Do we need another duplicate posting? First post was 12 hours ago[0] (469 points, 362 comments), and a dupe was 3 hours ago (116 points, 7 remaining comments - was merged into the first)
If this is true, it is impressive the door stayed attached to the plane as long as it did. If they really forgot to install all of the bolts intended to hold it in place, I would have expected to to fail much sooner.
One thing I don't understand is the why the door isn't designed so that pressure doesn't seal it tighter. Like a manhold cover. It seems at least the door needs to be "lifted" to be removed, so the missing bolts hold it down (but presumably do not hold against outward pressure).
I guess that's why it doesn't immediately fall out.
Has anyone seen mechanical or assembly drawings of these things?
My guess is a combination of it would be harder to install that way and designers never considered people would leave the bolts out! Hard to design a plane for such serious assembly oversight.
Competent designers and engineers consider the risks of incorrect assembly, and depending on risk assessment (probability vs severity of failure), and cost-benefit assessment (cost of redesign vs cost of inspection), and based on that decide on if further mitigations are required.
Generally if it's possible that a complete failure of process happens, designers will plan for that. Anything outside of fully automated assembly processes with interlocks and safety checks warrant that to a decent designer. You cite cost of redesign but we're talking design in the first place. What's on the table in this discussion is design, not response to failures.
My understanding, reading this, is that it is the pressure that holds it in place. The bolts just keep is from sliding off.
The doors are very heavy, the pressure keeps them in place in only one direction. Without the bolts, only friction from the pressure would hold them up.
based on the Blancolirio explanation, it would have been held in place by the pressure differential, except that on this flight it got jostled just right at a low enough altitude.
What seems odd is the springs that assist opening an emergency exit are left in place. Might not have made a difference.
Yes exactly, that's what makes me super doubtful of this story. Im mostly familiar with testing in the manufacturing process, but still, even for "in the field" post sales repair this would be very very hard to believe. Badly installed, not up to spec repairs? Sure! Forgetting about... literally all the bolts, in the factory, and then lying about it (???) for a repair that can take a day longer with 0 issues? Then tampering records? This is like 4 tiers of management, workers, inspectors colluding. Again, still possible, not super familiar with that but very very unlikely especially since they claim the door was basically not attached and painted over them bolts (again, ???). What would the endgoal even be at that point for the people who covered this up before it got sent to the airline ?
I would completely believe it if they claimed that they didn't manufacture up to specs, didn't install it correctly, cheaped out on the materials.. but painting over the plugs? Yeah....
I think you read Boeing engineer's post a bit too quickly, because that's definitely not quite what it said.
That said, I worked in commercial aircraft engine component repair for years and everything I read in this is 100% plausible to me.
One time we shipped a part back to the customer before it was even done being repaired. It had labels and stickers marking stuff that needed to be done still on it. When the customer got it they emailed us some photos of the markings and said "It doesn't look done to us...". We had to tell them to send it back so we could finish it. Whoops!
What did I get wrong? I read his two comments but it's very possible I misunderstood something.
I agree that the situation you are describing is very plausible. But correct me if I'm wrong but he's saying that Boeing knew (as in, their systems were aware) about it, even after the botched repair that there was no bolt. In this case, it would be closer to the customer knowing the repair wasn't done and still going forward with not sending it back. No checks are done even after malicious actions from spirit:
>the reason the door blew off is stated in black and white in Boeings own records. It is also very, very stupid and speaks volumes about the quality culture at certain portions of the business
>This entire sequence is documented in the SAT, and the nonconformance records in CMES address the damaged rivets and pressure seal, but at no point is the verification job reopened, or is any record of removed retention bolts created, despite it this being a physical impossibility. Finally with Spirit completing their work to Boeing QAs satisfaction, the two rivet-related records in CMES are stamped complete, and the SAT closed on 19 September 2023. No record or comment regarding the retention bolts is made.
To me that's just hard to believe. Errors happen, but this involves fraud from spirit and also just no process at all from Boeing. I get that the system that deals with this for the 737 is apparently bad, but this just very very basic process management (aka step y needs to be done so that step x can be marked as started or confirmed). In this case, Spirit earlier lied about repairing it, then still didn't repair it... for reasons (hard to procure parts for a manufacturer is an... odd reason to just not fix a critical part especially since it would mean that they actively decided to not install the missing parts). And Boeing caught it the first time but then stopped caring the second? Idk, possible, but I'll wait for sure for more info before drawing conclusions
in the anonymous post the eng mentions that the plug has to lift up before coming out. so it could sit there, unbolted yet pressurized for a while.
the quote:
> there are 4 bolts that prevent the mid-exit door plug from sliding up off of the door stop fittings that take the actual pressurization loads in flight
Sure, but that would still be a worlds first basically. It's not an incorrect install, it's intentionally not doing it and then lying about it. For a random seat related fixture, sure. But withholding critical parts in manufacturing just because you don't want to wait and do a more involved rebuild basically hasn't ever happened in aviation. It's not just short sighted and it's basically impossible for a single individual to have done this and sweep it under the rug. It would involve quite a few people from Spirit to sign off on it and sign their ticket directly to jail because they didn't want to wait. And then openly lying about it to Boeing at first by "painting over the parts", admitting to that lie, and then just not do it even if it's documented that they haven't done it.
I mean sure maybe but at best I'd guess the comment is based on the truth but with heavy hyperbole. And no matter what is said about Boeing, they still have in house inspectors and Qa. Even the laziest QA with no process management other than a spreadsheet would've caught this or raised a red flag. Would've also been believable if QA was overridden by management or something, but nope, nothing was raised according to the comment
probably the banal explanation is that they have two record systems (one official, one where the real notes are) and things get lost in translation between the two. The seal gets fixed, but the bolts reinstall work order doesn't get tracked.
> Context:
CMES – Common Manufacturing Execution System
SAT – Situation Action Tracker
Spirit employees cannot get write access to CMES, by process.
Going back to P poor quality of fuselage deliveries in 2018 Spirit teams have been onsite in Renton & Moses Lake.
It was common that Spirit could be doing out of sequence work all the way after factory rollout.
Frequently Boeing Senior Managers demanded CMES access for Spirit teams, but it can’t be allowed. Access to SAT is fine because it is not a production system.
When Wichita was part of Boeing system access was not an issue.
The only way to integrate Spirit SAT entries & info into CMES is manually, and that is tedious and error prone. Information escapes are unavoidable when 2 companies are working in multiple systems of “record”.
Can someone contextualize this link? On its face, I'm seeing the equivalent of someone posting a big-tech expose on this HN, only for PG himself to reply saying "pls email me at pg@ycombinator.com"; and then another reply suggesting how to avoid being outed just like Journalist X's source was, only for Journalist X to reply with "if you know how my source was outed, please text me at <cell>"
Related question: I've seen news reports that the plane had experienced three recent pressurization issues (cabin pressure warnings). Does anyone know what kind of warning this was? Was it:
- too low pressure, meaning that the door might have been leaking before it blew out.
- too high pressure, meaning that the door had extra force on it.
In my opinion, the fact that this airplane isn't legally an airplane yet, just a collection of parts, is part of the problem. Boeing and Spirit should be criminally liable for mismanaged production, or at the very least have their production certificate for the type revoked.
Ah, so the mystery is solved: the doors were never bolted. Inspections done in Renton caught this in 99% of cases but nobody thought to follow the defect to source and get it fixed there. Classic case of "you can't inspect-in quality".
I don't know, that's very very unlikely. I'm mostly familiar with the 787 inspection process though. still the events listed here are basically harder, longer and involve more bickering than just doing the repair. but who knows... as I said, I just know that this would be impossible with the 787 and way way harder than what is in relative term a simpler repair. Spirit might be shoddy, incompetent, etc but not putting bolts at all and fraudulently tampering with records? Painting over them? Why? Airplanes take longer to repair, or ship routinely, 737 parts are extremely available especially for the literal manufacturer, it's not some sort of insane repair... Again, even assuming sheer incompetence this is too competent
The conclusion is still pretty spot on. The company known (to Boeing) for shoddy work does a warranty repair as a follow-up for a failed warranty repair, and while Boeing QA caught the first repair going wrong it didn't catch the second repair going wrong because those bolts weren't tagged as having been touched.
Still sounds like a clear case of an unbelievable amount of bad quality from Spirit, and Boeing not pressing the issue until something finally slips through the cracks
Even that could've been believable. But , and correct me if I'm wrong, the comment claims that Boeing's systems clearly show that the bolts were never installed. Like, they knew about it even after all of that, and knew about it upon delivery. Again, that just doesn't even make sense from incompetence, greed, laziness or any other angle. It's not like the same people on the factory floor who would want to sweep the bad repair under the
rug[0] also gets to control the delivery process, the massive paperwork, checks, automated checks, etc that are inherent to massively complex manufacturing. Again, this is claiming that the system knew about it, and that it was written there all along. So we are way beyond manufacturing issues here.
[0] (not that they would for what is claimed to be literally just painting over the plugs, without installing anything. that's just dumb, it's not something you can just get away with, it would've been visible from the very first flights and would he basically a world's first)
> the comment claims that Boeing's systems clearly show that the bolts were never installed.
As I read it, it's kind of the opposite. The records show a repair that required removal of the bolts, but there is no record of replacement of the bolts. And there is a record of discussion of if the repair required a formal inspection, which (I think) would have involved recording the status of the bolts.
This is taken as clear evidence that the bolts were not there, but IMHO, if the bolts were missing, that's where they went, but it might not be enough to show they were missing. Also, there's some clear process problems.
I'd sure like to know about the inspections on the other plug doors.
Stupid metaphor but please help me understand, if I ordered a chicken-sandwich and was served with raw-meat, I would not politely return it and request a correction, I would flee the scene.
How are they in any sense whatsoever still working with a company or process that would 'paint over the plugs'? That was the warning sign.
I'm sure the answer is money somehow but is that even accurate? Surely this is more costly? Different budget maybe?
> How are they in any sense whatsoever still working with a company or process that would 'paint over the plugs'?
Because there is no alternative supplier that could supply the same components in a shorter-than-years timeframe. And they don’t want to halt all production as it would bankrupt the company.
To the HN crowd, this is like AWS becoming broken and unreliable, but your entire build and architecture has AWS baked in. Moving to Azure or Google or whatever would take years of engineering and data migration. So you paper over the holes as best you can to keep the lights on.
Well mistakes happen. I wouldn't assume that my main supplier is malicious, because then you have much bigger issues than just a single defective part. The issue isn't that a part was defective here, the thing I have a hard time believing is that the part was not just defective, but that spirit just didn't do the job, knowingly. They didn't have the parts that an actual legit repair/rebuild would have required so they didn't do it. To me that makes very little sense, especially since they apparently tried to just lie about it first too, and then no systems or QA did anything about it.
Raw meat can happen, you catch it before you start eating it. In this case, it's more like you got a pack of used needles instead (pretty much equivalent to a non installed door in an airplane) because the chef didn't want to wait 5mins for the meat to cook, you basically just acknowledge the needles and start eating. That what is hard for me to believe. The original 737max controversy made sense, shortcuts were taken, and it made sense why that sort of disfunction would slowly creep into the design process. It also makes sense that some believed that the redesign would still work since hey, pilots just have to handle the errors if they happen! But in this case? The motivations don't make sense, especially since it apparently wasn't just a mishap or a failure, but intentional fraud.
> With that out of the way… why did the left hand (LH) mid-exit door plug blow off of the 737-9 registered as N704AL? Simple- as has been covered in a number of articles and videos across aviation channels, there are 4 bolts that prevent the mid-exit door plug from sliding up off of the door stop fittings that take the actual pressurization loads in flight, and these 4 bolts were not installed when Boeing delivered the airplane, our own records reflect this.
Although TFA doesn't explicitly say the bolts were there to begin with, the bigger issue is that repairs were required on this door, the repair team lied about what they did, which bypassed a QA check that would have likely caught the missing bolts.
Given the awful (criminal?) behavior of the repair team, it seems reasonable to assume the bolts were initially present and they just forgot to put them back - rather than that they were missing from the start.
I read 'they were never bolted' as 'upon delivery of the aircraft, the bolts holding in the plug were not installed' in order to differentiate from the case of 'the bolts loosened during use of the aircraft after delivery to customer'.
But I can see how we could read to different statements from the parent.
Sure, that's not unbelievable. Tampering with the records is the weird part. I can't think of any reason why spirit or Boeing would be willing to do that when they can just... fix the repair if they knew about it? And if they didn't, then why tamper with the process management systems?
> has in the past 365 calendar days recorded 392 nonconforming findings on 737 mid fuselage door installations
which prompted my comment.
then it says that the door in question was re-worked by Spirit employees.
It implies that the bolts were removed and not re-fitted, but I don't see evidence that they were ever fitted. Perhaps they were, but then removed by people working for a company that day to day doesn't fit bolts, or perhaps they were never there and the people working for the non-bolt-fitting company didn't see the problem with that.
Maybe you're in a company where you have to record the process of making instant coffee.
You have your screwed container of coffee you have a cup and a kettle of hot water.
Someone later finds the coffee container open and spilled by the resident cat.
There's no record of the container opened but there's requests for coffee and completed work orders for coffee have been billed which you can't do without opening the coffee container
[0]: https://news.ycombinator.com/item?id=39102021