Adding to this, I've heard on the grapevine that the parameters of the detonation sequence are generated from a hash of the arming code number. The arming code number isn't like a traditional code lock where there's a chip that's doing "strcmp(entered_code, CORRECT_CODE)", but instead the arming code number is a fundamental piece of how the bomb works. Without that you need the kind of state nuclear apparatus you'd need to build a bomb in the first place to reverse engineer (think a state run nuclear research lab), or at best you can remanufacture it into a much crappier bomb that probably duds to just being a dirty bomb.
Dunno, wikipedia says the parameters are encrypted.
Anyway it would probably be good enough to just rip all the safety/firing electronics out, and replace it with your own best effort attempt at controlling the fireset capacitors charging and firing.
What's there to vary anyway? Charge voltages, and discharge timing? Odds would be you could probably make it go boom in a more controlled manner than what happened in past accidents when the bombs exploded after hitting the ground.
The trouble would be getting past the tamper-proof barrier, because the modern bombs are probably made to self-destruct, from what I've read, so instead of a nice ready-made bomb internals waiting for a simpler arduino control circuit for a fireset, you'd be dealing with a hot mess, after trying to get in.
> Dunno, wikipedia says the parameters are encrypted.
Laymen refer to any use of crypto as "encrypted".
> Anyway it would probably be good enough to just rip all the safety/firing electronics out, and replace it with your own best effort attempt at controlling the fireset capacitors charging and firing.
It's really not though, it's very very specific to the design of the warhead, and the electronics are embedded in it in a way that you can't just take it apart and put it back together. It's all nasty stuff like FOGBANK that if you could remanufacture, you wouldn't need to steal a nuke in the first place.
> What's there to vary anyway? Charge voltages, and discharge timing? Odds would be you could probably make it go boom in a more controlled manner than what happened in past accidents when the bombs exploded after hitting the ground.
Yep, nanosecond precision detonation on the shaped charges, and not just "all at the same time", but a very specific sequence that's dependent on the exact geometries, isotopes and these days, how they've decayed over decades. That's what Oak Ridge uses those fancy super computers for.
> The trouble would be getting past the tamper-proof barrier, because the modern bombs are probably made to self-destruct, from what I've read, so instead of a nice ready-made bomb internals waiting for a simpler arduino control circuit for a fireset, you'd be dealing with a hot mess, after trying to get in.
They are very much not designed to self destruct. Really the opposite. And yes, the threat model is that without the codes, you leave someone with components that they essentially have to remanufacture into a new bomb, which takes a well funded state actor. Or they just use the bits as a dirty bomb, which is why that gets so much play in the age of terrorist cells. But you don't get a full on usable nuke by just sticking a microcontroller on a stolen bomb.
Old bombs have much simpler designs, which can be hacked. Moreover, A bomb have core and blanket. Blanket can be reused. The core can be replaced with a few kilos of plutonium and deflectors. Plutonium is easy to extract from waste of weapon grade reactor. Yeah, it's easier to remanufacture.
Yeah, I was astonished that they managed to get such a successful experiment on the first try with trinity, despite 40s technology and computers. I expected that there'd be a few failures before the first successful nuclear explosion.
