<jercos> anyway, cheese details?
<Shrdlu> I guess I can explain neutron activation analysis - here's the version I give on reactor tours.
<jercos> which is probably somewhere near low enough of a level for me to get :P
<Shrdlu> So everybody here knows what an atom's made of, right? Nucleus in the center made of protons and neutrons, electrons flying around the outside.
<jercos> (insert power point slide with smiley-faced elementary particles)
<RK> e:)
<RK> n_n
<Shrdlu> In most atoms found in nature, the nucleus is what we call 'stable' -- that is, the protons and neutrons are in the right balance.
<RK> p_p
<Shrdlu> If you have too many neutrons or too few neutrons, the atom is radioactive -- it can't stay like that forever. Eventually it will decay, emitting some kind of radiation and turning into some other type of atom.
<Shrdlu> For example, it could emit an alpha particle, which is two protons and two neutrons stuck together.
<Shrdlu> Or it could emit a beta particle, which looks like either an electron or a positron.
<RK> q: q: q:
<jercos> yay antimatter.
<Shrdlu> Most likely it will also emit energy in the form of a gamma ray.
<jercos> the classic particle used for making things grow, shrink, and gain superpowers.
<Shrdlu> Now, the cool thing about this is that the energy of that gamma ray is characteristic of the original nucleus.
<RK> ~~~~~:D>-<
<Shrdlu> So if you started with Sodium-24, you'll get gamma photons of one energy. If you started with Cesium-137, you'll get a different energy.
<Shrdlu> What that means is, you can take a radioactive sample and put it under a gamma spectrometer, and get a graph of the spectrum of gamma rays being emitted.
<Shrdlu> You look at that spectrum and you'll see peaks corresponding to these characteristic energies -- look that up on a table, and you can see exactly what you're dealing with.
<jercos> so, it's kinda like a heating up a sample and using a spectrometer, but on a nuclear scale?
<RK> Uuq~~~ D:
<Shrdlu> Essentially! But the spectrum you're looking at is entirely gamma rays. So you'll only see it with radioactive samples.
<Shrdlu> Stable atoms never decay, so they aren't emitting gamma radiation at all.
<Shrdlu> Which is inconvenient, because most of the time, if I want to perform elemental analysis on some kind of sample, it's not radioactive.
<Shrdlu> For example, I could take some drinking water and ask whether there's any arsenic in it.
<Shrdlu> This method works for extremely tiny trace amounts of various elements, by the way.
<Shrdlu> You can do a quantitative analysis, in some cases with parts-per-billion accuracy.
<jercos> cool.
<Shrdlu> But, okay, it only works for radioactive stuff.
<Shrdlu> That's where the reactor comes in.
<jercos> that's good accuracy. Water treatment only goes down to parts per million accuracy...
<Shrdlu> A nuclear reactor is essentially just a place where you have a crapton of neutrons flying around.
<Shrdlu> (If anyone's interested in how the reactor actually works, I can give that spiel too when I finish this one.)
<jercos> sounds good.
<RK> n_n n_n n_n n_n n_n n_n n_n n_n
<Shrdlu> So if I take a sample of something that's not radioactive and stick it down next to the reactor, it'll get smacked by all those neutrons.
<Shrdlu> So let's say I start with a sodium atom.
<RK> (>n_n)>H2O<(n_n<)
<Shrdlu> Sodium's stable abundant form is Sodium-23 -- all sodium atoms have eleven protons, and sodium-23 has 12 neutrons. 11+12=23, hence the name.
<jercos> hmmkay.
<Shrdlu> Now, if my sodium atom happens to get hit by a neutron, there's a decent chance that it will absorb it.
<Shrdlu> It doesn't have to happen that way -- the neutron might just bounce off. Each nuclide has a different probability of absorbing a neutron, which is why some things become more radioactive than other things when you irradiate them.
<RK> (>n_Na
<Shrdlu> Sodium activates pretty well, so it's fairly likely that our Na-23 atom will absorb that neutron.
<Shrdlu> So that means it's now a Na-24 atom.
<jercos> which now *is* radioactive?
<Shrdlu> Thing is, Sodium-24 is radioactive.
<Shrdlu> Exactly.
<jercos> that's a neat trick.
<Shrdlu> Now, Sodium-24 has a half-life of about fifteen hours.
<Shrdlu> What that means is, each atom eventually has to decay, but there's no particular time that has to happen. It's completely random, and completely unpredictable.
<Shrdlu> But on average, Sodium-24 atoms decay after fifteen hours.
<RK> (:e Na~~~~:D>-<
<Shrdlu> So that means if I have a chunk of radioactive sodium, after fifteen hours half of the atoms in it will have decayed.
<Shrdlu> So I'll have half as much Sodium-24 as I started with.
<Shrdlu> After thirty hours, I'll have half of that -- so a quarter of what I started with.
<Shrdlu> But anyway. So we've taken our stable sample and stuck it in the reactor, making it radioactive.
<Shrdlu> We shut down the reactor and pull out the sample, and stick it under the gamma spectrometer.
<Shrdlu> And we observe peaks at particular gamma energies, so we know exactly what radioactive elements we have.
<RK> :O Mg
<Shrdlu> We subtract one neutron to account for the fact that we probably made it radioactive, and what we're left with is a really good elemental analysis of what we started with.
<Shrdlu> But the best part is that after a while -- in this case, about a week -- almost all of the radioactive atoms have decayed, and the sample is no longer radioactive.
<Shrdlu> The atoms that used to be radioactive have decayed to something else, but of course that's the vast minority -- for all intents and purposes, you've got your original sample back.
<jercos> good stuff...
<Shrdlu> So the story with the cheese went basically that way. Somebody wanted to perform neutron activation analysis on a cheese sample.
<Shrdlu> Did some quick calculations to estimate how radioactive it would be when it was pulled out, and went ahead and irradiated it.
<Shrdlu> Trouble is, they botched the math or something (this was before my time, I don't know the details) and it came out far too radioactive, because of all the sodium -- like I said, it activates easily.
<Shrdlu> So what they got back was a chunk of extremely radioactive cheese that would take a while before it could be disposed of.
<RK> Na~~~~~8)<-<
<Shrdlu> Long story short, we wound up using it as a check-source around the lab -- it's good to have something radioactive around to test your instruments with, so for a few days we were using this chunk of extremely hot cheese.
<RK> (>o_o)>~~~~~~ D:
<jercos> hah, that is like, the best story about a nuclear reactor I have ever heard.
<RK> ~~~~~<(o_o<) ,_
<Shrdlu> We tend to accumulate a lot of those -- it's that kind of workplace, for some reason.
<Shrdlu> The folks who work there are extremely good at what they do, but they're also mostly ridiculous.
<RK> (>o_o)>~~~~ (>o_o)>~~~~
<jercos> it's probably a trend. people who are incredibly skillful, and yet hilarious to be around.
<jercos> so, ah, the reactor itself?
<Shrdlu> Oh right!
<Shrdlu> Actually, it's surprisingly simple.
<jercos> I would imagine.
<Shrdlu> So I mentioned how you can take an atom and whack it with a neutron.
<Shrdlu> In most cases, if the neutron is absorbed, the atom becomes radioactive.
<jercos> unh hunh.
<RK> [|||]
<RK> (reactor core)
<Shrdlu> But in a few cases, with certain heavy nuclei, if the neutron hits the nucleus with just the right energy, the atom will undergo fission.
<Shrdlu> The nucleus splits apart into two large chunks, each of which goes flying off with a tremendous amount of energy.
<Shrdlu> And you also get two or three stray neutrons which fly off at high speed as well.
<jercos> and that's why fizzy drinks are better than fission drinks.
<Shrdlu> Now, the thing is, those stray neutrons could hit other fissionable atoms.
<RK> (< >) ,`(>n_n)>
<Shrdlu> Which would then fission, producing more energy and more neutrons.
<Shrdlu> And those neutrons would cause even more fissions.
<Shrdlu> So what you get is a self-sustaining nuclear chain reaction.
<RK> <(n_n<) <(n_n<) (< >) (>n_n)>
<Shrdlu> Now, in some reactors, the interesting part is all that energy being produced.
<RK> [|||]
<Shrdlu> On the large scale, it manifests itself as heat.
<Shrdlu> If you use that heat to boil water, you can produce steam, which drives a turbine, which powers a generator, which produces electricity.
<Shrdlu> That's essentially how a nuclear power plant works.
<jercos> I'd hit that. with a beryllium rod.
<RK> [|:|]
<Shrdlu> (As you're probably aware, the fissionable material used in these cases is uranium-235.)
<Shrdlu> (Which fact I failed to stick in somewhere earlier.)
<jercos> ah, 235. a magical number. of joy.
<RK> (>U_U)> <(n_n<)`,
<Shrdlu> But the place I work is a research reactor. We don't boil water, and we don't have a turbine, because we're not interested in producing electricity.
<Shrdlu> We operate at much lower power levels -- a typical power reactor might run at three gigawatts; our maximum licensed power is 250 kilowatts.
<RK> <(n_n<)`, (>? ?)>
<Shrdlu> What heat we do produce, we carefully cool back off with a cooling system that releases the heat to the environment.
<Shrdlu> What we're actually interested in is the neutrons.
<Shrdlu> Some of those won't go on to cause more fissions; some will leak out of the core.
<Shrdlu> So we can put our experiments in some equipment set up in and near the core, and they'll get hit by all those neutrons leaking out.
<RK> [|:|] `,(>n_n)> D:
<Shrdlu> Trouble is, what I described isn't actually a reactor.
<Shrdlu> What I described, under some circumstances, is a nuclear bomb.
<Shrdlu> The reason being that the way I described it, the chain reaction isn't just self-sustaining, it'll accelerate.
<Shrdlu> It'll accelerate extremely quickly -- each of those generations takes place in about 10^-5 seconds.
<RK> ~~~~{:*:]~~~ D: D: D:
<Shrdlu> So you've got more and more heat building up faster and faster, and that's how a nuclear explosion work.s
<jercos> fun.
<Shrdlu> In a reactor, obviously, there are mechanisms to keep that from happening -- one is the control rods.
<RK> |||
<jercos> :D
<Shrdlu> We have three rods that we can move in and out of the core. They're made of boron, which is a neutron poison; basically, all that means is that it absorbs neutrons really well.
<RK> (>n_|||n<)
<Shrdlu> So when the control rods are fully inserted in the core, there's a pretty good chance that a neutron coming off a fission won't hit another uranium atom; it'll hit a boron atom instead.
<jercos> here I thought boron would be boring. :P guess not!
<Shrdlu> What that means is that while the control rods are fully inserted, the chain reaction can't sustain itself.
<Shrdlu> As we pull the control rods out, the probability of a fission neutron hitting another uranium atom increases.
<Shrdlu> And so that means the chain reaction gets stronger.
<jercos> so rods in = full stop
<jercos> rods out = warp 6
<Shrdlu> Not quite full stop - there are still a few neutrons bouncing around, we can't stop them completely. But mostly.
<jercos> (or at very least, full impulse power)
<RK> (>n_n)> (>n_n)> (>n_n)> (>n_n)> (>U_U;)> <(n_n<) <(n_n<) <(n_n<) <(n_n<) <(n_n<)
<Shrdlu> So as we pull the rods out, we'll eventually reach a particular point where the number of neutrons going on to cause fissions in a particular generation is equal to the number of neutrons that started that generation.
<jercos> so what happens when two neutrons hit each other? anything magical?
<Shrdlu> It's really unlikely. If it happened, they'd just bounce off.
<jercos> hmmm, okay.
<RK> `,(>n_n)> <(n_n<).`
<Shrdlu> Anyway, at that point, the chain-reaction is precisely self-sustaining. The neutron population is neither increasing nor decreasing.
<RK> <(n_n<).` `,(>n_n)>
<Shrdlu> When that's the case, we say the reactor is critical.
<RK> [|:|] `,(>n_n)> D:
<Shrdlu> If we pull out the rods just a hair, the chain reaction is more than self-sustaining; you'll have more neutrons out of each generation than you started with.
<Shrdlu> So the neutron population is slowly increasing, and we say the reactor is supercritical.
<Shrdlu> Pull the rods out more and the neutron population starts increasing faster, as does your power level.
<RK> over 9000!
<Shrdlu> So what we ordinarily do is go supercritical for a few minutes until the power level is where we want it for the day, then put the rods back in until we're critical again.
<RK> [:*:]~~~~~:D>-<
<Shrdlu> At that point the chain reaction will hold steady at that power level, and we can do whatever experiment we want to do.
<jercos> WHAT HAS SCIENCE DONE?
<jercos> cool enough.
<RK> ^^
<Shrdlu> The other cool thing, which is a research-reactor-only feature that power reactors don't have.
<jercos> :D like windows XP pro lets to use the control panel.
<jercos> you*
<Shrdlu> Our reactor is inherently safe -- it can't have a runaway chain reaction. Not just because of the safety features built into it, but because it was designed such that the laws of physics prevent it from happening.
<jercos> how so?
<Shrdlu> So I mentioned before that a neutron hitting a uranium atom will cause fission only if it's moving with just the right energy.
<jercos> unh hunh?
<RK> [|||] (:
<Shrdlu> If it's moving too fast or too slow, it won't cause fission.
<jercos> it doesn't work if it's moving too *fast*?
<Shrdlu> Specifically, the neutron has to be in thermal equilibrium with its surroundings -- it has to have the same energy as everything around it.
<Shrdlu> Exactly!
<jercos> hunh, cool.
<Shrdlu> There's a particular energy peak where you get fission -- too far left or right, it's a no go.
<Shrdlu> Now, the interesting part is, a neutron coming OUT of a fission event has far too much energy to cause fission.
<Shrdlu> So in order to actually get a chain reaction, you need to slow them down to the point where they're at thermal equilibrium and they can cause another fission.
<Shrdlu> The way you do that is with a moderator. In our case, we use zirconium hydride.
<Shrdlu> The zirconium isn't actually the interesting part, it just forms a matrix for the hydrogen atoms.
<Shrdlu> The hydrogen itself is the actual moderator.
<Shrdlu> Picture a neutron and a hydrogen atom for a sec.
<RK> n_n H
<Shrdlu> Hydrogen nucleus is just a proton -- so it's the same size as a neutron.
<RK> (>n_n)> <(p_p<)
<Shrdlu> If a neutron with high energy comes flying at it, they're going to collide.
<RK> (>n_p<)
<Shrdlu> Picture two billiard balls, you shoot one into the other. The second one speeds up, but the first one slows down.
<Shrdlu> So a high-energy neutron, by bouncing off a whole bunch of hydrogen atoms, will eventually slow down to the point where it's at thermal equilibrium.
<Shrdlu> Make sense?
<jercos> sure!
<RK> p_p p_p p_p n_n;
<Shrdlu> At that point, it will be able to cause fission and you'll get more fast neutrons, which will have to undergo moderation before THEY can cause fission.
<Shrdlu> Now, this is true at any reactor anywhere.
<RK> p_p p_p p_p (>n_n)> <(U_U;<)
<Shrdlu> The reason our reactor is cool is that zirconium hydride has what's called a prompt negative temperature coefficient.
<Shrdlu> That means that as the moderator gets hotter, it quickly stops being an effective moderator.
<Shrdlu> Which makes sense -- if those hydrogen atoms have higher energy, they won't be able to slow down the neutrons effectively.
<Shrdlu> So picture an uncontrolled chain reaction for a second.
<Shrdlu> Our reactor has mysteriously lost all control and safety mechanisms.
<RK> ~~~~{:*:]~~~ D: D: D:
<Shrdlu> The control rods have spontaneously vanished, and the chain reaction is unchecked,
<jercos> so it's self limiting. as the reactor starts going faster, it slows it down, like a governor on a steam engine?
<Shrdlu> Exactly!
<Shrdlu> As the neutron population increases, it causes more fissions, which produce more heat.
<Shrdlu> That heat makes the moderator less effective, which stops fissions from being produced.
<Shrdlu> The reactor actually shuts itself down.
<jercos> hunh, wow. cool. or hot, as the case may be.
<RK> [|||;]
<Shrdlu> Every reactor in the US has a negative temperature coefficient. Most research reactors have a prompt negative temperature coefficient, which effectively means that it shuts itself down faster.
<Shrdlu> By comparison, the reactor at Chernobyl had a positive temperature coefficient.
<Shrdlu> That meant that as the core got hotter, it got more efficient.
<Shrdlu> Which was only one of a whole bunch of design and protocol failures at Chernobyl.
<RK> <(U_U;;;<) <(n_n<) <(n_n<) <(n_n<) <(n_n<) (>n_n)> (>n_n)> (>n_n)> (>n_n)> (>U_U;;;)>
<jercos> note to self: don't make a machine that gets better at blowing up the closer it gets to blowing up
<jercos> amirite?
<RK> yeah
<Shrdlu> Yeaaaahhh...
<Shrdlu> If you decide to do that, though, you should probably train your operators in how to run it safely.
<Shrdlu> That might be a good idea.
<RK> the whole core melted down, turned to lava, flooded everywhere
<Shrdlu> @RK: In my line of work that's a pretty good sign that you've fucked something up.
<RK> and now it has black mold growing inside which uses radiation as a source of energy
<jercos> wait, what?
<RK> hehe
<Shrdlu> That's what I've heard. I don't know anything about it but it scares the shit out of me.
<RK> http://en.wikipedia.org/wiki/Radiotrophic_fungus
<jercos> http://digg.com/general_sciences/Black_Fungus_Found_in_Chernobyl_Eats_Harmful_Radiation
<jercos> oh god.
<RK> evolution is impressive
<jercos> and holy lagg.
<jercos> firefox has slowed my computer to a halt.
<RK> meep
<jercos> and my computer's only loaded to like 60%, so i assume it's not doing anything.
<jercos> there we go.
<jercos> anyway, would that be related to your algae?
<Shrdlu> Jesus I hope not.