Humility

Please read this brief except from the Wikipedia entry on the Castle Bravo nuclear test:

The yield of 15 megatons was two and a half times what was expected. The cause of the high yield was a theoretical error made by designers of the device at Los Alamos National Laboratory. They considered only the lithium-6 isotope in the lithium deuteride secondary to be reactive; the lithium-7 isotope, accounting for 60% of the lithium content, was assumed to be inert.

However, when a lithium-7 isotope is bombarded with energetic neutrons, it captures a neutron then decomposes to form an alpha particle, a tritium nucleus, and the captured neutron. This means more tritium was produced than expected, and the extra tritium is fused with deuterium. In addition to tritium formation the extra neutron released from lithium-7 decomposition produced a larger neutron flux. This caused more fissioning of the uranium tamper and increased yield.

… The test used lithium with a high percentage of lithium-7 only because lithium-6 was then scarce and expensive; the later Castle Union test used almost pure lithium-6. Had more lithium-6 been available the usability of the common lithium-7 might not have been discovered.

This story should give you pause the next time you’re asked to accept a result that can’t be empirically tested, that was arrived at after a long chain of reasoning, that depends on the output of one or more models, or that is simply vouched for by all the experts.

Whoops

The lithium-7 goof was not a small or irrelevant error. It was not made when working on an unimportant problem. The scientists and engineers who made it were not starved for resources, or operating under any sort of moral hazard. The mistake wasn’t even centered in an obscure part of the mechanism. If it were possible to get things right as a general matter, this would never have been missed.

Code

My thoughts turned to Castle Bravo after finding another surprising bug in my code. I’d like to think that my views on rationality and its limits grow out of my experience with code, that any good engineer would apply Knuth’s famous caution to “[b]eware of bugs in the above code; I have only proved it correct, not tried it” to other areas of life. Ultimately, I must admit that this is unlikely to be the case; there are way too many lefty CS types to believe that the discipline imparts any wisdom. It would be nice if it did, though.

Editorial note: The Castle Bravo test is not the one depicted in my blog header; that’s the Ivy Mike test.

Share and Enjoy:
  • Twitter
  • Facebook
  • Digg
  • Reddit
  • HackerNews
  • del.icio.us
  • Google Bookmarks
  • Slashdot
This entry was posted in Jack Handy. Bookmark the permalink.

Comments are closed.