopher_space]

This was sort of a variation on the “don’t screw around, the idiots will see you” speech I give to the idiots. I missed the mark on tone.

[GuB-42]

It tends to be the opposite. Kids are usually fine, if they cross a construction site once, they would be really unlucky to have something fall on their head, even if they are careless. Professionals who work on site for thousands of hours will have something fall on their head eventually, even if they are careful. That's just probabilities. Take 10 times the risks for 1/10000th of the time and you are still 1000 times less likely to get injured.

[6SixTy]

OSHA 30 hour here: no the f** it isn't. You only lose an eye once before it's just gone. Hearing can only get worse. Some stuff will just kill you, some more slowly than others. Only literal children can bounce back from a what would otherwise be a fatal injury, but that's a very narrow slice of how you can get hurt.

[0x457]

> It's entirely possible to build up skills allowing you to avoid using PPE

Yes, but you only need to mess up once and your skill doesn't save you from other people mistakes.

[marcosdumay]

No, it's not.

Humans don't have perfectly consistent attention, and by the time you think you have any skill like that your attention is even less consistent than before you started "practicing".

[quickthrowman]

No it isn’t. Over an 80,000 hour career in construction/other dangerous field, you will eventually have an incident that will make you thankful for PPE.

[michaelt]

> multiple other manhattan project people had already told him he was going to die if he kept doing it

To me it seems quite reasonable that the people hired

to work on a bomb intended to kill 150,000 people in one go

against the backdrop of a war where 70-85 million died

might not place the greatest value on health and safety, and the sanctity of human life.

[Joker_vD]

On other human lives. They'd probably damn well care about their lives; otherwise, they wouldn't even make an effort to create "better weapons".

[GTP]

Regardless of the outcome, this still looks like a poor demonstration: what's the point of showing how it is done, if you're not following the protocol anyway? My understanding is that those in the room where nuclear experts, so they didn't need a demonstration to know that, the closer the two cores where, the higher the radiation.

[SoftTalker]

Same reason that chemistry professors demonstrate dramatic reactions in front of the class from time to time. It's fun, and keeps things interesting, even if you already know the chemical processes that are happening.

[dale_glass]

I've read that and it doesn't really answer those questions. How can you measure the core's neutron multiplication rate if you're not exactly controlling the distance? Isn't the measurement going to be all over the place?

[kstrauser]

In a demonstration, not an experiment, it’s sufficient to have the Geiger counter go clicky at different rates while the demonstrator plays the sphere like a theremin.

The point was to show it to people, not to collect data.

[eichin]

> plays the sphere like a theremin

Wow, that's a brilliantly horrifying image. (Are there other analogous ones? Does anyone do musical timing of building demolitions, or something like that?)

[Terr_]

I have the vague feeling like the last 40 years of movies must have contained at least one scene where a villainous figure makes conductor-motions as things explode to music, but I can't recall anything specific.

There was the V For Vendetta movie where landmarks exploded to the 1812 Overture, but no gesturing was involved.

[Terr_]

P.S.: I was wrong, the ending scene didn't have that because [spoiler omitted], but there's a much earlier scene [0] that fits the bill.

[0] https://www.youtube.com/watch?v=VCaT6tU7V8Q

[account42]

It's not just V For Vendetta, 1812 Overture is supposed to include explosions (artillery/cannon fire) in its composition.

[kstrauser]

The forgettable “Virtuosity” had Russell Crowe doing something like that, but it was awful.

[lm28469]

It was a boy's club with unlimited funding working on things that were never attempted before, a lot of things weren't exactly done by the books, even their original "safe" protocol would seem completely insane by modern standards. As long as it went boom in the end and they kept it secret I doubt they had many rules

[michaelt]

Imagine if you'd invented the world's first modern sink, in a world that had never seen a faucet or a plughole before. And you're training some new guys on the details of what you're working on.

Sure, some of that training is going to involve blackboard calculations and careful measurements.

But you're also probably going to demonstrate a sink to them and say "As you can see, when we turn this knob more hot water is added to the mix. Note how, after I put the plug in the plughole, the water level starts rising."

The purpose of the demo isn't to precisely measure the depth of the water or the temperature at the faucet or the angle the tap is turned to. It's just to let them see the thing in practice, so as they study it in theory they know what to imagine and how the model maps onto the real world.

[grecy]

Another thing I've always wondered - what would have happen if everyone in the room freaked out and just ran away, leaving the two reflector halves completed closed?

Would it have actually gone bang like a bomb, or more like just get insanely hot and give off an insane amount of radiation, but over the span of many seconds?

[sho_hn]

My understanding is that he was demonstrating a technique for how to bring the system to near supercriticality, without causing it. I.e. the objective was to look at the measurement devices they had and monitor them, and build an understanding of what the data was showing. This would then (in principle) be repeated by others with more specific objectives later.

Obviously they should've built a rig for that (at least), but I guess there was a "ain't nobody got time for that" attitude.

[dale_glass]

Right, but shouldn't distance be a critical part of such a measurement?

Like if we measure the amount of noise a device makes, we do it in a quiet room and at a standard distance. Without precision there's no useful data being generated.

So that's the part that I don't get. Shouldn't there be a screw being turned precise amounts, precisely made shims, or at least calipers be involved?

[StableAlkyne]

The honest truth is there's just a certain acceptable level of jank in a scientific lab.

Not everything needs to be measured to a high precision to be useful, and it's always a balance of how much effort you want to expend versus how useful that extra accuracy/precision is.

If all you care about is "when you're getting close to a critical mass, your instruments will look like this," you don't care that you have a wide swing in your data. You just want to show a difference from baseline.

[davisp]

Sometimes science doesn’t have to be precise to demonstrate a result.

Consider trying to measure feedback from a microphone and speaker. You don’t have to be an expert to know that there’s a quick change in system behavior when the microphone gets too close to the speaker.

[stetrain]

If the goal is to collect precise data and use it after the experiment to draw conclusions, update a model, etc. then sure.

If the goal is to demonstrate to observers how the neutron output (reaction rate) increases as the reflectors are moved closer together over the source, then that isn't really necessary.

This seems more like an incredibly dangerous version of a demonstration you might see at an interactive science museum or a classroom. You don't need precise measurements to demonstrate the relationship between two phenomenon.

[sho_hn]

I think it was more about being able to understand/read what the measurement devices were showing. The exact distance probably wasn't as important as the criticality could also vary with other variables (e.g. geometry).

As in, you're trying to understand the situation as "shouldn't they have precisely nailed down all the parameters, if the goal is to measure when X starts happening". But it seems Slotin was more demonstrating "this is what you're going to see on the monitors when stuff is close to going boom". It wasn't about "this specific distance is a safe gap" and more "here's how you can tell whether the gap is safe".

He was about to be reassigned out of the lab, and was demonstrating equipment to his designated successors.

[rocqua]

This is just proving 'move spheres closer means more neutrons'. It's something you quickly show someone to explain what you are going to do. The people watching will then get most of the same ideas you are suggesting, and figure out how to design a proper experiment around it.

Presumably the experiment to be done later is about characterizing different cores. They had already done it for this core, and wanted to teach the principles to others.

[Grimblewald]

The problem with this image of science is that in order to properly collect things we have to properly understand things we could be collecting. We can certainly do our best, but sometimes unexpected things happen. Take x-rays and their incidental discovery via the effect of x-rays on nearby photo-film. Totally accidental data collection from work at the bleeding edge, which was work that has transformed society as we know it. Another point is that in order for things like the lab equipment to be sleek and well built, we need to understand needs. This means that any research that meets your criteria is quite likely not at the cutting edge of anything. Most cutting edge labs I know look more like someone raided the hardware/electronics store to build some abomination than the Hollywood sleek and shiny labs you might be picturing. The sleek well built shiny labs you are picturing tend to be corporate labs and the like, doing work on safe and predictable things with well defined scopes. Translating existing knowledge into marketable products is a lot easier than discovering genuinely new knowledge.

[dale_glass]

This was after Hiroshima and Nagasaki, in a secret, government lab staffed by the utmost experts.

It wasn't the "oh look, something funny happens if I do this!" stage of experimentation. This was after they understood what they were dealing with well enough to build and successfully use two bombs. And Slotin was supposedly about to move elsewhere and was working on passing on knowledge.

That's why it's so weird to me.

[Grimblewald]

Experimentation is _ALWAYS_ at the "oh look, something funny happens if i do this" stage, otherwise it isn't experimentation/research but rather research/development. If we know what to expect, why experiment? Why waste resources?

[alt187]

Nothing. Richad Feynman said they were "tickling the tail of a sleeping dragon". The only goal of these experiments was to see how close they could flirt with criticality.

[Aardwolf]

Put a half-sphere reflective shield around a nuclear substance which will make the substance more reactive due to neutron reflections. Due to a slip-up with the screwdriver that was supposed to hold it up, accidentally fully drop down the shield, causing the too large radioactive reaction

I think with his reaction afterwards to remove it again, he saved the others in the room, but not himself