I've mentioned before that I have doubts that Russia's long-range nuclear missiles have a very high rate of reliability after decades of demonstrable weakness in Russian military reliability. It is one reason I assume Russia is willing to cheat on theater nuclear weapon limits in order to build simpler nukes that can reach the important non-US targets.
But is America facing a similar problem since we halted underground nuclear testing to make sure our nukes actually work?
[After the Cold War], a 1992 Presidential Moratorium prohibited underground nuclear testing. Since then we've relied on computer simulations. To ensure that our weapons stockpile remains reliable and effective, we require the three nuclear lab directors and the commander of the Strategic Command to certify this to the president each year. ...
Now, suddenly, a highly responsible, scientifically sound report appears which seriously questions whether the computer codes used in assessing nuclear performance have the capability to accurately evaluate decades-old weapons which have undergone unknown aging changes.
And this computer simulation problem is separate from the Fogbank problem.
It might be that India, Pakistan, and North Korea are the only people who have reason to believe their nukes will explode, because they have actually detonated warheads in tests.
A dangerous problem of even resuming our underground testing is that if a number of the warheads don't work, will enemies draw the conclusion that none of them work?
