Alaska Airlines 737-900 MAX loses a door in-flight out of PDX
Join Date: May 2010
Location: SOF BG/EU
Age: 63
Posts: 90
Likes: 0
Received 0 Likes
on
0 Posts
Wow, is NTSB Chair saying it could be airframe itself?
"He (Calhoun) called me and said they've made errors in the past, and they want to rectify that," she said. "Great, but my focus is less on the executive team and more on what happened here with this aircraft"
Homendy said the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure. So far the investigation has not been able to establish whether the door plug was outfitted with the four bolts that prevent it from vertical movement, but Homendy said it is too early to say whether the root cause was missing or wrongly installed bolts.
"We're also looking at the seal. We're looking at, was there any sort of structural flexing of the aircraft?" she said. "It may not be bolts.
https://www.reuters.com/business/aer...st-2024-01-18/
Homendy said the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure. So far the investigation has not been able to establish whether the door plug was outfitted with the four bolts that prevent it from vertical movement, but Homendy said it is too early to say whether the root cause was missing or wrongly installed bolts.
"We're also looking at the seal. We're looking at, was there any sort of structural flexing of the aircraft?" she said. "It may not be bolts.
https://www.reuters.com/business/aer...st-2024-01-18/
Wow, this IS significant. Paraphrasing:
Boeing to NTSB: Hey, we're going to fix them bolts, put them all and make double sure they're nice and tight. And we're ready to go! Nothing else to see here.
NTSB to Boeing: Well, it may not be them bolts, but some structural flexing of y'r airframe.
Could it be that NTSB meanwhile found some more clues in or around the doorframe, in addition to the ones possibly observed in the inside corners of the doorframe near the top part of the seal (not the ones at the top corners of the door plug itself)....
Last edited by D Bru; 20th Jan 2024 at 09:51. Reason: typo + precision
Wow, this IS significant. Paraphrasing:
Boeing to NTSB: Hey, we're going to fix them bolts, put them all and make double sure they're nice and tight. And we're ready to go! Nothing else to see here.
NTSB to Boeing: Well, it may not be them bolts, but some structural flexing of y'r airframe.
Could it be that NTSB meanwhile found some more clues in or around the doorframe, in addition to the ones possibly observed in the inside corners of the doorframe near the top part of the seal (not the ones at the top corners of the door plug itself)....
Boeing to NTSB: Hey, we're going to fix them bolts, put them all and make double sure they're nice and tight. And we're ready to go! Nothing else to see here.
NTSB to Boeing: Well, it may not be them bolts, but some structural flexing of y'r airframe.
Could it be that NTSB meanwhile found some more clues in or around the doorframe, in addition to the ones possibly observed in the inside corners of the doorframe near the top part of the seal (not the ones at the top corners of the door plug itself)....
Probability of losing sleep over the cause of the release of the door being flex of the structure... Low.
Last time I looked at an hull cutout like that, there was one striking feature, and that was the amount of structure that surrounds the aperture. When doing letter checks, I spend time in the mess for the pleasure of learning more about the aircraft, and watching engineers in their playpen. It is easy to deride errors that occur, but generally no matter where in the world you are, the engineers are professional, engaged and capable. There have been far more aircraft bent by drivers than spanners.
Same for the design groups, they are generally professional, and from my perspective, pretty damned conservative. Trying to open up their views even with video evidence is not an easy chore, they are conservative, and understand their responsibilities. The MAX, KC46 and pretty much any other product in recent times that came from the loins of Bills progeny aside, their products are not bad. Between this being a structural live load issue or someone doing a brain freeze while overworked keeping shareholders at bay, I'd go for the razor. IMHO.
To be clear, there would need to be around 1" minimum movement to misalign the stops around the periphery, and the door still must rise to free the upper guides from precluding the top of the door rotating outwards. That lift required looks to be around 1 1/2" or a bit more, it isn't less. Had that happened, I would lay odds that the bolt would have deformed the guide rather graphically, and there is no evidence of that in the images so far which are of less than forensic clarity, but good enough.
Colour me skeptical on fuselage flexing, which is actually talking about the aperture frame flexing by those levels, - it could be made of fettuccine and flex less than that amount.
Homendy said the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure. So far the investigation has not been able to establish whether the door plug was outfitted with the four bolts that prevent it from vertical movement, but Homendy said it is too early to say whether the root cause was missing or wrongly installed bolts.
The Reuters article quoted a couple of posts previously said "So far the investigation has not been able to establish whether the door plug was outfitted with the four bolts that prevent it from vertical movement".
"He (Calhoun) called me and said they've made errors in the past, and they want to rectify that," she said. "Great, but my focus is less on the executive team and more on what happened here with this aircraft"
Homendy said the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure. So far the investigation has not been able to establish whether the door plug was outfitted with the four bolts that prevent it from vertical movement, but Homendy said it is too early to say whether the root cause was missing or wrongly installed bolts.
"We're also looking at the seal. We're looking at, was there any sort of structural flexing of the aircraft?" she said. "It may not be bolts.
https://www.reuters.com/business/aer...st-2024-01-18/
Homendy said the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure. So far the investigation has not been able to establish whether the door plug was outfitted with the four bolts that prevent it from vertical movement, but Homendy said it is too early to say whether the root cause was missing or wrongly installed bolts.
"We're also looking at the seal. We're looking at, was there any sort of structural flexing of the aircraft?" she said. "It may not be bolts.
https://www.reuters.com/business/aer...st-2024-01-18/
Homendy said the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure. So far the investigation has not been able to establish whether the door plug was outfitted with the four bolts that prevent it from vertical movement, but Homendy said it is too early to say whether the root cause was missing or wrongly installed bolts.
"We're also looking at the seal. We're looking at, was there any sort of structural flexing of the aircraft?" she said. "It may not be bolts.
"We're also looking at the seal. We're looking at, was there any sort of structural flexing of the aircraft?" she said. "It may not be bolts.
I've was in this business long enough to know you 'never say never' - even a 10-9 event will happen one in a billion times.
That being said - that basic door plug design is on thousands of aircraft with millions of flight hours and cycles. If the design was truly deficient, the probability that it would show up - for the very first time - on a brand-new aircraft with just a few months of service is really, really low. Since it happened on a newly built aircraft, it almost certainly needs to be an assembly error of some sort - although not necessarily the bolts. It is possible some part of the interfacing hardware was no done properly.
I'm sure that's the focus of the NTSB if they can't verify that it was the bolts.
That being said - that basic door plug design is on thousands of aircraft with millions of flight hours and cycles. If the design was truly deficient, the probability that it would show up - for the very first time - on a brand-new aircraft with just a few months of service is really, really low. Since it happened on a newly built aircraft, it almost certainly needs to be an assembly error of some sort - although not necessarily the bolts. It is possible some part of the interfacing hardware was no done properly.
I'm sure that's the focus of the NTSB if they can't verify that it was the bolts.
What I found interesting in Homendy’s briefing was this: “the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure”.
One assumes that the original door plug testing data to meet Certification requirements is available, so this “destructive testing” must be something different. Could it be applying failure loads to a door, perhaps even in an installed condition, without the 4 locking bolts? On the face of it this wouldn’t prove anything so maybe the test might be on a door partly disengaged from its stop pads, and/or from the guide roller tracks? Any thoughts?
Either way setting up such testing would I imagine be complex and time consuming; not ideal for Boeing.
One assumes that the original door plug testing data to meet Certification requirements is available, so this “destructive testing” must be something different. Could it be applying failure loads to a door, perhaps even in an installed condition, without the 4 locking bolts? On the face of it this wouldn’t prove anything so maybe the test might be on a door partly disengaged from its stop pads, and/or from the guide roller tracks? Any thoughts?
Either way setting up such testing would I imagine be complex and time consuming; not ideal for Boeing.
Join Date: Jan 2024
Location: LA
Posts: 26
Likes: 0
Received 0 Likes
on
0 Posts
What I found interesting in Homendy’s briefing was this: “the NTSB will move next week onto destructive testing of the door plug, or testing to the exact point of failure”.
One assumes that the original door plug testing data to meet Certification requirements is available, so this “destructive testing” must be something different. Could it be applying failure loads to a door, perhaps even in an installed condition, without the 4 locking bolts? On the face of it this wouldn’t prove anything so maybe the test might be on a door partly disengaged from its stop pads, and/or from the guide roller tracks? Any thoughts?
Either way setting up such testing would I imagine be complex and time consuming; not ideal for Boeing.
One assumes that the original door plug testing data to meet Certification requirements is available, so this “destructive testing” must be something different. Could it be applying failure loads to a door, perhaps even in an installed condition, without the 4 locking bolts? On the face of it this wouldn’t prove anything so maybe the test might be on a door partly disengaged from its stop pads, and/or from the guide roller tracks? Any thoughts?
Either way setting up such testing would I imagine be complex and time consuming; not ideal for Boeing.
The stops didn't fail, so the only other option would be the bolts. Probably going to put the bolts in and a hydraulic jack and see how hard they have to push on the door to break all the bolts. Seems excessive as simpler CNC parts could be used , even actual door components, and build a rig around those. The test rig could also be used to estimate fatigue life, though that is unlikely to have been a problem source.
What is evident is that the door springs are strong enough to lift the door if the bolts are not in place, unless some other retaining method is used. The only other retainer is the stop screws - suggesting that someone, instead of using the bolts, just tightened a couple of the stop screws. Fitting the door back into place should be able to identify if any screws are over adjusted. The fall into the trees won't be enough to bend the door. If only the stop screws were holding it, some good thumps could nudge the door along until the screws were right on the edge of the stops. I don't know if this is a worthwhile direction to look as it only helps if using the stop screws to retain the door from shifting up is the design goal.
What is evident is that the door springs are strong enough to lift the door if the bolts are not in place, unless some other retaining method is used. The only other retainer is the stop screws - suggesting that someone, instead of using the bolts, just tightened a couple of the stop screws. Fitting the door back into place should be able to identify if any screws are over adjusted. The fall into the trees won't be enough to bend the door. If only the stop screws were holding it, some good thumps could nudge the door along until the screws were right on the edge of the stops. I don't know if this is a worthwhile direction to look as it only helps if using the stop screws to retain the door from shifting up is the design goal.
Join Date: Jan 2024
Location: LA
Posts: 26
Likes: 0
Received 0 Likes
on
0 Posts
I'd be surprised if they're doing structural testing on the plug itself, since there's no evidence of structural failure. If they want to know if the bolts failed, they'd have to pull a bunch of samples from inventory at Spirit and/or Boeing and they'd probably test them against the standard they're supposed to meet to make sure they meet it if they haven't done such a thing already.
The admission that they still don't know if the retaining bolts were installed means that they may not have found witness marks on anything that they can inspect with it intact. Which probably also means that there's no indication on the edge of the holes that the bolts were yanked against them and failed.
The admission that they still don't know if the retaining bolts were installed means that they may not have found witness marks on anything that they can inspect with it intact. Which probably also means that there's no indication on the edge of the holes that the bolts were yanked against them and failed.
Good question. I was brought up to tell the truth, so wanted to find a gentle way of conveying the reality to this entrepreneur CEO of a startup airline. Like most pax, he had little idea of what was between a sidewall panel and the outside of the aircraft. After a sketch and explanation of fuselage frames, the skin, and window positioning constraints he thought for a while and said "Aha !"
The admission that they still don't know if the retaining bolts were installed means that they may not have found witness marks on anything that they can inspect with it intact. Which probably also means that there's no indication on the edge of the holes that the bolts were yanked against them and failed
Join Date: Jan 2024
Location: LA
Posts: 26
Likes: 0
Received 0 Likes
on
0 Posts
The only argument I see against that is if the springs are strong enough to lift the plug over the stop fittings. If that's true, it should have launched on its first flight.
My working hypothesis is that the bolts were installed but the cotter pins either absent or not bent to lock the castle nuts. The cotters (if present but not bent) eventually vibrated out, then the nuts off, then the bolts out. But then NTSB should be seeing some signs if the bolts having been there at some point. You'd also think there would be marks on the paint around the bolt holes on the upper alignment channels on the plug from the nuts/washers pressing in.
So it's a mystery.
My working hypothesis is that the bolts were installed but the cotter pins either absent or not bent to lock the castle nuts. The cotters (if present but not bent) eventually vibrated out, then the nuts off, then the bolts out. But then NTSB should be seeing some signs if the bolts having been there at some point. You'd also think there would be marks on the paint around the bolt holes on the upper alignment channels on the plug from the nuts/washers pressing in.
So it's a mystery.
Join Date: Jun 2009
Location: Montreal
Age: 65
Posts: 41
Likes: 0
Received 0 Likes
on
0 Posts
I'd be surprised if they're doing structural testing on the plug itself, since there's no evidence of structural failure. If they want to know if the bolts failed, they'd have to pull a bunch of samples from inventory at Spirit and/or Boeing and they'd probably test them against the standard they're supposed to meet to make sure they meet it if they haven't done such a thing already.
The admission that they still don't know if the retaining bolts were installed means that they may not have found witness marks on anything that they can inspect with it intact. Which probably also means that there's no indication on the edge of the holes that the bolts were yanked against them and failed.
The admission that they still don't know if the retaining bolts were installed means that they may not have found witness marks on anything that they can inspect with it intact. Which probably also means that there's no indication on the edge of the holes that the bolts were yanked against them and failed.
The conclusions they draw will lead to actions costing a lot of money, so they have to be able to defend their methods to the nth degree. If there wasn't so much money invloved they could probably just look at the parts they have and say, based on general principles, there's no way a fastener sheared off in those holes, we would see *something*, but we don't.
I looked at the article and the "or testing to the exact point of failure" was not a quote from Homendy, but looks like a clarification written by the author. I haven't found a transcript or video from the actual session. It puzzles me if it's a quote because they don't know the failure mode, so they can't really simulate it in testing, and sticking it on a frame and pulling on it wouldn't be high on my list of tests. But if it's a clarification by the author I think it's a misinterpretation - in my experience "destructive testing" just means there's no "undo" and the part will no longer be the same as it was before. It could be a full up "mount it on a frame and break it" but it's more likely that it involves cutting out parts and subjecting them to analysis in ways that you can't apply to the whole plug. Electron and high resolution optical microscopy of the insides of the through-holes would require destruction. As would any metallurgical analysis, though from everything we've seen so far, there doesn't seem to have been a failure of any of the comonents that have been found. If they know the bolt material they likely should see residue inside the through-holes, but if it's at trace levels they'd need to use an SEM.
Last edited by Mach2point7; 21st Jan 2024 at 04:42. Reason: typo
Join Date: Jan 2024
Location: LA
Posts: 26
Likes: 0
Received 0 Likes
on
0 Posts
If my memory isn't failing me, Homendy said earlier that lab examination of the plug components would reveal whether or not the bolts were installed. I think that was a little optimistic. As far as i know there is no atlas of post-failure fastener holes that can be compared to the indident hardware to draw an indesputable conclusion. There are atlases of fractography for categorizing parts that have fractured, but this situation isn't like that, we are probably talking about deformed/damaged holes. So the testing they are talking about doing now is probably to generate such deformed holes to see what they look like.
The conclusions they draw will lead to actions costing a lot of money, so they have to be able to defend their methods to the nth degree. If there wasn't so much money invloved they could probably just look at the parts they have and say, based on general principles, there's no way a fastener sheared off in those holes, we would see *something*, but we don't.
The conclusions they draw will lead to actions costing a lot of money, so they have to be able to defend their methods to the nth degree. If there wasn't so much money invloved they could probably just look at the parts they have and say, based on general principles, there's no way a fastener sheared off in those holes, we would see *something*, but we don't.
I think that you might be on the right path. Since nobody has ventured a transcript or video of the session, perhaps it is a misunderstanding. Could it mean NDT (non destructive testing) and a) arose from a mis-speak, or b) perhaps the Reuters journalist misheard ? Destructive testing of the key piece of evidence seems rather drama dramatic to me.
Destructive testing is fine if they've gotten all they think they can get from it as-is and have thoroughly documented the configuration.
Join Date: Jan 2024
Location: UK
Posts: 45
Likes: 0
Received 0 Likes
on
0 Posts
On the testing, there was some discussion earlier about the relative strength of the upper and lower lock bolts. The suggestion from a few posters including me was that the lower bolts are much stronger than the upper because of the way they work. Perhaps the testing would be to see what happens if only the upper lock bolts were installed.