Or, how long do you have to wait before you can hit the site with another nuke?
Obviously the nuclear explosion itself lasts a very short time indeed, but the after-effects - the fireball, mushroom cloud, etc - last a lot longer. If you want to be sure of destroying a site, you’re going to want a second hit, but you don’t want your first hit to destroy the second. So how long do you have to wait?
I don’t think a mushroom cloud, firestorm, etc. will significantly damage a falling bomb before it gets a chance to explode. You’re probably good to go as soon as the shock front (which travels at the speed of sound) dissipates.
Username / post combo win! 
What about the fireball within the mushroom cloud?
Don’t have a cite, but I have heard two hours post strike. This was based on a target getting hit by an ICBM initially, with follow up bombers doing a Bomb damage assessment and second strike if needed. How long the actual mushroom cloud lasts, since heat rises and it gets to the stratosphere and starts to break up and drift away, no idea.
This opinion piece from the American Physical Society newsletter asserts that detonations of multiple warheads attacking a single target would need to be separated by “3-5 seconds” (to avoid the immediate effects of the first detonation–radiation, heat, and blast–destroying or deflecting the second warhead) but could be no more than “6-8 seconds” apart, as otherwise the dust cloud kicked up by the first warhead could destroy the second warhead–a very narrow window. Otherwise, you would need to wait as long as 20 minutes for the dust cloud to settle enough to allow a second warhead to get through to the target. (Those calculations are for low-altitude detonations of the sort needed to destroy hardened targets like underground missile silos.)
Thanks.
Why would a dust cloud destroy a warhead?
Dust particles are small, but there are a very large number of them, and the missile is moving very fast.
There may also be an issue with nuclear EMP.
Per Wikipedia, the E1 pulse lasts a tiny fraction of a second, the E2 pulse lasts up to a second, and the E3 pulse lasts quite a bit longer.
It’s the first two the 2nd nuke might have a problem with. Good news: they are over quickly. The bad news: you may not want your missile in the thick of the high atmosphere’s magnetic field when it happens. That’s where the E1 induced electrons of the E2 pulse are going crazy.
Let’s see, an incoming ICBM might reach 5k m/s. So you need to be over 2 seconds away on a straight inbound approach. Add several more seconds due to the angle. So probably over 5 seconds, maybe 10.
But in regards to the OP, if the goal of the first hit is to destroy a target mechanically, then there’s not going to be much of an EMP.
It’s not that different from a spacecraft re-entering the atmosphere. (And ICBMs basically ARE spacecraft re-entering the atmosphere…) The first explosion would push millions upon millions of dust and sand particles to extreme speeds. The second weapon would impact these. That would mean a lot of heat and friction. From the bomb’s perspective, it would be almost as bad as hitting a solid wall.
Previous thread.
After that thread I read of a tactic called X-ray pindown. The idea is that hard X-rays emitted by nuclear explosions can neutralize incoming warheads. Thus, detonating warheads 60-120km above enemy missile fields frequently enough, you could neutralize all the warheads launched from that site.
Back-of-napkin calculations… if we accept that ICBMs travel at 6.5 km/s terminal velocity, and we accept the effective pindown range is 60-120km as reported above, that means a warhead can traverse the pindown range in 9-18 seconds.
So it seems like the “safe” interval is 9-18 seconds. To avoid locking yourself out of the target site, you’d fire less frequently. To pin down the enemy launch side, you’d fire more frequently.
I did wonder if it would be possible to design a nuclear warhead reentry vehicle that was sufficiently well-shielded that it could punch through such a debris cloud from an immediately preceding blast. Even if it could be done, I suspect it would be prohibitively expensive from a mass point of view–would you rather have one super-duper, very heavy (but not necessarily particularly high yield, as these things go) RV on your ICBM or SLBM, or would you rather have 8 or 10 or 12 RVs? (I’m also having flashbacks to Late Cold War discussions of the phrase “throw weight”.)
Wasn’t ‘Dense Pack’ suggested as an ICBM basing method for this reason? That clustering them close together meant the incoming warheads would fraticide each other because of the effects of the first impacts.
Ah, here we go: https://en.wikipedia.org/wiki/Dense_Pack
Absolutely, and I was going to point that out until I saw your post.
I assume that the concept was rejected on ‘first principles’, i.e. under NO circumstance put all your (nuclear) eggs in one basket.
They hoped this would be the case, but then it was recognized that a “dense pack” field could easily be neutralized via X-ray pindown.
‘X-ray pindown’ was more of a theoretical objection. But I don’t think there was much doubt that if a good fraction of even a single MIRVed missile’s payload near-simultaneously hit the ‘Dense Pack’ of the US missile silos, the latter would be toast(ed).
Nuclear warfare strategies are all theoretical, and hopefully they’ll stay that way. But I think we’ve shown that the answer to the OP, the maximum bombardment frequency to avoid fratricide is 9-18 seconds on a given missile site.
That was my thought when I first heard of it, it seems at best an extraordinarily large risk to take.
It would probably have been too expensive but the train-based race-track basing idea always seemed like a good one to me.