The History of Nuclear Secrecy in the U.S. | Hacker News Reader

The History of Nuclear Secrecy in the U.S.

inference-review.com

56 pointstorritest23 comments

Open articleView on HN

The book author is also the author of NukeMap.

https://nuclearsecrecy.com/nukemap/

Who needs Resident Evil when you have a use of Google maps like that. Terrifying.

> I was told that the mushroom clouds were clearly visible from the casinos.

Just warming up for Armageddon in the distance, NBD.

But on another note, Bernstein says that the barrier to armament is only political, not technical. How does a nation-state that has previously not had a thermonuclear program go about designing and testing a weapon in the age of test-ban treaties? (Ignoring countries that obviously flaunt it,N̶K̶ ̶&̶ ̶I̶r̶a̶n̶). Say Germany wanted to build up a weapons program - wouldn't testing be an absolute prerequisite?

Edit: ok, so hypothetically for countries that flaunt it.

Probably the same way that nations that previously tested nuclear weapons currently develop nuclear weapons: with supercomputers and ICF facilities.

Making a bomb isn't very tricky. The nuclear bomb tests were a combination of learning physics and threatening adversaries. The understanding of making nuclear weapons doesn't see significant benefit from blowing up more bombs.

The physics is freely available. The gist of the designs are public. The rest is plug n chug.

Making a gun-type fission bomb, like the Hiroshima bomb, takes a lot of highly enriched uranium, but it isn't complicated. Making a plutonium implosion bomb is hard, for a number of reasons. Making a good implosion is moderately hard. It takes two kinds of explosives and lots of tricks to make it work. Plus techniques for making X-ray movies of an explosion as it happens to debug the process. Los Alamos went through tons of explosives getting that right. "The Curve of Binding Energy", by John McPhee, covers all this.

Making a fusion bomb is quite hard, and there are still parts of that which are classified. US bombs use some material called "FOGBANK", which is not only classified but so hard to make that there was a period of over a decade during which the US couldn't make it.

There's some grumbling in the nuclear establishment that there were too many smart people working on bomb designs in the 1950s and 1960s, and as a result, the designs are too complicated. This is a problem for building new bombs without testing them.

On the enrichment side, the centrifuges keep getting better. The original gaseous diffusion plant at Oak Ridge was a mile long. Everybody who built nuclear weapons in the 1950s had to build up a huge amount of plant. Today, an enrichment facility is about the size of a WalMart SuperCenter or a medium to large data center.[1]

Laser enrichment requires even less plant, but it's hard to make it work. The technology is mostly classified, but the NRC has a self-training course on it.[2] The current leading company in that area is SILEX, in Australia. Their investor presentation indicates that silicon isotope separation, which has some use in quantum computing, has more business potential than uranium separation.

[1] https://earth.google.com/web/search/Eunice,+New+Mexico/@32.4...

[2] https://www.nrc.gov/docs/ML1204/ML12045A051.pdf

I think there is a large discrepancy between hypothetically viable, and demonstrably viable.

If you only care about fission weapons, a practically viable weapon is easy, and does not require testing.

This was demonstrated by South Africa, who developed a fission weapon, using an indigenous and entirely novel isotope separation technique and no outside help. They produced a stockpile of around five weapons, with absolutely no testing. The Vela incident often referred to is outwith the relevant timescale, being in 1979 and the weapons program ran from 1984 to 1994. The devices themselves were gun-type bombs using uranium, similar to the 'little boy' (Hiroshima) design, with improvements in safety primarily.

This was as a hedge against Soviet agression in the Angolan civil war, and they were intending to detonate one of the bombs as proof of their capability if the situation ever became desperate. As a pariah state due to Apartheid, the idea was that the US and allies would be forced to abandon their arms embargos and step in to help with conventional assistance in order to prevent World War III. This never happened, and South Africa is still the only state ever to have developed a viable nuclear weapons capability and then voluntarily given it up.

The details are available in an excellent book by David Albright of ISIS (not that one, the Institute foir Science and International Security...) at https://isis-online.org/uploads/isis-reports/documents/Revis...

The designs are demonstrably viable. Hundreds of nukes were detonated unnecessarily to prove this point. Nations don't need to make new designs. Gun and super designs are common knowledge.

If you want to be on the UN Security council you need to detonate a nuke somewhere in a desert. It's a requirement.

> If you want to be on the UN Security council you need to detonate a nuke somewhere in a desert.

No, you don’t.

Every UN member rotates through UNSC seats. If you want a permanent Security Council seat, you need to be or be generally recognized successor in interest to one of five powers that got those seats after WW2, when only one had detonated nuclear weapons (and the important one were on Japanese cities, not desert.)

Much of the testing is due to refining designs to make them more efficient, smaller, and lighter. If the Ukraine decided to build a bomb that wasn't state of the art they could be reasonably assured that it would work, but substantially less clear on final yield.

Yield only really matters in theory and for planning MAD. Tank divisions are equally destroyed with a 100 kiloton device as a 10 megaton device.

But as a deterrent, you would need this to be reasonably accurate, yes?

Why? All you really care is whether the boom is big enough, and that you can deliver it on target. The latter was a big reason for optimising bomb construction, but you can just as well optimise your launchers - which leads to missiles like R-36, where the main difference between it and a space launcher is replacing the payload. Yes, early USA ICBMs also serve as base for space launches, but unlike R-36 were much less storable (except for Titan, iirc) and less powerful.

If we're talking about devices with megaton or hundred-kiloton yields, accuracy isn't super important. And simpler, unminiaturized weapons are probably going to be bomber-dropped anyway.

> Ignoring countries that obviously flaunt it,N̶K̶ ̶&̶ ̶I̶r̶a̶n̶)

Don't forget Israel

> How does a nation-state that has previously not had a thermonuclear program go about designing and testing a weapon in the age of test-ban treaties?

Openly defy or repudiate the treaty, if you are a party in the first place.

> Say Germany wanted to build up a weapons program - wouldn't testing be an absolute prerequisite?

Arguably not, but the main reason to have nukes is to be able to influence other state’s behavior with the implicit or explicit threat of their use, which is a lot more credible with clear demonstration of their functionality (and, also, with clear demonstration of the will to violate international norms.)

>Ignoring countries that obviously flaunt it, NK & Iran

When has Iran flaunted it? Three countries have made tests ignoring the treaty. North Korea, India, and Pakistan, all non-signatory states.

Also one of: France, Israel, and or South Africa [0], of whom the latter two were signatories to the 1963 agreement.

[0] https://en.m.wikipedia.org/wiki/Vela_incident

I should have specified the 96 treaty, as both the USSR and USA have violated the 1963 agreement as well. Still no violations by Iran though.

Iran doesn't have the nuclear bomb, as far as we know (?)

Iran has never really had much of an active nuclear weapons program, despite the media coverage. That's why I object to claims they violated testing treaties.

I would love to read this (presumably excellent) article, but this horrible javascript monstrosity of a website has inexplicably disabled my ability to navigate the page with my j key.

To be more clear, it seems they've disabled the native browser scrolling via javascript.

I have wrist strain issues so I have a programmable keyboard. I don't use a mouse to scroll and click, I use a layer on my keyboard. I don't map the arrow keys or page up/page down because they generally aren't useful. I literally have to change my setup to scroll on the page. I imagine there are others in need of even more specialized equipment than mine that is broken by this implementation.

I have yet to find a good reason when developing a webpage to reimplement scroll. Certainly not for a fancy progress bar which is presumably the reason this was done. This is a legitimate accessibility concern.

Reply on news.ycombinator.com