In 1961, a B-52 carrying two Mark 39 hydrogen bombs went into an uncontrolled spin over North Carolina.
In the cockpit of the rapidly disintegrating bomber was a lanyard attached to the bomb-release mechanism. Intense G-forces tugged hard at it and unleashed the nukes, which, at four megatons, were 250 times more powerful than the weapon that leveled Hiroshima. One of them “failed safe” and plummeted to the ground unarmed. The other weapon’s failsafe mechanisms — the devices designed to prevent an accidental detonation — were subverted one by one, as Eric Schlosser recounts in his new book, Command and Control.
Then there’s the curious case of the Titan II missile at Launch Complex 374-7, a few miles north of Damascus, Arkansas:
Plumb watched the nine-pound socket slip through the narrow gap between the platform and the missile, fall about 70 feet, hit the thrust mount, and then ricochet off the Titan II. It seemed to happen in slow motion. A moment later, fuel sprayed from a hole in the missile like water from a garden hose.
“Oh man,” Plumb thought. “This is not good.”
What Schlosser fails to mention, either because he is aggressively ignorant or because controversy is good for sales (flip a coin) is that there was still no chance of that bomb actually going off. The documents he got access to talk about the possibility that due to the faulty failsafes (definitely a problem to be fixed) the bomb might have exploded on impact.
What that means is that the high explosive in the weapon could have gone off, creating a nasty dirty-bomb explosion at the worst. With nuclear weapons explosion means something very different from initiation, which is the word for the mushroom-cloud producing event.
Lots of people have tried to explain this to Schlosser (and to the idiots editing the wiki page on the incident, who happily ate up his account of the event) but there is a catch 22. If you are not someone who has directly, personally worked with nuclear weapons in the government they plug their fingers in their ears and go “la la la you don’t count because you’re not official” (as if the laws of physics care). If you are, then you’re not authorized to say anything about it.
Upon rereading, I also noted that he uses the ever-popular (and tic-inducing) trick of labeling a bomb as “X times more powerful than the Hiroshima weapon” by taking its yield and dividing by the yield of Little Boy.
The problem is that this is completely misleading. The thermal radiation intensity produced by a fission or fusion weapon is proportional to the square of its yield, and the blast overpressure is proportional to the cube of yield. The first is due to the inverse-square law, and implies that a bomb with 250 times the yield of Little Boy would produce a heat wave 15 times as intense at a given distance. The second is due to the overpressure being a volumetric characteristic and implies that increasing the yield by 250 times would increase the blast damage at a given distance by a factor of around 6.
Under what conditions would the conventional explosives detonate without initiating the nuclear explosion?
The report specifically concluded that physical damage to the bomb might have caused a partial detonation of the high explosives. Even a complete detonation due to external damage to the device would not have had the precisely timed synchronization needed to compress the fissile material of the primary in such a way as to produce initiation.
As far as I’m able to determine, modern journalists are pretty useless with anything relating to science or technology. This aspect is only exaggerated when the “N” word is involved.
Isegoria wrote “Under what conditions would the conventional explosives detonate without initiating the nuclear explosion?”
I think for natural kinds of efficient bomb designs, even designs which didn’t prioritize prevention of accidental nuclear explosion, the answer is “under almost any conditions other than having the electronics trigger synchonized detonations at multiple points.” The designers want a shock wave travelling inward from all directions to compress the fissile core. (Alternative design approaches exist, like the Hiroshima “gun” design, but apparently require a lot more fissile material. Also ISTR the speed of implosion also can help work around a tendency of spontaneous fission of plutonium to trigger fizzles.) A natural way to create such an implosive pattern is by going to a lot of trouble to set off a lot of detonations at just the right microsecond at multiple points around the system. (And the antiproliferation people talk with the bomb makers, and from what I’ve heard, go to a lot of trouble to make it hard to buy technologies related to superfast accurately timed detonation, though I don’t know how effective that given the enormous advances in fast electronics since the 1950s.) Accidental shocks will tend to start detonation at the wrong place(s) and time(s) so that you never get a critical mass at all, indeed so that you never even come close, because the detonation pattern is some ordinary random pattern starting from a single detonation source rather than the many-detonations implosion that the designers went to so much trouble to arrange.
That said, if you *try* you should be able to able implosion bombs where the propagation of explosion from a single detonation is lensed and delayed and reflected around until it’s sufficiently symmetrical. Then any accident which set off a detonation at the appropriate point would tend to set off the bomb as long as no other interlocks or failsafes intervened. But as long as you have any concern for preventing accidental detonation, making the implosion depend on precisely timed multiple detonations around the outside seems natural: it’s an effective failsafe that arises naturally from other design considerations.
FWIW, there is at least one other highly-synchonized event that’s likely to go on, squirting a pulse of neutrons into the fissile material just at the moment it is suitably compressed. But I don’t know how tight the timing requirements are for that, and in any case even if that burst doesn’t happen you might end up with a nuclear explosion which is big in absolute terms even though it’s relatively small compared to what the bomb could do if everything went right.
TL;DR version of my previous comment: See the image with all the cables in http://en.wikipedia.org/wiki/Exploding-bridgewire_detonator and contemplate how likely it is that an accidental detonation will cause the pattern of implosion that the cabled signals are designed to cause.
SA is correct. It is very, VERY unlikely that a thermonuclear bomb would have detonated in the scenario Schlosser describes. Not that an HE detonation wouldn’t have been serious, but you would not have gotten a nuclear fireball. But beating up on SAC/Curtis LeMay is just too much for leftist journalists to pass up. They’ve been doing it for sixty years, and doubtlessly see no reason to stop now.
If a nuclear device cannot survive the impact of being dropped, does that mean that a nuclear weapon cannot use an impact fuze? It must be fuzed for an airburst?
Long story short (I know, a first for me): it depends on the design of the bomb. That bomb was not designed for impact.
In somewhat more detail, a lot of it has to do with the casing. Some tactical-yield weapons are specifically hardened to be able to take an impact or penetrate the ground before going off, but in the case of a strategic thermonuclear weapon there’s no reason for that. At most you’d fuse it to go off around 30-50 m up for a groundburst, to scrub a hard target like a silo or telephone exchange off the ground (ever wonder why those Bell buildings look so substantial?).
I believe Scipio Americanus is correct that one would seldom have a reason to want a nuclear bomb to detonate on impact — instead, some distance above the ground for ordinary targets, or perhaps significantly below the ground for very hard subterranean targets. I would add that plus or minus a meter is usually good enough as far as nuclear bomb detonation location is concerned, and it should be possible to make a jam-resistant electronic proximity fuse which reliably detects an enormous object like the earth one meter away. (Echo-based methods like radar get tend to get easier or harder in proportion to the fourth power of the distance.)
All that said, though, if for some reason impact detonation was somehow desirable, note how very fast all the relevant processes inside the bomb tend to be. From it seems we can expect the explosive effect to propagate at many kilometers per second. Typical bomb impact speed is probably less than 300m/s. (I have never heard of sonic booms from bombs.) So the detonation should easily outrace any collision-based damage as long as the electronics set off the detonation in time. In the detonator article I cited they are fussing about submicrosecond synchronization, so I would guess that the absolute delays are 10 microseconds or less. 1e-5s * 3e2 m/s means that the collision only has time to crumple 3e-3 m of the bomb nose, about an eighth of an inch, before the detonators go off. Even if you don’t go to any special trouble to make your nuclear innards resistant to crumpling of the outer bomb casing, that amount of crumpling is probably naturally harmless to the nuclear innards: it’ll probably be a slight flattening of a shape that matters for aerodynamics, safely many centimeters away from anything that matters for nuclear physics.
The correct fusing mode will depend on the type of target being attacked. A soft “countervalue” target (a city) would receive an airburst to maximize the radius of destruction since the overpressure requirements are low.
A counterforce target would be expected to be hardened and thus planners would want to maximize the amount of energy coupled into the ground. In the early days, that was done with “laydown” mode where the weapon would actually (and relatively gently) contact the ground before detonation, coming down on large parachutes.
The Mark 39 could be fused for airburst or laydown mode.
It is known to have large parachutes to enable laydown mode.
Now with the B61-11 nuclear bunker buster, the weapon is expected to penetrate the earth before detonation. This greatly increases the percentage of energy delivered and is why the US was able to remove the nine-megaton B53 from service and replace it with a weapon which is likely in the 300 kiloton class.
The only platforms able to carry a monster like the B53 were Titan ICBMs and B-52 bombers. The Titans were horribly unsafe and the chance of a B-52 getting through the defenses of a high value target were low.
The B61-11 is much lighter and can be carried by far more platforms (but not the F-22). In particular, the (former) Short Range Attack Missile seemed like a good carrier to me but it is likely going forward, the B-2 will be the only means of delivery.
From wikipedia:
Laydown delivery is a mode of deploying a free-fall nuclear weapon in which the bomb’s fall is slowed by parachute so that it actually lands on the ground before detonating. Laydown delivery requires that the weapon’s case be reinforced so that it can survive the impact, and generally involves a time-delay fuse to trigger detonation. Laydown mode can be used to increase the effect of the weapon’s blast on built-up targets such as submarine pens, or to transmit a shock wave through the ground to attack deeply buried targets. It has the additional advantage of allowing the carrier aircraft to fly very low – in order to avoid enemy defences – and still get away without being destroyed in the detonation. This is particularly important where high-yield nuclear weapons such as the B41 nuclear bomb and B53 nuclear bomb are concerned. An attack of this type produces large amounts of radioactive fallout.
Would you believe, one of the ideas considered for how to defend our missile bases was to lay in fields of telephone-pole like masts that were distributed in such a manner as to have a high likelihood of damaging an incoming groundburst-set warhead?