7 ms·

The author seems to think people aren't very familiar with electrical switches. But normal people are frequently only one switch away from likely death. Like if

by tlb 13y ago

The author seems to think people aren't very familiar with electrical switches. But normal people are frequently only one switch away from likely death. Like if you turn off your electric lawnmower and reach inside. It's recommended to have a fail-safe, but switches don't spontaneously turn on at even 1 in a billion and most people don't write breathless articles about the one time they were only one switch away from death 50 years ago.

wnevets 13y ago
> Jones found that of the four safety mechanisms in the Faro bomb, designed to prevent unintended detonation, three failed to operate properly 3/4 of the safety mechanisms on a 4 megaton bomb failed over US soil. I would hardly call this a breathless article.
- stevetursi 13y ago
  > and it was only that final, highly vulnerable switch that averted calamity What does "highly vulnerable" even mean? Could it have been triggered by the wind? If I'm the guy at the control panel, am I accidentally hitting that switch a couple of times per week? Maybe the only way that switch gets triggered is if I black out from the stress and my nose hits it on my way to the floor. I get the significance of three/four barriers failing, but if the fourth one has a one-in-a-billion chance, is it really fair to say North Carolina was "dramatically close" to being nuked?
  - wnevets 13y ago
    > the final switch that prevented disaster could easily have been shorted by an electrical jolt I would say its fair, would you feel safe if all bombs only had this switch from the 50s?
    - stevetursi 13y ago
      So he means something like what a surge protector blocks? I suppose they could have neglected to put a fuse in the circuit which should have protected it against a surge that would damage some very expensive equipment (not to mention cause nuclear detonation.) But I don't know. I wasn't there to see and I didn't read the report. And we all know fuses fail as well - would that be the fourth out of five failsafes? To answer your question, no - I honestly would not feel safe. But "dramatically close" is different altogether. When I think "dramatically close" I think about the time my kid ran into the street and I grabbed him from in front of a car... I needed to act urgently to avoid disaster. I could be wrong though (and often am.)
      - wnevets 13y ago
        According to the wiki it disintegrated on impact, I'm sure that played a role in the "dramatically close" statement
    - smsm42 13y ago
      That's like comparing losing a lottery by guessing only 3 numbers out of 4 and playing a lottery where you need to guess only one number. These are statistically very different events, since you don't guess other 3 numbers correctly every time - and all fail-safes aren't failing at the same time. That's the point of having many of them - if each of them fails one-in-a-million time, you get 10^24 of security.
      - nostromo 13y ago
        Here we have a sample of 8 safety measures across two bombs with 3 failures. That's not 1 in a million, that's 1 in 3.
        corin_ 13y ago
        That's not how probability works, though. I could go and buy a lottery ticket tomorrow for the first time in my life and win the jackpot, it wouldn't mean that the odds of winning are 100%.
        nostromo 13y ago
        No, this isn't a problem of probability, it's a problem of "not enough data." Based on this tiny sample, here's the lottery analogy: The lottery will flip 4 coins that have a 37.5% chance of being heads. If you roll 4 heads lots of people die. So, just using this very inadequate sample from a single declassified document, you could estimate that a misfired bomb has a 2% chance of detonation (37.5%^4). Obviously none of us know the real dataset, but my point is simply that it's not comforting that so many safety measures failed. (Which, as it happens, is also the opinion of the author of the original report in the article.)
        corin_ 13y ago
        > No, this isn't a problem of probability, it's a problem of "not enough data." That was exactly my point, that with such little data you cannot judge the probability.
        13y ago
  - coldtea 13y ago
    >What does "highly vulnerable" even mean? Could it have been triggered by the wind? Maybe the breeze of the wind is less potent than a impact of the drop of a heavy metallic object from an airplane.
  - m4x 13y ago
    From http://en.m.wikipedia.org/wiki/1961_Goldsboro_B-52_crash http://en.m.wikipedia.org/wiki/1961_Goldsboro_B-52_crash : Later, however, it was found that both bombs were fully functional and that a single low-voltage switch was indeed all that stood between detonation. The switch they're referring to is the arm/safe switch on the pilot's control panel. So the bombs didn't fail, one of them armed and tried to detonate exactly as designed when dropped from an aircraft. The only reason it didn't was because the pilot's arming circuit hadn't been activated. This isn't very reassuring because that arming switch may well have simply been a toggle switch under a cover, and even if it was better protected it still almost certainly boiled down to simply closing one or a pair of contacts, which can quite conceivably happen accidentally while bailing out of a spinning aircraft. Not exactly a one-in-a-billion chance. There is a discrepancy between what TFA says (three out of four safety circuits) and what the wiki says (five out of six arming circuits) and I can't find anything to clear that up on my phone. Either way it's a disconcerting piece of history
    - einhverfr 13y ago
      I think once you are dealing with an out of control aircraft that disintegrates in the air, all bets are off.
      - nitrogen 13y ago
        A nuclear bomb has to detonate in the right way in order to produce a nuclear explosion. Otherwise, it should just break apart. From that I assume it should be possible to design a bomb and corresponding delivery system that are highly unlikely to detonate by accident, even if your plane blows up or rips apart in midair.
        einhverfr 13y ago
        It should be possible to design a nuclear power plant to operate safely after a magnitude 9 earthquake and a tsunami following it, but I would still say in that case all bets are off. The problem is that there are certain sorts of disasters where you really can't be sure you have the unknowns down well enough..... That's why aircraft components are rigorously tested. The only really safe solution I can think of is to try to anticipate the problem and sabotage the warhead first. For example one might, say, design warheads to destroy portions of the electronics if they get a signal that the pilot ejected. So yes, it is probably possible. Is it possible to do so reasonably perfectly? I doubt it to be honest.
- cpleppert 13y ago
  If the safety mechanisms are failing catastrophically then how will the bomb even detonate? The timing mechanisms have to execute perfectly to get a full power detonation and should be designed to become inert in a catastrophic breakup anyway. If any part of the switching mechanism becomes damaged the bomb should should/could be designed not to be able to detonate without relying on any electrical switches.
  - TylerE 13y ago
    Because even if "fizzles" and goes off at 1% yield, Goldsboro, NC (which is, incidentally, about 20 miles from where I live) is a crate. Even at 1% yield that is 3x the Hiroshima bomb.
    - jlgreco 13y ago
      A fizzle of the (relatively small, fission) primary would be some percentage of a number much lower than 4 megatons. Most of that "oomph" comes from the fusion secondary which is set off by the primary detonating as designed. I don't know how big the primary on that bomb was, and can't find anything about that, but did find this (http://en.wikipedia.org/wiki/Castle_Koon http://en.wikipedia.org/wiki/Castle_Koon) test of a fusion bomb where the primary didn't fizzle but the secondary nevertheless did not ignite. The primary there was 110 kilotons. So the primary on that bomb, not fizzling, is around 5 or 6 Nagasaki blasts.
      - ballard 13y ago
        How much energy is yielded by "fusion" secondaries ([1st]fission - [2nd]fusion-fission) varies by design, but secondary's fusion is used for more efficient (eg more complete) fission. Only huge bombs have tertiary fusion stages that contribute much more yield than fission stages. http://en.wikipedia.org/wiki/Teller-Ulam_design http://en.wikipedia.org/wiki/Teller-Ulam_design
        jlgreco 13y ago
        I'm not sure what you mean, the W39 was a Teller-Ulam design bomb. It wasn't a boosted fission bomb (though maybe the primary was? I don't know). I believe the Castle bombs were as well.
joezydeco 13y ago
Let's try your lawnmower experiment in freefall at 10,000 feet and see if we can reach the same confidence level.
- Zikes 13y ago
  After surviving the breaking up of the plane that was carrying it.
cma 13y ago
>most people don't write breathless articles about the one time they were only one switch away from death 50 years ago. This article isn't about one person one switch away from death, it's about (potentially) millions of people one switch away from death after three of four switches failed.
- JackFr 13y ago
  The book ain't gonna sell itself -- someone's gottta do the shilling.
salem 13y ago
Yes, people are frequently one switch away from death, the difference is the consequences of that switch. Also, these are pretty much standard these days: http://en.wikipedia.org/wiki/Residual-current_device http://en.wikipedia.org/wiki/Residual-current_device
coldtea 13y ago
>It's recommended to have a fail-safe, but switches don't spontaneously turn on at even 1 in a billion and most people don't write breathless articles about the one time they were only one switch away from death 50 years ago With the difference that law-mowers usually aren't dropped from 10,000 feet to the ground in airplane accidents. And, when I'm not in my "let's cheerlead for engineering / throw-all caution-to-the-wind cavalier attitude", I wouldn't really guarantee that they'd stay turned off if that happened. There's also the little known fact that lawn mowers usually don't even have the potential to kill 3 million people in an instant. So, I'd rather go with the opinion of the expert who prepared the accident report and who was actually "a senior engineer in the Sandia national laboratories responsible for the mechanical safety of nuclear weapons". As for the accuracy of the cheerleading: I've had fickle switches turn on on me. Much more frequently than one in a thousand even. Try some low cost electric products and you'd be surprised (nay, shocked).
cynwoody 13y ago
FWIW, according to Daniel Ellsberg†, the switch in question was the pilot's arm/safe switch. I would think that switch would be more like the lawnmower switch and less like the other three switches, which were presumably operated by sensors and circuitry. E.g., an automatic system senses the bomb is dropping, so it deploys the chute and charges up the firing capacitors. Then another system detects the bomb has reached firing altitude and tries to detonate, but nothing happens because the arm/safe switch is on safe. †http://en.wikipedia.org/wiki/1961_Goldsboro_B-52_crash http://en.wikipedia.org/wiki/1961_Goldsboro_B-52_crash
anigbrowl 13y ago
I would feel less comfortable about the reliability of switches in a plane that was spiraling out of control and presumably heading for a catastrophic impact. The switch might never turn itself on, but it has no way to prevent things palling onto it. as for your example about home appliances, surely you are aware that there are many accidents every year from people carelessly or accidentally activating switches on power tools or the like.
mikeash 13y ago
Electrical switches become somewhat less reliable when subjected to violent physical stresses in an airplane crash like this one was. The bomber broke up in mid-air. It's not too hard to imagine how the switch could have been smashed in such a way as to close its contacts during this event. Also, do people really switch off their electric lawnmowers and root around inside with the thing still plugged in? Sounds like a great way to get an unnecessary amputation to me. When I work with electricity which must not be turned on, I always either cut it off with something that can't possibly be undone by accident e.g. unplugging the device entirely, or I'm at least using a much more robust circuit breaker.
- XorNot 13y ago
  Except what I'm not seeing is a description of the design of this electrical switch. It's certainly not a normal switch - dynamo-technology implies some kind of motorization to close the contacts. Of course if there's a lesson here it's already been learned - this is more a commentary on "things we did in the Cold War".
  - mikeash 13y ago
    That's an interesting point. However, I'm not sure if it matters too much. Even if the switch were absolutely indestructible, the wiring going to it is not, and a short elsewhere could easily cause the switch to "close" even if the physical switch itself is not affected.
auctiontheory 13y ago
But normal people are frequently only one switch away from likely death. How often are hundreds of thousands of people simultaneously one (single) switch away from likely death? Do you think that a switch on a device is more likely to fail/flip when that device has been dropped from a plane (than when it is sitting still in your garage)?
Dylan16807 13y ago
I'd be pretty uncomfortable about that situation, too. There should be a nice place for me to unplug the power supply entirely before I do anything around the blade.
einhverfr 13y ago
I dunno. I try hard to avoid trusting switches with my life, or if I must, I try to make sure I have at least two or more mechanically triggered switches between myself and impending doom. The switches in this case are likely soft switches or relays, if they could be triggered electrically. If the other three failed, there could be all sorts of reasons to be concerned about this. Interestingly I had been told about this incident growing up in the 1980's by pro-nuclear-disarmament folks. But the question in my mind is what would have happened if the bomb had detonated? Would we have jumped to the conclusion that the Soviets were somehow responsible? Could the accident, have triggered nuclear war?

by tlb 13y ago

wnevets 13y ago

> Jones found that of the four safety mechanisms in the Faro bomb, designed to prevent unintended detonation, three failed to operate properly 3/4 of the safety mechanisms on a 4 megaton bomb failed over US soil. I would hardly call this a breathless article.

stevetursi 13y ago
> and it was only that final, highly vulnerable switch that averted calamity What does "highly vulnerable" even mean? Could it have been triggered by the wind? If I'm the guy at the control panel, am I accidentally hitting that switch a couple of times per week? Maybe the only way that switch gets triggered is if I black out from the stress and my nose hits it on my way to the floor. I get the significance of three/four barriers failing, but if the fourth one has a one-in-a-billion chance, is it really fair to say North Carolina was "dramatically close" to being nuked?
- wnevets 13y ago
  > the final switch that prevented disaster could easily have been shorted by an electrical jolt I would say its fair, would you feel safe if all bombs only had this switch from the 50s?
  - stevetursi 13y ago
    So he means something like what a surge protector blocks? I suppose they could have neglected to put a fuse in the circuit which should have protected it against a surge that would damage some very expensive equipment (not to mention cause nuclear detonation.) But I don't know. I wasn't there to see and I didn't read the report. And we all know fuses fail as well - would that be the fourth out of five failsafes? To answer your question, no - I honestly would not feel safe. But "dramatically close" is different altogether. When I think "dramatically close" I think about the time my kid ran into the street and I grabbed him from in front of a car... I needed to act urgently to avoid disaster. I could be wrong though (and often am.)
    - wnevets 13y ago
      According to the wiki it disintegrated on impact, I'm sure that played a role in the "dramatically close" statement
  - smsm42 13y ago
    That's like comparing losing a lottery by guessing only 3 numbers out of 4 and playing a lottery where you need to guess only one number. These are statistically very different events, since you don't guess other 3 numbers correctly every time - and all fail-safes aren't failing at the same time. That's the point of having many of them - if each of them fails one-in-a-million time, you get 10^24 of security.
    - nostromo 13y ago
      Here we have a sample of 8 safety measures across two bombs with 3 failures. That's not 1 in a million, that's 1 in 3.
      - corin_ 13y ago
        That's not how probability works, though. I could go and buy a lottery ticket tomorrow for the first time in my life and win the jackpot, it wouldn't mean that the odds of winning are 100%.
        nostromo 13y ago
        No, this isn't a problem of probability, it's a problem of "not enough data." Based on this tiny sample, here's the lottery analogy: The lottery will flip 4 coins that have a 37.5% chance of being heads. If you roll 4 heads lots of people die. So, just using this very inadequate sample from a single declassified document, you could estimate that a misfired bomb has a 2% chance of detonation (37.5%^4). Obviously none of us know the real dataset, but my point is simply that it's not comforting that so many safety measures failed. (Which, as it happens, is also the opinion of the author of the original report in the article.)
        corin_ 13y ago
        > No, this isn't a problem of probability, it's a problem of "not enough data." That was exactly my point, that with such little data you cannot judge the probability.
      - 13y ago
- coldtea 13y ago
  >What does "highly vulnerable" even mean? Could it have been triggered by the wind? Maybe the breeze of the wind is less potent than a impact of the drop of a heavy metallic object from an airplane.
- m4x 13y ago
  From http://en.m.wikipedia.org/wiki/1961_Goldsboro_B-52_crash http://en.m.wikipedia.org/wiki/1961_Goldsboro_B-52_crash : Later, however, it was found that both bombs were fully functional and that a single low-voltage switch was indeed all that stood between detonation. The switch they're referring to is the arm/safe switch on the pilot's control panel. So the bombs didn't fail, one of them armed and tried to detonate exactly as designed when dropped from an aircraft. The only reason it didn't was because the pilot's arming circuit hadn't been activated. This isn't very reassuring because that arming switch may well have simply been a toggle switch under a cover, and even if it was better protected it still almost certainly boiled down to simply closing one or a pair of contacts, which can quite conceivably happen accidentally while bailing out of a spinning aircraft. Not exactly a one-in-a-billion chance. There is a discrepancy between what TFA says (three out of four safety circuits) and what the wiki says (five out of six arming circuits) and I can't find anything to clear that up on my phone. Either way it's a disconcerting piece of history
  - einhverfr 13y ago
    I think once you are dealing with an out of control aircraft that disintegrates in the air, all bets are off.
    - nitrogen 13y ago
      A nuclear bomb has to detonate in the right way in order to produce a nuclear explosion. Otherwise, it should just break apart. From that I assume it should be possible to design a bomb and corresponding delivery system that are highly unlikely to detonate by accident, even if your plane blows up or rips apart in midair.
      - einhverfr 13y ago
        It should be possible to design a nuclear power plant to operate safely after a magnitude 9 earthquake and a tsunami following it, but I would still say in that case all bets are off. The problem is that there are certain sorts of disasters where you really can't be sure you have the unknowns down well enough..... That's why aircraft components are rigorously tested. The only really safe solution I can think of is to try to anticipate the problem and sabotage the warhead first. For example one might, say, design warheads to destroy portions of the electronics if they get a signal that the pilot ejected. So yes, it is probably possible. Is it possible to do so reasonably perfectly? I doubt it to be honest.
cpleppert 13y ago
If the safety mechanisms are failing catastrophically then how will the bomb even detonate? The timing mechanisms have to execute perfectly to get a full power detonation and should be designed to become inert in a catastrophic breakup anyway. If any part of the switching mechanism becomes damaged the bomb should should/could be designed not to be able to detonate without relying on any electrical switches.
- TylerE 13y ago
  Because even if "fizzles" and goes off at 1% yield, Goldsboro, NC (which is, incidentally, about 20 miles from where I live) is a crate. Even at 1% yield that is 3x the Hiroshima bomb.
  - jlgreco 13y ago
    A fizzle of the (relatively small, fission) primary would be some percentage of a number much lower than 4 megatons. Most of that "oomph" comes from the fusion secondary which is set off by the primary detonating as designed. I don't know how big the primary on that bomb was, and can't find anything about that, but did find this (http://en.wikipedia.org/wiki/Castle_Koon http://en.wikipedia.org/wiki/Castle_Koon) test of a fusion bomb where the primary didn't fizzle but the secondary nevertheless did not ignite. The primary there was 110 kilotons. So the primary on that bomb, not fizzling, is around 5 or 6 Nagasaki blasts.
    - ballard 13y ago
      How much energy is yielded by "fusion" secondaries ([1st]fission - [2nd]fusion-fission) varies by design, but secondary's fusion is used for more efficient (eg more complete) fission. Only huge bombs have tertiary fusion stages that contribute much more yield than fission stages. http://en.wikipedia.org/wiki/Teller-Ulam_design http://en.wikipedia.org/wiki/Teller-Ulam_design
      - jlgreco 13y ago
        I'm not sure what you mean, the W39 was a Teller-Ulam design bomb. It wasn't a boosted fission bomb (though maybe the primary was? I don't know). I believe the Castle bombs were as well.

joezydeco 13y ago

Let's try your lawnmower experiment in freefall at 10,000 feet and see if we can reach the same confidence level.

Zikes 13y ago
After surviving the breaking up of the plane that was carrying it.

cma 13y ago

>most people don't write breathless articles about the one time they were only one switch away from death 50 years ago. This article isn't about one person one switch away from death, it's about (potentially) millions of people one switch away from death after three of four switches failed.

JackFr 13y ago
The book ain't gonna sell itself -- someone's gottta do the shilling.

salem 13y ago

Yes, people are frequently one switch away from death, the difference is the consequences of that switch. Also, these are pretty much standard these days: http://en.wikipedia.org/wiki/Residual-current_device http://en.wikipedia.org/wiki/Residual-current_device

coldtea 13y ago

>It's recommended to have a fail-safe, but switches don't spontaneously turn on at even 1 in a billion and most people don't write breathless articles about the one time they were only one switch away from death 50 years ago With the difference that law-mowers usually aren't dropped from 10,000 feet to the ground in airplane accidents. And, when I'm not in my "let's cheerlead for engineering / throw-all caution-to-the-wind cavalier attitude", I wouldn't really guarantee that they'd stay turned off if that happened. There's also the little known fact that lawn mowers usually don't even have the potential to kill 3 million people in an instant. So, I'd rather go with the opinion of the expert who prepared the accident report and who was actually "a senior engineer in the Sandia national laboratories responsible for the mechanical safety of nuclear weapons". As for the accuracy of the cheerleading: I've had fickle switches turn on on me. Much more frequently than one in a thousand even. Try some low cost electric products and you'd be surprised (nay, shocked).

cynwoody 13y ago

FWIW, according to Daniel Ellsberg†, the switch in question was the pilot's arm/safe switch. I would think that switch would be more like the lawnmower switch and less like the other three switches, which were presumably operated by sensors and circuitry. E.g., an automatic system senses the bomb is dropping, so it deploys the chute and charges up the firing capacitors. Then another system detects the bomb has reached firing altitude and tries to detonate, but nothing happens because the arm/safe switch is on safe. †http://en.wikipedia.org/wiki/1961_Goldsboro_B-52_crash http://en.wikipedia.org/wiki/1961_Goldsboro_B-52_crash

anigbrowl 13y ago

I would feel less comfortable about the reliability of switches in a plane that was spiraling out of control and presumably heading for a catastrophic impact. The switch might never turn itself on, but it has no way to prevent things palling onto it. as for your example about home appliances, surely you are aware that there are many accidents every year from people carelessly or accidentally activating switches on power tools or the like.

mikeash 13y ago

Electrical switches become somewhat less reliable when subjected to violent physical stresses in an airplane crash like this one was. The bomber broke up in mid-air. It's not too hard to imagine how the switch could have been smashed in such a way as to close its contacts during this event. Also, do people really switch off their electric lawnmowers and root around inside with the thing still plugged in? Sounds like a great way to get an unnecessary amputation to me. When I work with electricity which must not be turned on, I always either cut it off with something that can't possibly be undone by accident e.g. unplugging the device entirely, or I'm at least using a much more robust circuit breaker.

XorNot 13y ago
Except what I'm not seeing is a description of the design of this electrical switch. It's certainly not a normal switch - dynamo-technology implies some kind of motorization to close the contacts. Of course if there's a lesson here it's already been learned - this is more a commentary on "things we did in the Cold War".
- mikeash 13y ago
  That's an interesting point. However, I'm not sure if it matters too much. Even if the switch were absolutely indestructible, the wiring going to it is not, and a short elsewhere could easily cause the switch to "close" even if the physical switch itself is not affected.

auctiontheory 13y ago

But normal people are frequently only one switch away from likely death. How often are hundreds of thousands of people simultaneously one (single) switch away from likely death? Do you think that a switch on a device is more likely to fail/flip when that device has been dropped from a plane (than when it is sitting still in your garage)?

Dylan16807 13y ago

I'd be pretty uncomfortable about that situation, too. There should be a nice place for me to unplug the power supply entirely before I do anything around the blade.

einhverfr 13y ago

I dunno. I try hard to avoid trusting switches with my life, or if I must, I try to make sure I have at least two or more mechanically triggered switches between myself and impending doom. The switches in this case are likely soft switches or relays, if they could be triggered electrically. If the other three failed, there could be all sorts of reasons to be concerned about this. Interestingly I had been told about this incident growing up in the 1980's by pro-nuclear-disarmament folks. But the question in my mind is what would have happened if the bomb had detonated? Would we have jumped to the conclusion that the Soviets were somehow responsible? Could the accident, have triggered nuclear war?