Wednesday, August 25, 2010

A Primer on Nuclear Weapons Development

With the idiotic question of whether or not to bomb Iran's uranium enrichment facilities being passed around the news outlets these days (wanna make sure Iran feels the need to make a nuclear deterrent? BOMB THEM), I figure Train readers could use a refresher on just how these weapons are produced.

First off, let's set one thing straight: no one – up to and including Iranian leaders and scientists – knows if and when Iran will produce enough weapons-grade uranium to make a bomb. Any set date you see printed anywhere is an estimate based on insufficient information, and probably advanced primarily to advocate a particular political viewpoint or action. But I will say that unless things change radically, Iran will very likely have enough material for a bomb within the next ten years. Although I'm skeptical that Iranian nuclear weapons will radically alter the equations of world power, a new state entering the nuclear club is always alarming – and of course, overreaction on the part of the United States or our allies is as much of a danger as the bomb itself.

But now the facts:

You can make a bomb from either plutonium 239 or uranium 235 (as well as some other isotopes, though none of them work as well). U235, although it exists naturally, makes up less than 1% of natural uranium (most of which is U238), and because it is the same element as U238 (and thus chemically identical) it is tremendously difficult to separate. Plutonium (the first man-made element) is a byproduct of U238 fission (usually in a reactor), and because it is a different element it is quite easy to chemically separate from contaminants. Neither production process is quick or easy, but the difficulty of U235 separation makes it considerably more expensive than plutonium.

Plutonium bombs, however, are vastly more complex and finicky (though also more powerful), and of course safely working nuclear reactors is non-trivially difficult. The US has managed this fairly well (though we still have tons of radioactive waste to deal with), but early Russian reactors were deathly dangerous: the majority of reactor workers received excessive dosages during the first year of operation, roughly 7% in the life-threatening greater-than-100-rem range, as compared to 3-7 rem during the LIFETIME of a US worker.

A good deal of the difficulty with making a nuclear explosion is the process of creating the critical mass that needs to be present for a chain reaction to occur. Uranium bombs can generally be extremely simple in their design because U235's chain reaction proceeds relatively slowly and reliably, and so you can simply fire one barely sub-critical mass at another with a gun. This will detonate with such fantastic reliability that they didn't even test the Little Boy (uranium) design before dropping it on Hiroshima.

Plutonium is more inclined to fission spontaneously and its fission proceeds more quickly, so the gun design won't work: the parts of the two subcritical masses that are closest to one another will fission first as they are brought together, reducing the core below true critical mass and blowing the two pieces of the bomb apart again before they have a chance to come completely together and detonate. It will explode with considerable force, but nothing like that of a full critical mass. Los Alamos scientists solved this in the '40s by strapping complexly shaped explosive "lenses" around a solid mass of plutonium and setting them all off at once, which compressed the core sufficiently to create the density needed for a chain reaction. But Iran appears to be working towards a uranium bomb for now, so this is irrelevant for the moment.

The more U238 (the non-chain-reacting kind) you have in the mix, the more material you need to reach critical mass. A uranium-based bomb requires something like 50 kg of U235 to be compressed into a space smaller than seven inches in diameter. A contaminant like U238 acts as a tamper on the reaction while simultaneously adding excess material that makes it physically more difficult to compress the core into such a small space. 90% U235 is the number that gets thrown around a lot, and this is why: below that, it just gets to be too difficult.

Though it is impossible to know for certain, Iran seems to be mostly pursuing U235 separation through a vast sequence of centrifuges. Like other refinement methods, you need hundreds or thousands of centrifuges working at peak efficiency in a "cascade" for long stretches of time to produce kilogram-levels of refined U235. But refinement is not a linear progression: it tends to be more difficult at lower concentrations than higher ones. Iran has stated that they intend to produce 20% refined U235, which is a huge leap up from the ~4% required for civilian nuclear reactors, and I believe closer to 90% than 4% is to 20%.

In this sense, there isn't a solid line between civilian and military refinement: the components are the same for each. But also, you don't need large numbers of high efficiency centrifuges to produce 4% U235, so essentially Iran is flirting with this line to see how much the rest of the world will tolerate.

If they are pursuing a straightforward U235 gun-type bomb, they don't need much of anything besides refined uranium to make it work: the technology required for the gun portion isn't significantly different from a standard artillery piece. Because of this, things like rushed advances in warhead and missile designs that are capable of supporting the size and weight of a nuclear payload are perhaps the best indication of nearing capacity. The list of components for a plutonium or a hydrogen bomb, however, is considerably longer and more specific, and some of the pieces would probably be trackable from abroad. A lot of it is classified, obviously, but components like the initiator used in a plutonium bomb would likely only be useful for a weapon. Then again, it's science: who knows what these things can be innocently used for?

In short, an Iranian bomb is probably inevitable, but also probably not right around the corner: they almost certainly still face tremendous challenges.


  1. Fascinating stuff.

    I also liked this disclaimer: "wanna make sure Iran feels the need to make a nuclear deterrent? BOMB THEM"

    If Israel decided to attack them unilaterally I honestly think it would be a more self destructive move than the Iraq war... and that's saying something!

  2. A valid point, but with the existing support-Israel-no-matter-what-or-you're-antisemitic logic floating around in American politics, do you think they'd really have to do it unilaterally?