Centerfield

Interesting article. I guess we should all be glad that making nukes is a very difficult process. I think people are a little leary about growing complacent though. After-all North Korea wasn't supposed to be working on a bomb, until one morning we all woke up to find out that they already had one. People are frightened of a repeat of that...... especialy with a culture where martyrdom appears to be a popular pastime.

I can't seem to remember if the North Korean bomb was a plutonium or uranium device.
In order to make a plutonium device you need a functioning nuclear reactor. You "burn" enriched (but not bomb grade) uranium fuel rods then you harvest the resulting plutonium. It's a completely different process with it's own pitfalls.

It's been a couple of years since the North Koreans were caught, but their activities were detectible.

Here is a nice article on the North Korean program.

I can't seem to stop plugging numbers into that SWU equation. What's unclear to me is the dimension (if any) of the number yielded by the equation. If it's work/unit, what is a unit? Since the only variables are percentages, it would seem that there would be no dimension. Is there a constant, with a dimension, that has to be used for a given unit of product?

I don't think there is a dimensional unit in the normal sense of the word. What you're doing with a centrifuge is increasing the proportion of U-235 in the output gas. So a normal unit like "cubic feet" would cancel out.

(Please note: I've never actual worked at an enrichment plant so I'm guessing here.)

That was my impression, but I wasn't quite understanding the real significance of the SWU. I think I pondered it long enough to get it.

The thing that threw me off was the "calculate how many SWU it takes to make 25kgs of HEU" proposition. After thinking about it, I would term the SWU as a "process factor." You still have to know the amount of feed and the feed assay before you can really consider the SWUs required to make 25kgs of HEU, not to mention the required assay of the product.

I know this wasn't really the point of the post, but I figured a physicist would be willing to entertain the diversion.

The missing pieces for me are the derivation of the equation for SWU for the enrichment process and, more importantly, the details of what goes into determining the SWU capabilities of a centrifuge.

I would assume the latter involves the geometry and size of the centrifuge and the angular velocity it can obtain and how fast it can obtain it.

I would also assume that the equation for SWU required for a given enrichment accounts for whatever iterative processing is required given the centrifuge used. The process might require thousands of SWU, but the centrifuge might only be capable of a few SWU, so the tails would have to be used as feed in successive enrichments to obtain the needed amout of product from the original feed material.

I could probably go on for a while with everything I was thinking about, but why bother. This isn't a physics or engineering blog. I just needed to vent a bit to move past the minor obesession I've had over the last day or so about the concept of SWU. You just never know what's going stick in you craw, do you?

Actually your comment got me thinking about why the equation is unitless.
I have to wonder if when you do all the math, the centrifuge has to be a particular size, shape and speed for use with Uranium hexafluoride.

It's always those little “obvious” details that get left out when you talk tech.

My guess is the bigger, more numerous and faster the better, taking into account cost and safety. The more material you can put into the thing, all other things being equal, the faster you can enrich, but that doesn't enter into the SWU in my estimation.

It seems to me that size matters with regard to SWU only as it affects centripetal acceleration. The gas under a given acceleration for some amount of time should reach a (quasi-)steady state with a final gradient of U-235 concentration.

My guess is that a given centrifuge would have some acceleration-dependant time constant, tau, associated with it. The heavier molecules would migrate at an exponentially diminishing rate, so the process would be similar to charging a capacitor through a resistor, following a [1-(e to the t/tau)] curve. The final gradient, in practial terms, would be achieved after five time constants. So, given some feed assay, where you draw the line between tails and product would depend on the concentration gradient.

If I'm not out in left field, what you really end up doing with SWU is find out how much feed with a given assay you need relative to the amount of product, with some desired assay, you wish to produce in a reasonable amount of time given the number, size and SWU of the centrifuges you have.

Of course, I could be completely wrong about all of this.

That should be 1-(e to the -t/tau). Partial credit.

WHQ: You do realize that enriching uranium in your garage is frowned upon by the NRC. Not to mention the whole baldness and toothless side effects.

I fused my DNA with that of a cockroach many years ago, so I'm immune.

Cool!
Could you post the instructions.
I've been wanting to build a thorium reactor but was afraid of the side effects.
On the other hand maybe I'll try grisly bear dna. I always wanted to sleep through the winter.

ping