Hacker News Reader

Comment by blantonl | Hacker News Reader

These systems are so vastly complicated, old, and rarely if ever launched. These aren't like data center generators which have testing schedules etc, and STILL there are failure points.

I really wonder what the failure rate would be if they were all actually launched today. And I mean failure, from not lifting off, to failure in flight, to misguided warheads etc.

View on HN

But they do have testing schedules. They certainly tests the electronics regularly. And every so often they randomly pick a missile and launch them from Vandenberg.

https://www.spaceforce.mil/News/Article-Display/Article/3796...

> I really wonder what the failure rate would be if they were all actually launched today

I hope we will never find out. But certainly there would be many duds. But this is calculated into the effectiveness of the system by simply having more missiles. That is how it achieves its goal of dettering a would be attacker. (Not even talking about how there are two other totaly separate legs of the nuclear triad with dissimilar personel and technical solutions.)

> randomly pick a missile and launch them from Vandenberg

Jeff! Did you remember to take out the warheads? Jeff?!

Somewhat soberingly, earlier in the war with Ukraine, Russia managed to launch a ballistic missile with a concrete "Warhead simulator" installed, literally a mass designed to test the launch vehicle without using a nuclear device.

If that was an accident, that means they didn't properly know which warheads are where. That's.... upsetting.

That reminds me of the 2007 incident when the Air Force meant to transfer twelve unarmed cruise missiles from North Dakota to Louisiana but loaded up six nuclear-tipped missiles by mistake. The missiles remained unprotected on a B-52 for 36 hours before someone noticed the missiles were nuclear. A whole lot of people including 8 generals got in trouble over this.

https://en.wikipedia.org/wiki/2007_United_States_Air_Force_n...

You can watch video of the most recent test, a little more than a month ago, of a Minuteman III. This is a short clip:

https://www.youtube.com/watch?v=zUg7x1zo7D0

They can test the missiles and reentry vehicles, everything except the nuclear warhead. The closest those come to a test is a supercomputer simulation, since those tests are forbidden by treaty.

Minuteman III has an excellent but not perfect [0] failure rate. Some other older systems, like the UK's submarine launched Trident missiles... not so much.

[0] https://www.airandspaceforces.com/icbm-test-failure-nuclear-...

Th UK Trident 2 missiles are literally the same missiles used by the US submarines operating in the Atlantic - both sets of submarines are supplied by a shared pool:

https://en.wikipedia.org/wiki/UGM-133_Trident_II

However, the warheads on the UK missiles are designed and manufactured by the UK.

Interesting. Are recent US missile tests at sea as bad as the recent UK tests? I don’t remember seeing news to that effect.

I suspect that just indicates that it wasn't just the missiles that were causing the failures...

I had the same thought. They lovingly take Minutemen out of their silos and launch them after reassembly from Vandenburg AFB or someplace for a test. I’m pretty sure trident tests happen at sea.

Strategic weapons are tested regularly, except for the warhead. The Trident II D5 has had about 200 tests in the last 35 years. That's about the same number of test missiles expended as are actively deployed at any given time. While there are a theoretical maximum of 344 deployed at any one time (14 Ohio-class, each with 20 tubes, plus 4 Vanguard-class with 16 tubes) some number of those boats are unarmed in refit at any given time.

They have operational models with statistical failures included so they understand what the interception and failure probabilities likely are for each item at component level. You can design that into a system and test for it.

Obviously you can't factor in unknown problems but that's what drills and test flights are for.

Air force takes one at random, transports it to Vandenberg sans warhead and lights the fuse to see if it works. Or at least they used to. I also heard that the Soviet equivalent of this random testing process was to simply retarget a random missile and launch it straight out the operational silo.

Reply on news.ycombinator.com