4 ms·

I think we'd be having a similar discussion if a door opened unexpectedly during flight.

thaumasiotes 3y ago
I think the message of the rest of the comment thread is that a door has two states, "open" and "closed", and the processes around those states know about both of them and handle them appropriately. Whereas a door plug can't be opened and only has one state, "installed", but the internal processes treated it as if it were a door, generating the documentation that would be appropriate for opening and closing a door (nothing) rather than the documentation that would be appropriate for removing and reinstalling a structural element. Then, when the door plug wasn't reinstalled correctly, there was no system checking whether it had been "closed", because that concept doesn't apply to door plugs. Whereas an actual door, if it had been left open, would still have generated no relevant documentation, but the plane wouldn't have taken off in that state because the problem would have been obvious.
- RF_Savage 3y ago
  Oh the process and procedure for generating that documentation exists, it was just not followed. The process for checking that the door plug bolts are correctly installed was simply skipped, as nominally no work had been done on it due to it not being marked as worked on. So it was not re-inspected.
  - thaumasiotes 3y ago
    Because closing a door doesn't mean working on it, right?
  - amluto 3y ago
    I’m sure the process exists. But I can understand how an employee could mess it up. If some fairly common work on an airplane involves opening a door, I imagine it’s okay to open and close the door without documenting it. (Just like (I imagine) turning on and off some lights or locking and unlocking a door, etc.) But now there’s a special kind of door (that’s really a plug but works a bit like a door and is literally called a “door plug”), and opening it involves removing four bolts but does not involve removing it completely from the fuselage. So, in a culture where there isn’t a checklist that gets followed for everything, they think “I opened the door (plug)” instead of “I partially removed the door (plug)”, and then they access the rivets (and document that!), and then they sloppily “close the door (plug)” and it is indeed “closed”. But they might not be thinking “I reinstalled the door (plug),” and they don’t check the checklist or fill out the form, and no one ever tracked those four bolts, and a near-catastrophe occurs. So one might argue that calling this thing a door (plug) and “opening” it is a mistake. And one might argue that a design that “closes” a bit like a door but isn’t correctly installed when it looks closed is a poor design.
    - RF_Savage 3y ago
      It is not the rivets, but the door plug gasket that got damaged while dealing with the damaged rivets. That gasket got changed, but they did not mark that they had done work on the door, requiring reinspection.
djmips 3y ago
You can't normally open a regular door in flight as the air pressure prevents it. This is a good feature!

I think the message of the rest of the comment thread is that a door has two states, "open" and "closed", and the processes around those states know about both of them and handle them appropriately. Whereas a door plug can't be opened and only has one state, "installed", but the internal processes treated it as if it were a door, generating the documentation that would be appropriate for opening and closing a door (nothing) rather than the documentation that would be appropriate for removing and reinstalling a structural element. Then, when the door plug wasn't reinstalled correctly, there was no system checking whether it had been "closed", because that concept doesn't apply to door plugs. Whereas an actual door, if it had been left open, would still have generated no relevant documentation, but the plane wouldn't have taken off in that state because the problem would have been obvious.

RF_Savage 3y ago
Oh the process and procedure for generating that documentation exists, it was just not followed. The process for checking that the door plug bolts are correctly installed was simply skipped, as nominally no work had been done on it due to it not being marked as worked on. So it was not re-inspected.
- thaumasiotes 3y ago
  Because closing a door doesn't mean working on it, right?
- amluto 3y ago
  I’m sure the process exists. But I can understand how an employee could mess it up. If some fairly common work on an airplane involves opening a door, I imagine it’s okay to open and close the door without documenting it. (Just like (I imagine) turning on and off some lights or locking and unlocking a door, etc.) But now there’s a special kind of door (that’s really a plug but works a bit like a door and is literally called a “door plug”), and opening it involves removing four bolts but does not involve removing it completely from the fuselage. So, in a culture where there isn’t a checklist that gets followed for everything, they think “I opened the door (plug)” instead of “I partially removed the door (plug)”, and then they access the rivets (and document that!), and then they sloppily “close the door (plug)” and it is indeed “closed”. But they might not be thinking “I reinstalled the door (plug),” and they don’t check the checklist or fill out the form, and no one ever tracked those four bolts, and a near-catastrophe occurs. So one might argue that calling this thing a door (plug) and “opening” it is a mistake. And one might argue that a design that “closes” a bit like a door but isn’t correctly installed when it looks closed is a poor design.
  - RF_Savage 3y ago
    It is not the rivets, but the door plug gasket that got damaged while dealing with the damaged rivets. That gasket got changed, but they did not mark that they had done work on the door, requiring reinspection.

djmips 3y ago

You can't normally open a regular door in flight as the air pressure prevents it. This is a good feature!