Alaska Airlines 737-900 MAX loses a door in-flight out of PDX
An investigation? Good for them. I'd suppose that this door was not intended to depart the plane in midair, so I can intuit that, yes, Boeing failed.
Better close those barn doors. The horse is out.
I get that the enforcement by the FAA has to follow this path, but this is a path that can be put onto a post-it note.
I feel sorry for the FAA employees. They will once again be used as a punching bag by the politicians that short them necessary funds to both fully staff and attract enough top people to oversee every aspect of modern aviation. I guess we can expect a bunch of very concerned Representatives openly asking if they should cut money from the FAA budget until they can prove they can do better with less, knowing that if FAA can the Reps will see that as a reason to keep the funding low, and possibly cut it further.
Better close those barn doors. The horse is out.
I get that the enforcement by the FAA has to follow this path, but this is a path that can be put onto a post-it note.
I feel sorry for the FAA employees. They will once again be used as a punching bag by the politicians that short them necessary funds to both fully staff and attract enough top people to oversee every aspect of modern aviation. I guess we can expect a bunch of very concerned Representatives openly asking if they should cut money from the FAA budget until they can prove they can do better with less, knowing that if FAA can the Reps will see that as a reason to keep the funding low, and possibly cut it further.
Join Date: Apr 2010
Location: Alexandria VA USA
Posts: 40
Likes: 0
Received 0 Likes
on
0 Posts
1) How much upward force they exert on the plug when it's in the closed position? Not sure what you mean. In the intended "closed position" the answer is zero. In the pulled-closed but not bolted position the answer is > 31.5 lbs for each hinge.
2) There should be zero interference between the stop pads and stop pins on the ground at zero differential pressure
3) IMO there is no way the plug was sealed behind the interior plastic trim without at least one "placed" bolt. Whether that bolt also had a castellated nut with or without a split pin is another question.
2) There should be zero interference between the stop pads and stop pins on the ground at zero differential pressure
3) IMO there is no way the plug was sealed behind the interior plastic trim without at least one "placed" bolt. Whether that bolt also had a castellated nut with or without a split pin is another question.
This would mean when its closed with stop bolts installed, there will be some pressure on one or more of the stop bolts resisting the upward movement of the door due to the assist spring pressure.
And if what you are saying is correct, it supports the theory that the airplane was flying around with at least one stop bolt somewhat in place, or the door would have never stayed closed. However, pressure on any of the bolts, nut or not, would make it more difficult for the bolt to fall out.
I find it unlikely that all four stop bolts, even with the nuts removed could fall out in less than 200 cycles. The loose bolt theory works better without the stop bolts having to resist the springs forcing the door off the stops.
I keep retuning to the sticky door no bolt theory. This would require a mechanism that would allow the door to be closed and held in place without the stop bolts. That would be an interference fit with the stop pads or maybe friction from the seal.
We shall see.
Last edited by Old Ag; 12th Jan 2024 at 02:31.
Join Date: Dec 2014
Location: Schiphol
Posts: 479
Likes: 0
Received 0 Likes
on
0 Posts
Production-, transport-, final assembly configurations
As expected you can find multiple production-, transport-, and final_assembly-configurations. Would take a lot of time to validate and sequence them.
Prelim - plug config rolling out of Spirit - note opening angle and delivery without escape hatches (papered over):
Prelim plug during transport:
Prelim 737-9 door - posted earlier but - note items as sign ‘no slide installed’ and wiring:
Prelim - plug config rolling out of Spirit - note opening angle and delivery without escape hatches (papered over):
Prelim plug during transport:
Prelim 737-9 door - posted earlier but - note items as sign ‘no slide installed’ and wiring:
Join Date: May 2011
Location: U.S.
Posts: 84
Likes: 0
Received 0 Likes
on
0 Posts
What I mean under your point number 1 is: How much force is needed to hold the plug into the closed position without any stop bolts? If what you are saying is correct, if you pull the plug into the closed position, then let it go, it will pop up and over the stop pads on its own. Then a little push toward the outside opens it.
This would mean when its closed with stop bolts installed, there will be some pressure on one or more of the stop bolts resisting the upward movement of the door due to the assist spring pressure.
And if what you are saying is correct, it supports the theory that the airplane was flying around with at least one stop bolt somewhat in place, or the door would have never stayed closed. However, pressure on any of the bolts, nut or not, would make it more difficult for the bolt to fall out.
I find it unlikely that all four stop bolts, even with the nuts removed could fall out in less than 200 cycles. The upper stop bolts might even have the threaded end pointed slightly downward making the loose stop bolt theory even more unlikely. The loose bolt theory works better without the stop bolts having to resist the springs forcing the door off the stops.
I keep retuning to the sticky door no bolt theory. This would require a mechanism that would allow the door to be closed and held in place without the stop bolts. That would be an interference fit with the stop pads or maybe friction from the seal.
We shall see.
This would mean when its closed with stop bolts installed, there will be some pressure on one or more of the stop bolts resisting the upward movement of the door due to the assist spring pressure.
And if what you are saying is correct, it supports the theory that the airplane was flying around with at least one stop bolt somewhat in place, or the door would have never stayed closed. However, pressure on any of the bolts, nut or not, would make it more difficult for the bolt to fall out.
I find it unlikely that all four stop bolts, even with the nuts removed could fall out in less than 200 cycles. The upper stop bolts might even have the threaded end pointed slightly downward making the loose stop bolt theory even more unlikely. The loose bolt theory works better without the stop bolts having to resist the springs forcing the door off the stops.
I keep retuning to the sticky door no bolt theory. This would require a mechanism that would allow the door to be closed and held in place without the stop bolts. That would be an interference fit with the stop pads or maybe friction from the seal.
We shall see.
I was thinking the spring force upwards always remains the same on the plug, until a lower stop bolt is placed, at which time the stop bolt receives all of the spring force. One of the questions (which I think you are getting at) is how much resistance to upward motion is present from the static friction of the seals.
I just had a thought when looking at the photo below. Could a passenger ever dislodge an unbolted plug upward, say with an aggressive window shade lift? Not sure if there is another trim layer that goes on top of the white window cover that we see, which is attached to the plug...
Last edited by lateott; 12th Jan 2024 at 03:22. Reason: Typo
Inbound pax 26A
I think I have read every post assiduously but didn't see whether 26A was occupied on the inbound to PDX prior to the incident flight. If so, has the occupant been identified and interviewed to see whether they noticed any whistling or temperature change? I recall that the Singapore A380 door incident and Baku diversion which led to the door redesign program was preceeded by pax complaining of cold and whistling noises into LHR. It would help solve the prior px alarm issue which, if you subscribe to occam's razor, is a stretch to disconnect from the incident flight.
I have been wondering all along just how they get those upper guide locking bolts into position, without specially made pliers like a forceps it would be the job given to a small fingered Apprentice (at least in my training days) of course some blue tack holding it temporarily in the outer side of the track before the plug was pulled into the aperture might work.
I considered the guide roller being placed in the track before the plug was offered up and then bolted when mated to the frame serrations, but that would entail many loose parts.
My crusty size XL hands are no longer dexterous for such a job; my similar task as an Apprentice was wire locking the recharge
points on a Trident main oleo whilst the night shift rested in the crew room!
I considered the guide roller being placed in the track before the plug was offered up and then bolted when mated to the frame serrations, but that would entail many loose parts.
My crusty size XL hands are no longer dexterous for such a job; my similar task as an Apprentice was wire locking the recharge
points on a Trident main oleo whilst the night shift rested in the crew room!
Having reread Seattle Times article and watched Chris Brady’s latest video, there’s sufficient confusion about which way’s up that I’ll risk sticking with my assumption at #678 that thrust from “Lift Assist Springs” is less than weight of plug. Though I thank lateott for telling us that plug weighs 63 lbs., much less than I guessed.
Certainly on proper doors the springs must lift the guide fittings clear of their rollers and stop fittings clear of their pads, so that door can open. But such an arrangement would be quite illogical for plug that must (should) stay shut.
On a side issue raised by Brady and others, the door configuration used on 737 used to be known as “semi-plug”. It was allowed only because door had to move up before it could move out. Whereas a true plug is bigger than its hole.
Certainly on proper doors the springs must lift the guide fittings clear of their rollers and stop fittings clear of their pads, so that door can open. But such an arrangement would be quite illogical for plug that must (should) stay shut.
On a side issue raised by Brady and others, the door configuration used on 737 used to be known as “semi-plug”. It was allowed only because door had to move up before it could move out. Whereas a true plug is bigger than its hole.
Last edited by ozaub; 12th Jan 2024 at 08:11. Reason: Deleted spurious comparison with A320 door
Amazing that with 63 lbs a plug, and you have two, one each side, there is no option to get the tube frame without those openings at all. 1kg mass is about 100L Jet-A1 a year for normal airline use. If I assume that without an airframe opening you get 15 lbs we talk about 4T of Jet-A1 every year only two carry two plugs.
Join Date: Oct 2017
Location: Quimper
Posts: 2
Likes: 0
Received 0 Likes
on
0 Posts
Vertical motion capability
This is my thinking of what could appened in this case :
In the absence of bolts avoiding vertical motion of the plug , you have a system with
This give you a dynamic system with a aprox frequency of 50 Hz if the motion is not contradicted that could be the case by the vertical stop on the upper slot when the lug go down in contact.
Also there is no or very low delta pressure in the phases of ground or near ground opreration.
With the taxi, take off flight, landing even low vertical acceleration with low delta pressure, This the plug could ramp up the slot with the vibration frequncy, desangage the 12 pads and have only the slot resistance to conteract the delta pressure loads.
The upper guide seems to not have been designed for those loads and could break when the delta pressure loads exceeded the guide capability with the roller down on the guide.
It is just my thinking.
T
In the absence of bolts avoiding vertical motion of the plug , you have a system with
- a weight the plug (63 Lbs) , 28 Kg
- springs with a stifness of 50 Lbs force(the plug weight part taken by the ,springs versus the plug displacement to engage in the upper guide fitting slots (4 inches) ) aprox: 12, 5 lbs/force by inch (2180 N/m)
- damping, frictions loads provided by the 12 pads (stop fitting) with the delta pressure loads ) aprox 2400 Lbs/ Psi delta pressure
This give you a dynamic system with a aprox frequency of 50 Hz if the motion is not contradicted that could be the case by the vertical stop on the upper slot when the lug go down in contact.
Also there is no or very low delta pressure in the phases of ground or near ground opreration.
With the taxi, take off flight, landing even low vertical acceleration with low delta pressure, This the plug could ramp up the slot with the vibration frequncy, desangage the 12 pads and have only the slot resistance to conteract the delta pressure loads.
The upper guide seems to not have been designed for those loads and could break when the delta pressure loads exceeded the guide capability with the roller down on the guide.
It is just my thinking.
T
Join Date: Jul 2021
Location: FL95
Posts: 32
Likes: 0
Received 0 Likes
on
0 Posts
Too many JPG artifacts in that picture to be sure, could be there though as some edge needs to be present to cause these artifacts in the first place - in that cosine transformation which JPG is.
Last edited by C2H5OH; 12th Jan 2024 at 06:31. Reason: pic
Join Date: Jul 2021
Location: FL95
Posts: 32
Likes: 0
Received 0 Likes
on
0 Posts
In Chris Brady's videos, he rightly labels them "lift assist springs". It's someone either inside or outside of the aircraft, pulling down the actuation lever, that provides the additional force required to lift the door off the rollers. The springs were originally fitted because of the difficulty in opening the door purely by brute force.
Last edited by DaveReidUK; 12th Jan 2024 at 08:05. Reason: typo
...and it isnt just the fuel cost variable... Adding options (was it four in the current case?) increases complexity, maintenance cost, and the chances of a "quality escape" during assembly.
Obviously there's a tradeoff between seat density, weight, lifetime revenue, maintenance and regulatory compliance and you can be sure that TBC beancounters have client spreadsheets so that the airline can pick the option tha suits their route network/traveller profiles (F/J/Y) etc. and maximises profit. But I am willing to bet that the financial and reputational cost of this episode kicks the cost of simply giving the option of a standard door (lose 2 seats?) or a disabled/inop door with no slide empennnage (lose 1 seat?) into touch. Complexity is the enemy of quality and the mantra for TBC might be "Do it simply - do it right - time after time"
Obviously there's a tradeoff between seat density, weight, lifetime revenue, maintenance and regulatory compliance and you can be sure that TBC beancounters have client spreadsheets so that the airline can pick the option tha suits their route network/traveller profiles (F/J/Y) etc. and maximises profit. But I am willing to bet that the financial and reputational cost of this episode kicks the cost of simply giving the option of a standard door (lose 2 seats?) or a disabled/inop door with no slide empennnage (lose 1 seat?) into touch. Complexity is the enemy of quality and the mantra for TBC might be "Do it simply - do it right - time after time"
A poster earlier mentioned did the 757/767 design ever use Door Plugs, the answer is no.
Both of the types were offered with variable options for overwing Exit Window and Exit Door configurations whether ordered for Legacy, or Charter airline seating,
and this was determined on the build line by the Customer/Lessor.
Door Plugs (or window exit plugs) were seen as far back as the three BAC 1-11 400 series built in 1968 for Channel Airways who selected 4 overwing exit windows to enable 99 charter IT pax, but later Operators plugged/deactivated the 2nd extra pair of window exits.
The BEA Trident 3B also had Door Plugs fitted at the deactivated starboard Mid-Cabin door (the Mid-Cabin Pantry area).
Exit limit (as seen on BA Super Shuttle) was up to 146 pax, but could be increased to 170, with this Door activated for IT Charter flights.
180 pax is mentioned in the HS design brochures (I assume 180 pax would need some 7 abreast seating, but no airline ever did this on the Trident 3).
DaveReidUK will no doubt know more about BEA /BA Trident 3's
The Boeing 707-320C which had a 'Hat Rack' exit door hatch design akin to today's 737M-9, could have this exit plugged if the 707 Operator did not want to fly it as a Combi, nor want to seat more than 189 Pax, which was the Exit Limit without the Hat Rack door activated (same as 737M8-200 and M9).
At BMA British Midland we had 3 x 707C's reconfigured in 1982 with all new wide-look cabins/galleys/seats to enable 212 pax for IT Charter flights,
and although they were fitted with the original Boeing Hat rack doors, these was actually replaced by a new larger Boeing 757 Type 1 exit drop down door with a Slide.
***BMA 707C
The new aft exit door is now larger than the original hat rack door
***BMA 707C
You can see the larger new aft exit door fitted 1982
Both of the types were offered with variable options for overwing Exit Window and Exit Door configurations whether ordered for Legacy, or Charter airline seating,
and this was determined on the build line by the Customer/Lessor.
Door Plugs (or window exit plugs) were seen as far back as the three BAC 1-11 400 series built in 1968 for Channel Airways who selected 4 overwing exit windows to enable 99 charter IT pax, but later Operators plugged/deactivated the 2nd extra pair of window exits.
The BEA Trident 3B also had Door Plugs fitted at the deactivated starboard Mid-Cabin door (the Mid-Cabin Pantry area).
Exit limit (as seen on BA Super Shuttle) was up to 146 pax, but could be increased to 170, with this Door activated for IT Charter flights.
180 pax is mentioned in the HS design brochures (I assume 180 pax would need some 7 abreast seating, but no airline ever did this on the Trident 3).
DaveReidUK will no doubt know more about BEA /BA Trident 3's
The Boeing 707-320C which had a 'Hat Rack' exit door hatch design akin to today's 737M-9, could have this exit plugged if the 707 Operator did not want to fly it as a Combi, nor want to seat more than 189 Pax, which was the Exit Limit without the Hat Rack door activated (same as 737M8-200 and M9).
At BMA British Midland we had 3 x 707C's reconfigured in 1982 with all new wide-look cabins/galleys/seats to enable 212 pax for IT Charter flights,
and although they were fitted with the original Boeing Hat rack doors, these was actually replaced by a new larger Boeing 757 Type 1 exit drop down door with a Slide.
***BMA 707C
The new aft exit door is now larger than the original hat rack door
***BMA 707C
You can see the larger new aft exit door fitted 1982
Last edited by rog747; 12th Jan 2024 at 09:43.
" Are there aircraft in service which have restraining bolts fitted to exits, but which are meant to be operational ? ! "
Itis would be very poor engineering practice to use a disabling device which could be inadvertently misused to prevent a required operation.
The FAA letter identifies Part 21 problems, but the issue also involves Part 25 Mitigating human error in design, engineering, and compliance. Which additionally relates to the overall design, certification, and manufacturing process in both Spirit and Boeing, and also with FAA oversight.
Perhaps another instance of local approval (cf previous MAX grounding), where the FAA delegates their responsibilities to the manufacturer via DER oversight, in-house self-checking, which if in error enables the FAA to blame the manufacturer, whereas the higher responsibility is still that of the regulator.
Modern views of safety in complex systems (any system with human involvement is complex), indicate that active safety responsibility has to move upwards in organisational and certification processes; i.e. fewer delegations.
Ref for info: 'HF guidance for regulators ICAO Doc 10151 Manual on Human Performance (HP) for Regulators' … and manufacturers, and operators, and individuals.
https://www.icao.int/safety/OPS/OPS-...alltext.en.pdf
It
The FAA letter identifies Part 21 problems, but the issue also involves Part 25 Mitigating human error in design, engineering, and compliance. Which additionally relates to the overall design, certification, and manufacturing process in both Spirit and Boeing, and also with FAA oversight.
Perhaps another instance of local approval (cf previous MAX grounding), where the FAA delegates their responsibilities to the manufacturer via DER oversight, in-house self-checking, which if in error enables the FAA to blame the manufacturer, whereas the higher responsibility is still that of the regulator.
Modern views of safety in complex systems (any system with human involvement is complex), indicate that active safety responsibility has to move upwards in organisational and certification processes; i.e. fewer delegations.
Ref for info: 'HF guidance for regulators ICAO Doc 10151 Manual on Human Performance (HP) for Regulators' … and manufacturers, and operators, and individuals.
https://www.icao.int/safety/OPS/OPS-...alltext.en.pdf
Last edited by safetypee; 12th Jan 2024 at 08:52. Reason: Ref#800
The upper restraining bolts go through holes in the guide tracks attached to the door plug. There would be no reason for those holes to exist in the guide tracks on a live door (I don't know whether they do or not, does anyone?).
The lower bolts go through holes in the fitting that hinges down as the door opens. The jury still seems to be out re whether it's the same fitting on the live door, because of the debate on how far the respective doors are able to open and, if they're different, whether the holes exist on both.
So may turn out that it's not actually possible to lock the live door in place with those bolts. We need someone with an AMM.