Is this a fundamental design flaw or did United (or whatever airline) neglect maintenance?
Doesn’t seem like a fundamental design flaw. Maybe a flaw in the maintenance procedures, or neglect on the part of Spirit Aerosystems. I agree with LSLGuy in that I don’t see drilled bolt heads. So, no safety wire, and probably couldn’t be safety wired. Why is that? Dunno yet.
This also assumes that the problem shown here is the same as the problem as Alaska, though that’s seeming like a good bet at the moment.
Well I’ll be damned! So the 737 MAX has a “Check Engine” light! 
Yep, I have, and I get it. And what it often tells you is “don’t worry about it for now”, which is probably also analogous with aircraft.
Per @Racer72 upthread who used to do assembly and QC on Boeing fuselages, that plug is installed at the subcontractor in Wichita that builds the whole fuselage. That’s Spirit Aerosystems, a ~20 years ago spin-off from Boeing.
Once installed the plug is a life of the airframe thingy. It might need to be looked at every 10-15 years during the total teardown inspection (“D check”) , but otherwise its maintenance-free. If designed and assembled correctly. Good bet not a single employee or sub-contractor at UAL laid eyes on that work since the airplane was delivered new to them. Nor did they have any reason to want to.
Pretty much. Although it’s more of a "check complete jet from nose to tail light.
In the case of the airplane it often results in "“Reset relevant system unit 21.A.B.E.Q in accordance with maintenance procedure 21.2.3.4.5.6.B.4.e, performed BITE check per maintenance procedure 21.2.3.4.5.6.J.4.a, observed no faults, OK for service.”
If it keeps happening, or the gizmo flunks BITE, now that gets folks attention that they’ve got a real fault, not a glitch. Which in turn leads to more extensive testing, replacing boxes in hopes of making it go away, or actually getting out a wrench & sorting out which mechanical thingy isn’t thingifying properly.
In the US Navy, this is also known as “troubleshooting” aka “hit it hard with said wrench and see if it starts working”.
ETA: Surprisingly often, this is an effective maintenance evolution.
Looking even more closely at that pic, here’s the bottom bolt again:
See the gap between the shank of the bolt and the drilled hole it goes through? I can think of only two reasons for it to look like that: it was made into a slot intentionally for adjustment, or it started as a circular hole but stretched/wore through the bottom over time.
An adjustment slot doesn’t make much sense to me due to the spring taking up any minor misalignment. But given how the bracket appears to be shoved upward, and that the bolt is undoubtedly a harder material than the bracket, I could easily see it wearing through the bottom of the hole over time, with the tension provided by the spring and vibration providing the energy input.
Wasn’t that two months ago?
‘My old grandmother used to say anything mechanical, give it a good bash.’ (from The Longest Day (1962))
I dunno.
It’s 2 of 4 bolts holding 1 of only 4 attaching points to the door frame. This is holding a door in place that is under significant internal pressure to push into the slipstream. It’s not designed for this kind of stress and more importantly wasn’t tested for it. At 500 mph a loose part is just begging to resonate. And once a part starts to resonate the time to failure is measured in seconds.
This doesn’t speak well for Boeing at all.
Agreed. Looking at the photo of the door frame with that bracket still attached to the hinge pin, it’s not clear whether those bolts had loosened and come off or they were just sheared off by aerodynamic forces, but it’s clear that they’re more important than I first thought.
I thought all modern jet doors were plugs that cannot pop out without taking a chunk of the fuselage with it. Indeed, they are stronger than the rest of the fuselage (kinda like my front door is more durable than the walls around it…it’d probably be easier to break into my home through the walls around my door rather than through the door itself).
The aircraft carrier I worked on was built in 1943 and significantly upgraded in 1953-1955. I worked the arresting gear. Anything critical held together with blind bolts or bolt/nut combos was safety wired if subject to mechanical stress via vibration. There were exceptions - mostly covers for things that would be dangerous if uncovered, but lack of a cover wouldn’t cause failure and non-critical parts. I can’t tell you how many hours were spent cutting safety wire to take something apart and then putting the safety wire back after reassembly. I can tell you I got pretty damn good with the safety wire spinning pliers.
And anything not wired was regularly checked (i.e., usually twice daily during flight ops) for tightness and wear.
This is, in my opinion, an egregious design failure on the part of Boeing. New evidence may change my opinion, but it would have to be of the “even if all the bolts failed, the plug can’t pop out” kind of evidence.
737 mid-exit doors open outwards, so no, they don’t fit the same way as the plug-style doors.
My dad was Combat Aircrew on CV-47 Philippine Sea (‘Syphilitic Pee’) during the Korean War, then CV-16 Lexington on (I think) the first cruise after she got her angled deck.
I’m not an engineer so anything I make is over-engineered by necessity. With that said I’m a bit curious about the upper attaching door mechanism. It consists of what looks like a 1/2 inch or 3/4 stud that slides into the door channel. Pinning it in place is fine for keeping it from moving up and down but there’s nothing stopping it from pulling out of the channel sideways. If there’s a 1/16 gap between the stud in the door frame and the door channel it rides in then that means there’s 1/8 inch of play back and forth. How much flex is there on a fuselage? It doesn’t have to pull out of the channel to create stress on the channel because there’s less contact between the 2 parts while retaining the same stress load. I’d like to see a bolt sideways between door and frame so the gap reminds constant.
Alaska Airlines has also found “loose hardware” on other aircraft.
As our maintenance technicians began preparing our 737-9 MAX fleet for inspections, they accessed the area in question. Initial reports from our technicians indicate some loose hardware was visible on some aircraft.
Sorry for the hijack, but that’s my Lady. If your dad is still active enough, I highly recommend a visit to her at Corpus Christi. The museum society there is utterly fantastic. Contact them ahead of time and they’ll ring him on-board as “former shipmate so-and-so arriving”. My wife did it for me when we visited. I teared up.
And if he has passed, you should still visit. The last active-service Navy ship with WWII battle flags. She was a hell of a ship.
My point above was that, even in the '50s engineers knew that vibration and stress worked bolts loose and took steps to mitigate the issue. the fact that a modern airliner built in 2023 didn’t have that mitigation for a structurally key part in place is beyond ridiculous.
He died in November, 1998. It would be nice to visit Lex; and also USS Arizona, where my great-(great-?) uncle is. (A Doper is or was a docent at the memorial, and is quite familiar with his name.)
The plug has 12 mounting points that overlap in such a way that the door can’t pop out unless all the bolts are out and the plug is moved upwards to clear them. Also hinges and dogs. It seems there were multiple failures in this incident.
Or a “common cause” failure. If 4 of those attachment points are the same as each other, for example, and the design is such that it’s possible that they are able to come loose with vibration over time, then it should be considered a single failure mode and the design modified or mitigated, especially if the failure of all 4 could prevent continued safe flight and landing.
These types of failure analysis, with the goal of identifying latent, dormant, common cause and other more insidious failures is at the core of significant regulatory overhaul for the standards and guidance related to the subject matter. There’s also increased scrutiny on the so-called “changed product rule” which dictates when and why an aircraft redesign or modification is required to comply with the latest standards instead of the ones originally used on the program (analogous to building codes, and how you obviously don’t have to rewire your home to change a light bulb, or add a light fixture, though the new fixture has to meet new code, or as close as possible, etc.). The existing CPR guidance may be too permissive.
These issues were a significant contributor to the crashes of the MAX aircraft and I suspect will contribute to this one.
I can see from the aircraft type certificate data sheet that the doors largely complied with the second-most recent door standard. That standard was significantly rewritten at the most recent version and CPR would have allowed that compliance is “not practicable”.
I’m trying not to speculate too much on what specifically happened, but the spotlight on some of the trickiest aspects of airworthiness and certification is fascinating. Great opportunity to learn from, even only based on hypothetical scenarios.