An Unanswered Question at the Heart of America's Nuclear Arsenal | Hacker News Reader

An Unanswered Question at the Heart of America's Nuclear Arsenal

blogs.scientificamerican.com

25 pointsvo2maxer37 comments

I think anyone who has any engineering experience would suspect that never testing something will eventually result in it not functioning.

Please, please DO NOT test a nuclear weapon. The world will be far better off if everyones nuclear weapon turned out to be a dud.

Nukes may shorten an otherwise prolonged war. "Better off" is obviously very subjective, but I don't see how nukes aren't better off if they end up saving lives. Let's also consider all the direct wars avoided in the last ~75 years, an outcome of nuclear MAD.

The reality of our situation if that we will keep fighting each other, that's a given. Best to test our nukes to make sure they work. Nukes that won't work can't save lives.

MAD is indeed an equilibrium. But it is an unstable one. That is the problem. So far the US-USSR/Russia detante has worked, but it might not work in other situations. India and Pakistan are a case in point. They have fought at least two limited conflicts (the most recent one being last year,) despite being declared nuclear weapons powers. The possibility of catastrophic miscalculation in the fog of war is very, very real.

True, but if just one side's nuclear weapons are actually duds, and the other side figures that out, that's not such a great situation.

I know. I know. But imagine the comedy of epic proportions that would ensue if WW3 consisted of all sides desperately pushing the big red button and nothing happens!

Yes, that would be hilarious. Instead of half of the worlds population being wiped out by nukes, we'd be wiped out by chemical weapons, airstrikes, and ground warfare.

Alternative scenario: The imbeciles are lynched in the street by angry crowds, many enquiries follow, life continues as normal.

I know anyone alive today, not suffering from a mental illness, that would accept their leaders starting a global nuclear war.

I would wager roughly 49% of the American voting public would be fine with Trump starting a Nuclear war. They didn't bat an eye at him assassinating an Iranian general. Or praising leaders like Kim in North Korea. They don't mind his rampant corruption to date or abuses of power.

Mind you, that's not a majority of the actual population given how low voter turnout is, but that's enough to be really scary.

I think the number is far too high, in my experience partisan people sober up remarkably quickly once their ass is on the line.

None of those are anywhere near as destructive as thermonuclear weapons.

What if that's already happened?

Russia and China remelt their pits every few years. That's more or less a public knowledge. I'm surprised that US doesn't do the same

As a child I thought it would have been completely hilarious if the leaders of the free world solemnly decided to destroy it by a large scale nuclear exchange, only to find much to their surprise that none of the warheads functioned due to age and lack of testing.

I say leave the missiles and warheads exactly the way they are and let anyone with the temerity to order a first strike deal cognitively with the risk that the world might survive their decision.

I think that might be the established thought behind most of the world's nuclear arsenal. No one knows if they might all work. But the fact that some countries have them will give their adversaries second thought. See North Korea as an example. They tested one crude device few years ago. But no one outside of the NK regime know for sure if their nukes are deployable with a delivery system. But it doesn't matter. The fact that they have the capability makes even the U S pause and consider if regime change is a viable option there.

That's a special case.

The most likely case for somebody other than a great power making a nuclear arsenal, and ever thinking of nuclear first strike will be somebody:

1. Living in a very bad neighbourhood

2. Being extremely secretive

3. Not having reputation for bluffing, considering winning a war a matter of existential national importance

If we are to see military use of nukes happening this century, most likely you will see it where you never saw it coming, in a scenario of extreme power imbalance.

Saying this again, for anybody who is desperate enough to bet on a nuke for a military advantage, is already in such a dire straight, that they are already pass the point of using atomic weapons as a deterrent.

They will hold it as the top military secret till the "day X"

A world with broadly-held would have civil wars fought with them.

The blame slinging seems a bit unwarranted given the understandable constraints on detonation tests.

The two plauisble outcomes even as a laymen are either the plutonium's decay results in less energetic nuclear explosions as lost to atomic decay or the breakdown products "poison" it to a dud which cannot sustain fission. The other unlikely result is a "meltdown state". The decay chains would presumably be calculable without any destruction.

The composition could potentially be destructively tested by centrifuges and chromatography suitable for weapons grade radioactives - possibly resulting in just some low level possibly atomic waste instead of a detonation.

That has more to do with metallurgy than with nuclear physics. It's clear that decay and fission products need at least 100 years to accumulate in any significant amount to affect munition performance.

Pure plutonium has 6 metallurgical phases, and alloys used in pits add even more to metallurgic issues.

Imagine an implosion design going through a brittle failure instead of an orderly compaction of a hollow pit? Or worse, one side of a pit went through phase transition, and the other don't and you don't even get a fizzle.

On an unrelated note, I highly recommend this episode about nuclear weapons from Dan Carlin's Hardcore History: https://www.dancarlin.com/hardcore-history-59-the-destroyer-...

I have this naive fantasy where every country that has nuclear weapons independently and secretly came to the conclusion that keeping them operational is ultimately more dangerous than not (due to accidents, human error, clinically insane leadership, ...), and just removed the detonators and unplugged the big red buttons. The nuclear deterrent remains, but is now actually based on a collective fiction. We've discovered the ultimate defensive weapon, one that only needs to exist in our imaginations ;)

The major problem with 'global zero' initiatives is that if one nation can achieve nuclear primacy they become the de-facto owners of the world. No one wants to risk that.

Alternatively, the problem with not pursuing global zero is that the probability that someone will eventually use them approaches 100% in long enough timescales.

When pits degrade, what is happening? Is it merely a function of the radioactive isotopes decaying below the point at which they'll detonate in a big ol' fireball?

Pu is an unusual metal. It has at least 6 different solid allotropes (phases), each of which have different density, hardness etc. The phase can change over time, kind of like how an old chocolate bar can become white and crumbly or soft and chewy. You build your bomb expecting your Pu to be in the δ phase (delta phase). If your pit has changed into a different phase due to age, the explosive lenses will not compress it correctly and you will get a dud or reduced yield.

  https://en.wikipedia.org/wiki/Allotropes_of_plutonium

Can these things be tested in space? Instead of doing a test on earth, push something way out to space and test it, what would the impact be on earth if it was far away?

Probably yes, and probably safely once in space, but the difficulty is that it has to travel from the ground through the atmosphere to get to space, and sometimes rockets fail on the way up, creating a bigger problem when they fall back down.

This is also the solution for nuclear waste: Just send it on a direct course to the sun and the problem will solve itself. But getting it reliably on that route is the (still) problematic thing.

The effect would be a massive EMP damaging infrastructure over hundreds or thousands of miles.

https://en.wikipedia.org/wiki/Starfish_Prime

It's inevitable that if the US is to keep a nuclear arsenal, thousands of new nuclear weapons will have to be manufactured...

... and tested?

The U.S. government could announce to the world a temporary resumption of atmospheric nuclear tests, in violation of several binding treaties, in the name of practical necessity. Such a step would not be taken lightly. A series of tests could be conducted over a few months to test new bomb designs, as well as the older Minutemen and B61s. Those tests could take place at the old Nevada Test Site and in the Pacific.

Why atmospheric? Just cost? Ground test or sending it out to L1 seem reasonable.

In a real context, nuclear weapons will be detonated above ground, and you can't test heat damage and overpressure as a function of distance from the blast, with underground tests.

We already have sufficient models for how these warheads will perform if they work. The question at hand is whether they work - an underground test would give a measure of yield.

What is L1?

A lagrange point in space. On further thought l1 isnt stable. 3 and 4 are closer which isn't great but trash there would stay there.

We probably just don't need as many nukes - the world will be much less populated with near-peer adversaries in 2040.

For example in 20 years time there will be as many Russian military-age males as in 1969, and as many Chinese as in 1973, and the populations will be declining continuously.

sure, but they will have just as much land to distribute their launch sites across. part of why the US has so many missiles is to have a decent chance of hitting all enemy launch sites

Can we not simulate these weapons, instead of actually blowing them up?

The physics seems to be pretty well understood, is this just impossible with current tech? Doesn't the US government have access to some pretty amazing supercomputers?

Reply on news.ycombinator.com