Alaska Airlines 737-900 MAX loses a door in-flight out of PDX
Join Date: Jun 2009
Location: Montreal
Age: 65
Posts: 41
Likes: 0
Received 0 Likes
on
0 Posts
What it does tell us is that the door is intact and didn't collapse. So, the failure was *not* collapse of the door which is a huge relief. Unless the other side shows the stop fittings failed, which would be shocking, it's a functional failure not a structural failure.
Join Date: Jul 2014
Location: UK
Posts: 120
Likes: 0
Received 0 Likes
on
0 Posts
PureVidaTransport made an interesting point about the low probability 26A and 26B would be empty. (Apologies if others have commented on this but I'm not finding it by search.)
Choosing to ignore other factors, there is randomly > 95% probability that 26A would NOT be empty, and > 99% probability that 26A and 26B would BOTH not be empty. In a root cause investigation, probabilities above 95% are usually considered significant.
Putting this together with the 3 prior auto-pressurization events I expect they will look closely to see if the seats were de-prioritized/blocked (at least 26A) due to intermittent noise, and my speculation is they are, and they are related.
It is possible the cause of the auto-pressurization warning was transient during certain flight phases, therefore when switched to ALT things looked fine. Uncomfortable noise may have also occurred transiently, again depending on flight phase. There is more insulation covering the plug than there would be covering a normal exposed door, but apparently air leaks are very loud.
Main point is it is extremely unlikely the empty seats were randomly empty.
(If you are going to reply that 2 people reserved the seats and missed the connection, please include a source.)
Choosing to ignore other factors, there is randomly > 95% probability that 26A would NOT be empty, and > 99% probability that 26A and 26B would BOTH not be empty. In a root cause investigation, probabilities above 95% are usually considered significant.
Putting this together with the 3 prior auto-pressurization events I expect they will look closely to see if the seats were de-prioritized/blocked (at least 26A) due to intermittent noise, and my speculation is they are, and they are related.
It is possible the cause of the auto-pressurization warning was transient during certain flight phases, therefore when switched to ALT things looked fine. Uncomfortable noise may have also occurred transiently, again depending on flight phase. There is more insulation covering the plug than there would be covering a normal exposed door, but apparently air leaks are very loud.
Main point is it is extremely unlikely the empty seats were randomly empty.
(If you are going to reply that 2 people reserved the seats and missed the connection, please include a source.)
Being loose should not be a problem as the bolts aren't clamping anything. They are functioning as pins to stop motion. As long as they are in place, no motion takes place, loose or not. Until the last bolt leaves it's hole the door should not move and there would not be evidence of any problem.
Having been next to a door with a bad seal, the engine sound is piercing even though the gap was only an eighth of an inch. It was a genuine 737 plug style door and, when the pressurization came up, the door moved to seal the gap. That was over wing and this is behind the engine, so even greater noise, though the interior panel might dampen some of it.
That photo is as close to a burlesque fan dance as an aircraft photo can be. Show it. Show it!
Having been next to a door with a bad seal, the engine sound is piercing even though the gap was only an eighth of an inch. It was a genuine 737 plug style door and, when the pressurization came up, the door moved to seal the gap. That was over wing and this is behind the engine, so even greater noise, though the interior panel might dampen some of it.
That photo is as close to a burlesque fan dance as an aircraft photo can be. Show it. Show it!
Join Date: Apr 2010
Location: Alexandria VA USA
Posts: 40
Likes: 0
Received 0 Likes
on
0 Posts
My guess is that none of the bolts were installed sometime during or afree interior fit. If any of the bolts were installed, the door would not have been able to translate upward by the necessary 4cm and disengage the lugs that prevent it from opening. There are two bolts at the top, as discussed, that prevent the door’s guide channels from riding up on the upper guide rollers. A single bolt on either side would have prevented that. But there are two bolts at the bottom of the door that prevent the hinge guide fitting from riding up the hinges. These two bolts pass through both the hinges and the hinge guides, locking the two together. So even if the hinge springs were misrigged, they would be unable to push the door upward if either of the lower bolts were in place. Any single one of these four bolts would prevent the door translating upward. On the ground, the sheer force on the bolts would be zero as the weight of the door is greater than the spring assist pressure. So the door could be closed and appear closed without the bolts installed. Even if the bolts were not fully tightened, so long as the castellated nut and locking pin were installed, the bolt would do its job and not fall out. I can’t imagine the forces on these bolts could overcome even a single, properly installed bolt.
In flight, with no bolts, I think there is a small window where differential pressure acting on the door is enough to force it to translate up, but not enough differential pressure to lock the door to the lugs. At higher differential pressure, there is too much pressure on the lugs to prevent it from translating up. So maybe with a bit of aerodynamic force or turbulence, and no bolts in the door, there is a diff pressure number at which it will open with just a small external force.
so my guess is someone left all the bolts out.
In flight, with no bolts, I think there is a small window where differential pressure acting on the door is enough to force it to translate up, but not enough differential pressure to lock the door to the lugs. At higher differential pressure, there is too much pressure on the lugs to prevent it from translating up. So maybe with a bit of aerodynamic force or turbulence, and no bolts in the door, there is a diff pressure number at which it will open with just a small external force.
so my guess is someone left all the bolts out.
Join Date: Sep 2007
Location: UK
Posts: 23
Likes: 0
Received 0 Likes
on
0 Posts
From BBC
Bolts in need of "additional tightening" have been found during inspections of Boeing 737 Max 9s, United Airlines has said.
Inspections began after a section of the fuselage fell from an Alaska Airlines plane of the same type on Friday.
United Airlines said "installation issues" relating to door plugs would be "remedied" before the aircraft type would return to service.
Some 171 planes of the same type remain grounded by the US regulator
Bolts in need of "additional tightening" have been found during inspections of Boeing 737 Max 9s, United Airlines has said.
Inspections began after a section of the fuselage fell from an Alaska Airlines plane of the same type on Friday.
United Airlines said "installation issues" relating to door plugs would be "remedied" before the aircraft type would return to service.
Some 171 planes of the same type remain grounded by the US regulator
Join Date: Nov 2010
Location: Canada
Posts: 133
Likes: 0
Received 0 Likes
on
0 Posts
Such a tease! I know Portland is known for some titillating dance venues but come on....show us the other side!
Also, looks to be in good shape - at least Alaska wont have to wait for a back ordered plug door to get back in the air. May need order a few pieces of hardware however.....
Also, looks to be in good shape - at least Alaska wont have to wait for a back ordered plug door to get back in the air. May need order a few pieces of hardware however.....
Join Date: Jun 2020
Location: UK
Posts: 17
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Jul 2009
Location: France
Age: 62
Posts: 98
Likes: 0
Received 0 Likes
on
0 Posts
My guess goes along this lines: most likely a procedural quality issue, similar to what resulted in the hasty, not well thought through introduction of the infamous MCAS. Spirit mounts all doors provisionally for rail transport of the fuselage to keep rain and grit out. Boeing unmounts, re-adjusts and finally inspects all doors. Except the fake door replacement plug of the -9. Nobody feels in charge of this one, since it is not a real door, just plugged fuselage...
Last edited by spornrad; 8th Jan 2024 at 21:11.
Join Date: Nov 2015
Location: UK
Posts: 37
Likes: 0
Received 0 Likes
on
0 Posts
Being loose should not be a problem as the bolts aren't clamping anything. They are functioning as pins to stop motion. As long as they are in place, no motion takes place, loose or not. Until the last bolt leaves it's hole the door should not move and there would not be evidence of any problem.
Having been next to a door with a bad seal, the engine sound is piercing even though the gap was only an eighth of an inch. It was a genuine 737 plug style door and, when the pressurization came up, the door moved to seal the gap. That was over wing and this is behind the engine, so even greater noise, though the interior panel might dampen some of it.
That photo is as close to a burlesque fan dance as an aircraft photo can be. Show it. Show it!
Having been next to a door with a bad seal, the engine sound is piercing even though the gap was only an eighth of an inch. It was a genuine 737 plug style door and, when the pressurization came up, the door moved to seal the gap. That was over wing and this is behind the engine, so even greater noise, though the interior panel might dampen some of it.
That photo is as close to a burlesque fan dance as an aircraft photo can be. Show it. Show it!
https://pbs.twimg.com/media/GDWRH6OX...g&name=900x900
Last edited by Thruster763; 8th Jan 2024 at 21:10. Reason: addede link
PureVidaTransport made an interesting point about the low probability 26A and 26B would be empty. (Apologies if others have commented on this but I'm not finding it by search.)
Choosing to ignore other factors, there is randomly > 95% probability that 26A would NOT be empty, and > 99% probability that 26A and 26B would BOTH not be empty. In a root cause investigation, probabilities above 95% are usually considered significant.
Choosing to ignore other factors, there is randomly > 95% probability that 26A would NOT be empty, and > 99% probability that 26A and 26B would BOTH not be empty. In a root cause investigation, probabilities above 95% are usually considered significant.
I was close on a previous prediction; from the comments in the door recovery video, apparently snipped from the Oreganian article:
Even got a near hit on the level of surprise on the part of the finder; someone up to something with that headrest. And I nailed it -the door was up in a tree!
Diane Flaherty said she found a charcoal gray headrest on the ground of her back patio Sunday morning about 8:30 a.m. She thought at first: “Gosh it’s just like an Alaska headrest,” she said, having flown Alaska Airlines to Las Vegas two weeks ago. She dismissed the idea as probably trash that someone had thrown in her yard until she got a message from a friend that NTSB workers were in her neighborhood. She lives next door to the house where the door plug was discovered.
Last edited by MechEngr; 8th Jan 2024 at 21:46.
Join Date: May 2011
Location: U.S.
Posts: 84
Likes: 0
Received 0 Likes
on
0 Posts
I suggested earlier in the thread that the absence of passengers might be an indication of 'issues' that made the seats unbookable. There was a response suggesting that US pax prefer aisle seats rather than window, so not so surprising. But if, as now appears, there may have been loose bolts, I wonder if there had been complaints?
The least preferred seats are middle seats towards the rear, so one might think the empty 26B is not exactly random, a probability of being filled less than 96%. However, there are 14 middle seats behind 26B on that flight and we know at least 9 of them were occupied, with those middle seats further to the rear selected by travelers or gate agents instead of the more forward 26B middle seat.
And of course, window and isle seats are much more in demand, so the empty 26A window seat probability is biased in the other direction, probability of being filled much greater than 96%.
The bottom line is it is statistically extreme that both 26A and 26B were vacant.
Interesting extract from the NTSB 7th Dec statement (ref AvHerald):
” During the explosive decompression the cockpit door flew open as designed and impacted the forward lavatory door shutting it jammed, there were concerns somebody might have been trapped in the lavatory, nobody was in the lavatory at that time however”.
The ‘as designed’ wording seems surprising, and would seem to introduce doubt about any blowout panel built in to the cockpit door?
” During the explosive decompression the cockpit door flew open as designed and impacted the forward lavatory door shutting it jammed, there were concerns somebody might have been trapped in the lavatory, nobody was in the lavatory at that time however”.
The ‘as designed’ wording seems surprising, and would seem to introduce doubt about any blowout panel built in to the cockpit door?
The photos on theaircurrent.com clearly show loose fastners (bolts) but these are NOT the locking bolts! One photos shows two bolts attachung a spring assist bracket (machining) that are several turns loose!
https://pbs.twimg.com/media/GDWRH6OX...g&name=900x900
https://pbs.twimg.com/media/GDWRH6OX...g&name=900x900
That's not good. Those should have prevailing torque nuts on the hidden ends so that they cannot become several turns loose. Those joints will be adjusted at the initial install and that should be at Spirit, and not removed or touched anywhere else for any reason. That looks like someone put the bolts and washers in place, put the nuts on finger tight, until the internal deformation grabbed, and then forgot or was distracted before finishing the task. The prevailing torque nuts must still be in place or the bolts would have vibrated out of the holes.
Join Date: Apr 2010
Location: Alexandria VA USA
Posts: 40
Likes: 0
Received 0 Likes
on
0 Posts
That's not good. Those should have prevailing torque nuts on the hidden ends so that they cannot become several turns loose. Those joints will be adjusted at the initial install and that should be at Spirit, and not removed or touched anywhere else for any reason. That looks like someone put the bolts and washers in place, put the nuts on finger tight, until the internal deformation grabbed, and then forgot or was distracted before finishing the task. The prevailing torque nuts must still be in place or the bolts would have vibrated out of the holes.
None of this is good.
Join Date: Apr 2010
Location: Alexandria VA USA
Posts: 40
Likes: 0
Received 0 Likes
on
0 Posts
Correct me if you think I'm wrong, but play in those two bolts, particularly if repeated at the corresponding assembly at the other (aft/forward) end of the door plug, could have the effect of allowing play in the door position, in a longitudinal direction. That could compromise the stops on one side of the door or other and/or the engagement of the guide roller in the guide on that side.
Edit: They do hold the bottom of the door tight against the lugs, but with the upper guides and pins in place, the door can’t move too far inward.
Last edited by Old Ag; 8th Jan 2024 at 21:54.
Further to my post at #347 and Old Ag at #450, without at least one of four safety bolts plug is only held in place by gravity; resting on rollers.
Then under low or negative G, lift assist springs can eject the plug?? That’s neglecting any friction against stops due to cabin pressure.
So I’m visualising the plug jiggling upwards in minor turbulence before cabin is pressurised and ejecting when pressure builds up.
Then under low or negative G, lift assist springs can eject the plug?? That’s neglecting any friction against stops due to cabin pressure.
So I’m visualising the plug jiggling upwards in minor turbulence before cabin is pressurised and ejecting when pressure builds up.
Last edited by ozaub; 9th Jan 2024 at 21:49. Reason: Minor corrections
Correct me if you think I'm wrong, but play in those two bolts, particularly if repeated at the corresponding assembly at the other (aft/forward) end of the door plug, could have the effect of allowing play in the door position, in a longitudinal direction. That could compromise the stops on one side of the door or other and/or the engagement of the guide roller in the guide on that side.