7 ms·

On at least 5 airplanes? Of one airline? Are loose bolts a minor issue, or is it as insane as it reads?

Syonyk 3y ago
No, it's as insane as it reads. It means people were careless and sloppy during assembly. Think of it like seeing "one cockroach in your kitchen." It's not one cockroach. You just haven't opened up the rest of the walls.
- Scoundreller 3y ago
  Or it means the torque specs were a little too low, or too high and stripped, or the bolt threading was defective/spec’d wrong. Or bad metallurgy. Or the spec didn’t specify which order to tighten the bolts, so which direction you go causes different outcomes. Lots of stuff is assembled consistently and carefully “wrong”, but as specified.
  - flamedoge 3y ago
    idk what aviation standards say but shouldnt loctite or spring washers at least be used
    - jacquesm 3y ago
      Bolts are torqued down to spec and if specified will have a retention mechanism, usually a pin, a foldover or a wire.
    - ddoolin 3y ago
      These bolts were fitted with a wire pin to prevent the nuts from spinning loose. The nuts were not locking nuts on their own.
  - 88913527 3y ago
    Some part of the process -- anywhere from specs to assembly-- had some cost cutting that led to the outcome. The root cause of this problem is something financially motivated.
    - eightysixfour 3y ago
      While I don’t disagree with you that it is possible, it isn’t a very “engineering” approach to declare that is the case without doing a root cause analysis. Stating it as fact is as bad as the MBAs…
    - crazygringo 3y ago
      You don't know that. Cost cutting could have been a factor. Or the root cause might have been something entirely different. It's quite possible to choose to spend more money on a process or method you believe is higher quality, but still discover it has some specific problem that the previous cheaper version didn't.
  - stefan_ 3y ago
    To be fair it seems there is an image of some of the loose bolts: https://nitter.net/ByERussell/status/1744460136855294106 https://nitter.net/ByERussell/status/1744460136855294106 I think this should stick out to an assembly person as not quite right.
    - CobaltFire 3y ago
      Note that, as a comment there states, there are tool marks indicating the bolts were both tightened and loosened at various points. Complicates the issue of who is at fault; where/when were they loosened?
      - Rapzid 3y ago
        Maybe it was opened for maintenance reasons after it was initially installed. Maybe it was torqued twice to get the lock pin through.
    - Rapzid 3y ago
      That lock nut may not have lined up correctly for some reason and was backed off a bit to get the pin in? That could make it torqued under spec. But also, it doesn't look like that bolt clamps anything so it may not matter if it was a little "loose"..
      - Syonyk 3y ago
        Are you looking at the center nut in the photo that's cotter pinned in place, or the two over on the right side that appear to be "loosely threaded on at best, washer rattling along the shaft"? The center one doesn't concern me. The two that aren't even finger tight do.
        Rapzid 3y ago
        I was and missed everyone was talking about those other bolts in this particular case. Those are much more problematic as the remaining tight bolts could fail leading to a cascade. Though, would expect the flange to still be connected to the post in the aftermath images and the upper pins .. how did they get out of the track? I wish we had close up pics of all the bits in the airframe opening.
  - TaylorAlexander 3y ago
    It could also mean their torque wrenches are not being properly calibrated and traced, and have become out of spec. Out of college I worked as a mechanical engineer at a torque testing firm. We made devices with high quality torque sensors which are calibrated against NIST certified weights every 3-6 months, and those sensors are supposed to be used to check every single torque wrench both before and after each work shift to ensure they are accurate. Someone could choose to ignore that process, if they were sloppily cutting costs.
  - thsksbd 3y ago
    So it's not the worker who ducked up, but there are loose bolts either way because the Walsh gang took over.
  - mint2 3y ago
    A recent example of your points: Toyota recently had wheels literally falling off cars due to mistakes like those.
    - jazzyjackson 3y ago
      Just because I'm a pedant and had to check, I can't find anything saying that wheels had actually fallen off, the recall was just that they could :D
- hef19898 3y ago
  Assembly was loppy, management either ignored or, even worse, encouraged it while QA was asleep. Which is really bad...
- tyingq 3y ago
  > It means people were careless and sloppy during assembly. Maybe. Could also be incorrect torque specs, bad parts, unexpected vibration in that area, etc. Edit: Yes, none of these are good either. Just saying there are many possibilities.
  - hmottestad 3y ago
    Those don’t sound very reassuring either. What’s to say that this is the only bolt with incorrect torque spec, or if this part is bad why not others, and if bad parts are not being caught during quality control then what other parts will be next to fail prematurely? And if there are unexpected vibrations in that area, why wasn’t it discovered during test flights? What other vibration issues would they not have accounted for if they weren’t capable of detecting this one?
    - hypothesis 3y ago
      To be fair, I read recently that another issue with MAX 7 was discovered through flight testing. The issue is that now Boeing wants to exempt plane from safety rules…
      - hmottestad 3y ago
        The de-icing button issue?
- dylan604 3y ago
  What's crazy to me is that these things do not roll of the factory line in large numbers. If it was a car plant or some other line where large numbers come off per shift, you could find that possibly one operator for one shift set their torque wrench to the wrong setting causing the finished items for that shift to be suspect. But seeing as not one plane rolls of the floor per shift, this is much more systemic like possibly the documentation was wrong or similar where it is persistently done incorrectly. Or maybe just the one plane for that one day that the worker incorrectly set their torque wrench that day.
  - hef19898 3y ago
    There is definetly one plabe rolling of FAL per shift. Airbus for example is aiming for a delivery rate of 60 single aisle planes per month.
  - johnwalkr 3y ago
    It can be opposite of your interpretation too. If you make 100,000 cars in a year, each step is done at least 100,000 times and it's justified to have a single wrench, with single non-adjustable torque value, and automated datalogging of applied value for a particular step. Or even a robotic process. Maybe even computer vision to check fastener length. If you make 10 or 500 planes in a year, you have to rely on multiple people to use multiple torque wrenches, with multiple attachments, to follow multiple procedures to assemble multiple planes using fasteners of multiple lengths, nearby multiple other processes and people. There is a big emphasis on procedures and traceability but there are still so many potential failure points that can go undiscovered for a long time.
- dendrite9 3y ago
  Well they left a ladder in the tail of a plane during assembly in the not distant past. It's not like just a couple screws or washers were left behind to rattle around. An ex worked on a QA-type project related to production for them. It makes me wonder if some of the issues were more fundamental than that project could ever have addressed. https://www.nytimes.com/2019/06/28/business/boeing-787-dreamliner-investigation.html https://www.nytimes.com/2019/06/28/business/boeing-787-dream...
- baz00 3y ago
  In software this is a reboot or an apology to a client. In aerospace this is dead people.
  - firen777 3y ago
    Server leaking sensitive personal info due to bug can easily lead to doxxing, which can be amplified to swatting, which can kill. Hospital system failing to parse patient's allergies can kill. MCAS Therac-25 That said, the stuffs I normally work in (or rather what I allow myself to work in) fall in to the reboot and apologize category.
- t-writescode 3y ago
  When a junior software engineer drops a table in production, we don’t say the junior engineer was sloppy. We say the protections and process in place were bad. I don’t like how the sentence in this parent comment blames individuals for process problems.
  - emeril 3y ago
    well, so long as a student's name isn't "Robert'); drop table students;"...
  - thsksbd 3y ago
    It doesn't matter if the worker is sloppy and should have been fired, or if Boeings training is poor and should be improved. Boeing has a responsibility to make sure its workers do their jobs as needed.
- itronitron 3y ago
  From Alaska Airlines... >> "Initial reports from our technicians indicate some loose hardware was visible on some aircraft." Friendly reminder that 'some' can also mean most or all.
  - mips_r4300i 3y ago
    In the same way Apple's "a small number of affected devices" means anywhere from 1 to 99% of devices in the field
baq 3y ago
It's a cliche thing to say, but if it was a known issue and management papered over it - as the alternative is incompetent engineering instead of just cost-center-managed engineering - someone should go jail?
- j-bos 3y ago
  Did anyone go to jail for the previous Boeing MAX crashes?
  - 0xffff2 3y ago
    At this point there's a pretty clear implication that if some Boeing people had gone to jail that time maybe this wouldn't have happened at all.
  - jerkstate 3y ago
    not only did nobody go to jail, they didn't even fix the problem: https://www.travelweekly.com/Travel-News/Airline-News/Omnibus-bill-Boeing-737-Max-certification https://www.travelweekly.com/Travel-News/Airline-News/Omnibu...
- fulladder 3y ago
  Nah. When you have an engineering culture problem, putting people in jail is not going to solve anything. It will scare employees, sure, but it won't fix your culture. I believe someone at Boeing actually was criminally prosecuted in connection with the earlier MAX 8 problems, but, as in all these situations, it is seriously unlikely that that fixed any actual problems. Most likely, somebody felt they needed a neck to wring and found a convenient scapegoat.

No, it's as insane as it reads. It means people were careless and sloppy during assembly. Think of it like seeing "one cockroach in your kitchen." It's not one cockroach. You just haven't opened up the rest of the walls.

Scoundreller 3y ago
Or it means the torque specs were a little too low, or too high and stripped, or the bolt threading was defective/spec’d wrong. Or bad metallurgy. Or the spec didn’t specify which order to tighten the bolts, so which direction you go causes different outcomes. Lots of stuff is assembled consistently and carefully “wrong”, but as specified.
- flamedoge 3y ago
  idk what aviation standards say but shouldnt loctite or spring washers at least be used
  - jacquesm 3y ago
    Bolts are torqued down to spec and if specified will have a retention mechanism, usually a pin, a foldover or a wire.
  - ddoolin 3y ago
    These bolts were fitted with a wire pin to prevent the nuts from spinning loose. The nuts were not locking nuts on their own.
- 88913527 3y ago
  Some part of the process -- anywhere from specs to assembly-- had some cost cutting that led to the outcome. The root cause of this problem is something financially motivated.
  - eightysixfour 3y ago
    While I don’t disagree with you that it is possible, it isn’t a very “engineering” approach to declare that is the case without doing a root cause analysis. Stating it as fact is as bad as the MBAs…
  - crazygringo 3y ago
    You don't know that. Cost cutting could have been a factor. Or the root cause might have been something entirely different. It's quite possible to choose to spend more money on a process or method you believe is higher quality, but still discover it has some specific problem that the previous cheaper version didn't.
- stefan_ 3y ago
  To be fair it seems there is an image of some of the loose bolts: https://nitter.net/ByERussell/status/1744460136855294106 https://nitter.net/ByERussell/status/1744460136855294106 I think this should stick out to an assembly person as not quite right.
  - CobaltFire 3y ago
    Note that, as a comment there states, there are tool marks indicating the bolts were both tightened and loosened at various points. Complicates the issue of who is at fault; where/when were they loosened?
    - Rapzid 3y ago
      Maybe it was opened for maintenance reasons after it was initially installed. Maybe it was torqued twice to get the lock pin through.
  - Rapzid 3y ago
    That lock nut may not have lined up correctly for some reason and was backed off a bit to get the pin in? That could make it torqued under spec. But also, it doesn't look like that bolt clamps anything so it may not matter if it was a little "loose"..
    - Syonyk 3y ago
      Are you looking at the center nut in the photo that's cotter pinned in place, or the two over on the right side that appear to be "loosely threaded on at best, washer rattling along the shaft"? The center one doesn't concern me. The two that aren't even finger tight do.
      - Rapzid 3y ago
        I was and missed everyone was talking about those other bolts in this particular case. Those are much more problematic as the remaining tight bolts could fail leading to a cascade. Though, would expect the flange to still be connected to the post in the aftermath images and the upper pins .. how did they get out of the track? I wish we had close up pics of all the bits in the airframe opening.
- TaylorAlexander 3y ago
  It could also mean their torque wrenches are not being properly calibrated and traced, and have become out of spec. Out of college I worked as a mechanical engineer at a torque testing firm. We made devices with high quality torque sensors which are calibrated against NIST certified weights every 3-6 months, and those sensors are supposed to be used to check every single torque wrench both before and after each work shift to ensure they are accurate. Someone could choose to ignore that process, if they were sloppily cutting costs.
- thsksbd 3y ago
  So it's not the worker who ducked up, but there are loose bolts either way because the Walsh gang took over.
- mint2 3y ago
  A recent example of your points: Toyota recently had wheels literally falling off cars due to mistakes like those.
  - jazzyjackson 3y ago
    Just because I'm a pedant and had to check, I can't find anything saying that wheels had actually fallen off, the recall was just that they could :D
hef19898 3y ago
Assembly was loppy, management either ignored or, even worse, encouraged it while QA was asleep. Which is really bad...
tyingq 3y ago
> It means people were careless and sloppy during assembly. Maybe. Could also be incorrect torque specs, bad parts, unexpected vibration in that area, etc. Edit: Yes, none of these are good either. Just saying there are many possibilities.
- hmottestad 3y ago
  Those don’t sound very reassuring either. What’s to say that this is the only bolt with incorrect torque spec, or if this part is bad why not others, and if bad parts are not being caught during quality control then what other parts will be next to fail prematurely? And if there are unexpected vibrations in that area, why wasn’t it discovered during test flights? What other vibration issues would they not have accounted for if they weren’t capable of detecting this one?
  - hypothesis 3y ago
    To be fair, I read recently that another issue with MAX 7 was discovered through flight testing. The issue is that now Boeing wants to exempt plane from safety rules…
    - hmottestad 3y ago
      The de-icing button issue?
dylan604 3y ago
What's crazy to me is that these things do not roll of the factory line in large numbers. If it was a car plant or some other line where large numbers come off per shift, you could find that possibly one operator for one shift set their torque wrench to the wrong setting causing the finished items for that shift to be suspect. But seeing as not one plane rolls of the floor per shift, this is much more systemic like possibly the documentation was wrong or similar where it is persistently done incorrectly. Or maybe just the one plane for that one day that the worker incorrectly set their torque wrench that day.
- hef19898 3y ago
  There is definetly one plabe rolling of FAL per shift. Airbus for example is aiming for a delivery rate of 60 single aisle planes per month.
- johnwalkr 3y ago
  It can be opposite of your interpretation too. If you make 100,000 cars in a year, each step is done at least 100,000 times and it's justified to have a single wrench, with single non-adjustable torque value, and automated datalogging of applied value for a particular step. Or even a robotic process. Maybe even computer vision to check fastener length. If you make 10 or 500 planes in a year, you have to rely on multiple people to use multiple torque wrenches, with multiple attachments, to follow multiple procedures to assemble multiple planes using fasteners of multiple lengths, nearby multiple other processes and people. There is a big emphasis on procedures and traceability but there are still so many potential failure points that can go undiscovered for a long time.
dendrite9 3y ago
Well they left a ladder in the tail of a plane during assembly in the not distant past. It's not like just a couple screws or washers were left behind to rattle around. An ex worked on a QA-type project related to production for them. It makes me wonder if some of the issues were more fundamental than that project could ever have addressed. https://www.nytimes.com/2019/06/28/business/boeing-787-dreamliner-investigation.html https://www.nytimes.com/2019/06/28/business/boeing-787-dream...
baz00 3y ago
In software this is a reboot or an apology to a client. In aerospace this is dead people.
- firen777 3y ago
  Server leaking sensitive personal info due to bug can easily lead to doxxing, which can be amplified to swatting, which can kill. Hospital system failing to parse patient's allergies can kill. MCAS Therac-25 That said, the stuffs I normally work in (or rather what I allow myself to work in) fall in to the reboot and apologize category.
t-writescode 3y ago
When a junior software engineer drops a table in production, we don’t say the junior engineer was sloppy. We say the protections and process in place were bad. I don’t like how the sentence in this parent comment blames individuals for process problems.
- emeril 3y ago
  well, so long as a student's name isn't "Robert'); drop table students;"...
- thsksbd 3y ago
  It doesn't matter if the worker is sloppy and should have been fired, or if Boeings training is poor and should be improved. Boeing has a responsibility to make sure its workers do their jobs as needed.
itronitron 3y ago
From Alaska Airlines... >> "Initial reports from our technicians indicate some loose hardware was visible on some aircraft." Friendly reminder that 'some' can also mean most or all.
- mips_r4300i 3y ago
  In the same way Apple's "a small number of affected devices" means anywhere from 1 to 99% of devices in the field

baq 3y ago

It's a cliche thing to say, but if it was a known issue and management papered over it - as the alternative is incompetent engineering instead of just cost-center-managed engineering - someone should go jail?

j-bos 3y ago
Did anyone go to jail for the previous Boeing MAX crashes?
- 0xffff2 3y ago
  At this point there's a pretty clear implication that if some Boeing people had gone to jail that time maybe this wouldn't have happened at all.
- jerkstate 3y ago
  not only did nobody go to jail, they didn't even fix the problem: https://www.travelweekly.com/Travel-News/Airline-News/Omnibus-bill-Boeing-737-Max-certification https://www.travelweekly.com/Travel-News/Airline-News/Omnibu...
fulladder 3y ago
Nah. When you have an engineering culture problem, putting people in jail is not going to solve anything. It will scare employees, sure, but it won't fix your culture. I believe someone at Boeing actually was criminally prosecuted in connection with the earlier MAX 8 problems, but, as in all these situations, it is seriously unlikely that that fixed any actual problems. Most likely, somebody felt they needed a neck to wring and found a convenient scapegoat.