Leeham News and Analysis
There's more to real news than a news release.
Leeham News and Analysis
- Why the IAM 751 rejected the Boeing contract and what is needed for approval September 16, 2024
- Boeing CFO: we want to negotiate; taking steps to preserve cash September 13, 2024
- Bjorn’s Corner: New engine development. Part 23. New versus old, GTF versus V2500 September 13, 2024
- IAM files Unfair Labor Practice against Boeing in thumping rejection of contract September 13, 2024
- IAM rejects Boeing contract, strikes at midnight September 13, 2024
Boeing defends 787, 777 against whistleblower charges
By Scott Hamilton
April 17, 2024, © Leeham News: A whistleblower appeared before the US Senate today recommending that the entire fleet of Boeing 787s be grounded until inspections can be performed to assure safety.
The whistleblower, Sam Salehpour, is a Boeing engineer who worked on the 787. He claims he was moved off the program by Boeing in retaliation for raising safety concerns about the 787 and the 777. Boeing denies this charge.
Salehpour went public with his safety charges a week ago. He focused on the small gaps between fuselage sections and other areas on the airplane that failed to meet Boeing’s own specifications. Production gaps, where parts of the airplane are mated, are common. Boeing and other manufacturers use shims to fill these gaps.
This illustration, which is not to scale, shows how gaps develop, how joins are pulled together and how shims fill gaps that remain. The gaps are 0.005 to 0.008 inches wide–about the thickness of a piece of paper. Source: Boeing.
In 2020, Boeing revealed that in some cases, the gaps were greater than the 0.005 inches of its own specifications. Gaps of 0.008 inches were found. The gaps are the thickness of a piece of paper. Boeing initially grounded eight 787s for inspection.
In October 2020, Boeing suspended delivery of the 787 for what would eventually be 20 months. Deliveries already had been deferred by customers because of the COVID-19 pandemic. Ultimately, Boeing has 110 787s parked that were completed. After a lengthy process with the Federal Aviation Administration, the FAA approved Boeing’s fix. The fuselages of the 110 airplanes have to be inspected and measured. If repairs are necessary, it takes longer (5-6 months and in some extreme cases, 7-8 months) to complete than it does to assemble the airplanes in the first place. There are about 40 787s still awaiting rework.
Responses and testing
Boeing, in a media briefing on Monday, said its inspections and repairs resolved many of the problems claimed by Salehpour. Other issues he raised are “inaccurate,” Boeing claims. Furthermore, inspections of the current in-service fleet revealed no fatigue or safety of flight issues, Boeing said. Testing from 2010-2015 through the equivalent of 165,000 cycles (a takeoff and landing is one cycle) revealed no fatigue issues. The 787 fleet flies an average of 600 cycles per year.
Steve Chisholm, Vice President and Chief Engineer for Boeing Mechanical and Structural Engineering, said that of the more than 1,100 787s in service, 671 completed 6-year maintenance inspections; eight completed 12-year maintenance inspections; 10 had “intensive structural maintenance evaluations;” and “Through all of this, there’s been zero airframe fatigue findings on the 787 fleet, and all these results have been shared with the FAA.”
“Service actions [bulletins] issued for an airplane that’s in service are not unique to the 787 or to Boeing,” Chisholm said during the media briefing. “One could go look at airworthiness directives issued on any program for any OEM and find those on a regular basis.
“When we’re looking at the gap management journey, to date we haven’t identified anything that is safety. We have not closed all of those items. We’re still in that discussion with the FAA,” he said. “We expect to be done by the end of the year. We have not seen anything in service related to gap management that would indicate that there is an issue with the in-service fleet.”
Performing inspections
Separately, a Boeing spokesman—replying to a question from LNA—said, We have done extensive test and analysis to cover the entire fleet and at this time have found no need for in-service action to ensure structural integrity and longevity for composite structure related to gap management. The evaluation continues….”
This figure outlines testing to validate gap solutions. Source: Boeing.
Boeing divides the 787 into three groups:
- Group 1: airplanes produced, delivered, and entered into service before join verification.
- Group 2: airplanes inspected and delivered through Boeing’s join verification process.
- Group 3: Airplanes produced and delivered since 2022 (planes that went into inventory).
Boeing said the third group was never found to be unsafe. “They did not conform, and we cannot deliver airplanes that do not conform – even if they are safe.”
The spokesman also said that “there is no work associated with gap management happening in the [in-service] fleet.”
777 claims
Salehpour, the whistleblower, in addition to urging the 787 in-service fleet be grounded for inspection, asserted that assembly procedures for the 777 violate standards. Boeing, noting the 777 has been in service for three decades, disputed the claims. “We are fully confident in the safety of the 777, which remains the most successful widebody airplane family in aviation history.”
Looking forward to seeing how this latest 787 / 777 story develops and is resolved, along with the John Barnett (RIP) allegations as that suit also moves forward.
I find the Boeing response to be convoluted and inconsistent.
– “…of the more than 1,100 787s in service, 671 completed 6-year maintenance inspections…”
* So, that means that more than 430 787s haven’t been inspected yet.
– “…“Through all of this, there’s been zero airframe fatigue findings on the 787 fleet…”
* He forgot the qualifier “so far”. There were also zero door plug blowouts on MAXs until the Alaska fiasco.
– “We have not seen anything in service related to gap management that would indicate that there is an issue with the in-service fleet.””
* Just wait a little longer: by definition, fatigue takes time to reveal itself. Just look at the rear pressure bulkhead on the JAL 747 that crashed: just because it hadn’t revealed itself yet doesn’t mean that the underlying problem wasn’t there.
___
Here’s the other side of the coin: a disgusting and alarming treatise of what the Senate hearing has revealed so far:
“Boeing’s Broken Safety Culture: 5 Key Insights From Senate 787 Dreamliner & B777 Whistleblower Hearing”
https://simpleflying.com/boeing-whistleblower-senate-hearing-testimony/
Of particular note:
“Following an inspection into gaps in the Dreamliner’s fuselage, Salehpour claims that up to 98.7% of gaps have not been shimmed at all in sections 41/43 or 46/87 of the 787, increasing the chances of a sudden fatigue failure which would be catastrophic. In his estimation, over 1,000 of the widebodies currently in service are affected.”
“Relax; we got this” is not a satisfactory response from that company,
given recent circumstances. Passengers (and airlines) seem to be voting with their feet WRT that company.
Yes, well, that company repeatedly told us that the MAX was the most scrutinized plane in history…and then Alaska came along and the cat was out of the bag.
I wonder: do they believe that PR themselves, of do they just spout it for the rest of us?
Two different problems. Don’t be simple minded. The max was the most scrutinised aircraft as per the design and that is true. Boeing did NOT put bolts in a door ffs. Don’t turn this into a max problem. It’s not
That’s a different problem. You don’t put bolts it doesn’t hold. That’s not a max problem. Let’s remove the bolts on all your chairs and tables and see how that works out for you.
They could do that on a 777 or 787 or 767.
There’s a broad manufacturing quality problem with the MAX.
Also a wiring problem.
Also an overheating engine nacelle problem.
That’s not what you’d expect after historic “scrutiny”.
Basic trend here: try to blind using waffle and PR, and just hope/pretend that all is well.
Finally, a commonsense answer.
I’m curious then, Opus – why do these things seem to happen exclusively to the Max line? As you mentioned, it COULD happen to a 777, 767, 787 or even a 737NG…but it hasn’t.
There are some 1,500 Max’s in service, in a sea of how many BA aircraft flying? 1,100 Dreamliners out there. 1,300 767’s. About the same number of 777’s. All are still being produced, but you haven’t seen them flying themselves into the ground or doors being blow out just months after coming off the line.
The Max had an EIS of 2017, right?
At the same plant, they made the NG right up until 2020. Even today they’re still being produced for the military.
Workers seem to have no problem making those jets – and making them safely.
So while you may protest that this is not a Max problem, and yes – BCA is beset with many issues regarding other models…one model seems to be bitten by the bad luck charm, a lot harder than the others.
Is it not, at the very least…curious?
@Frank P
“… flying themselves into the ground…”
Only if you consider the pilots so lacking in airmanship that they weren’t flying the airplane. That’s a bit uncharitable of you. The data shows that they had basic piloting skills; on each of the three flights that encountered a runaway stab trim they countered the uncommanded pitch deviation and restored pitch attitude and any altitude lost just as any pilot would—by using elevator and then electric trim to assist when the force became excessive, exactly as Boeing predicted. After all, reacting to uncommanded pitch changes—which occur all the time—is the first control action a pilot learns and it becomes a reflexive action. And they did so for each time the runaway manifested and they were paying attention. So they were in control of the airplane and maintaining level flight.
Also, if you are only informed by the news media you’re going to be at their reporting mercy and asking why anything “seems to happen” is moot.
@ Pete P
Hard to believe that there are people still trying to forward the argument of incompetent foreign pilots — that tired argument was shot down years ago by pilots’ unions in the US (see link).
The argument is also intrinsically inconsistent — since those same pilots were perfectly capable of flying 737NGs on a daily basis, without incident. And the (main) difference between a MAX and an NG is …(drumroll)…that amateuristic screw-up called MCAS!
https://www.theguardian.com/business/2019/may/23/boeing-737-max-crashes-american-airlines-pilots-union-mcas
Moreover, Boeing has since formally declared that the MAX crash pilots were blameless…surely you knew that?
Logically,if Boeing is correct about the in service fleet,they should now relax their production tolerances?
It sound like tolerances of 0.002″, 0.005″ are from old sheet metal standards for draftsmen. They have worked fine historically, but it is a huge task to calculate and verify new ones for carbon structures, especially as the autoclave baking quality can vary alot depending on size, number of plies and geometry. The expected live of a widebody is around 7000 cycles in normal intercontinental airline operations and then 3000-4000 cycles in cargo ops. To then design for 70 000 cycles leaves tons of margins.
What I am seeing is incompetence in manufacturing engineering at the top. So let’s think this through. In the old days, if we needed to hold tight tolerances on a part that was being produced one place and then shipped to another where it would have to fit properly with another one, a tooling system would be designed to assure that. The components would be certified tools, and metrology would set the recertification cycle appropriately. The tools would be serialized, since we are talking about multiple shipsets, and then they would be tracked in a database. Certified tool location was one of the infrastructure challenges I was working on when I retired in 2014. This is actually a very thorny problem due to some fundamental design flaws in the lower layers of the internet, battery size issues, and cost and size of active RFID tag options. This is actually a really interesting and quite complicated topic.
From the discussion so far, what I’m hearing is that even the recognition that a certified tooling system is needed hasn’t been happened. The leadership lights are literally out. That is a much bigger problem than the technical details of this one specific issue. It raises the question as to how many other certified tool requirements have not been recognized and addressed.
BTW, I would add that in the design of a really good certified tool used for transporting critically sized parts for alignment with others during assembly, the transportation/protective tool becomes part of the assembly tooling.
This gets back into the importance of manufacturing engineering, and getting rid of that word lean. The process should be though of as choreographing a dance. Things should just flow.
Maintaining a healthy weight manufacturing
Plenty of time manufacturing
Long term shareholder value
Jidoka,stop and fix in Airbus speak,having learnt a hard lesson.
Seems like western companies are picking and choosing Toyota principals
One would expect a measurement protocol of finished part in a fixture reproducing the same fit conditions as during assembly. Either the part meets blueprint tolerances or not and the printed certificate with measurements are attached to part at shipping for Boeing receiving inspection. Actually Airbus had similar problems of forcing wing panels on the A380 to ribfeets in the UK inducing too high loads into causing cracking discovered after the Qantas A380 wing repair after engine failure. Cheating with shimming is a common historical issue as with aluminium alloy airframes the axial loads into the rivets were not as high due to less stiff structure compared to carbon fibre structures forced to close gaps.
A structure the size of a 777 or 787 fuselage barrel Section 46 deforms under its own weight by more than 0.005 in.
That just makes the process of measuring them, and establishing the capability of the processes that make them, a challenge.
Try suspending it. Try fixturing it. Try a specific orientation relative to gravity.
You seem to be hinting that because the parts are flexible there is no ability, and should be no need, to control their dimensions.
This is not about deformation
but about insufficient “flatness”.
( in the overlapping region around fasteners,
see the image up in the article body from Bjoern)
CFRP is stiffer than Al.
.. and B seems unable to get to grips with that issue
even when using an inner male mandrel.
Diameter is a size dimension or measurement. Roundness or circularity is a form requirement or measurement. Two barrels(?) can be of correct mating size but (with or without flexing) not meet the form requirements necessary for mating in assembly.
Yeah, when the two don’t fit together on the line how can it be established whether the failure to fit is a size problem or a form problem?
(Is this a situation where a SPEEA whistleblower goes to make a report while a manager directs AIM machinists to jump on it?)
OK, Boeing has a lot of issues… Still, this crisis is now entering the realm of hysteria…
No. Serial all-fatalities crashes and door blowouts (!) are not evidence of “hysteria”, but of something much worse and more insidious in the culture at that company.
Evidence to the contrary is welcome.
This string makes my point…
The automatic presumption—insistence—that whistleblowers are all-knowing gurus certainly reveals some naivety. The fact that serial all-fatalities crashes were blamed on design deficiencies—by people not qualified to make such assessment, with absurd explanations to boot—is plenty of evidence of hysteria. A dozen or so people chose to ignore the data (likely didn’t understand it) and speculate with their very limited knowledge of the subject. The rest of the masses followed blindly in a hysterical stampede. The data showed quite clearly that the system performed as designed, and that performance was well within a pilot’s capabilities to handle it—and they DID handle it. It certainly wasn’t as the news media (and a few pilots giving fraudulent simulator reenactment demos) described—monstrous secret systems that wrested control from the pilots and could not be overridden, or stunned pilots who didn’t know what to do because training had been “hidden” from them.
I attribute it to prejudiced hysteria rather than a more sinister motive on the basis of the wise words: Never attribute to malice what can easily be explained by ignorance or stupidity.
When the choices that WEREN’T made weren’t chosen because they were expensive it is a pretty good indicator that quality was not the driving force behind the final decision. And from there, like in plumbing, it rolls down hill.
MCAS wasn’t a mistake. Door plugs weren’t a mistake. Running the line, and the business, too much for profit and not enough for quality, would seem to be the root cause of it all.
Where does a single angle-of-attack sensor come in? Is it a design deficiency? How about when a redundant sensor is offered only at extra cost?
“The data showed quite clearly that the system performed as designed,”
the only bit in your screed that holds some truth.
But that “truth” is damning to boot!
The design was criminally haphazard ignoring basic fault analysis and assuming nonexistent pilot performance.
This is fascinating:
“Stephanie Pope, who replaced Stan Deal last month as head of Boeing’s commercial aircraft business, is “committed” to speeding up late deliveries, O’Leary said.”
Any theories as to how Ms. Pope plans to achieve that? Has the “Just Ship It” banner been hung back up?
https://www.bloomberg.com/news/articles/2024-04-17/ryanair-says-boeing-taking-steps-to-accelerate-jet-deliveries
I still can’t quite believe the appointment of that unqualified person (in the works for some time, IMO) as head of BCA- along with their other duties.
Not a confidence-inspiring pick. / weird
Should they be slow forever? Of course they should speed up deliveries where they can. That does not equate to a quality drop. Airbus is building 60+ jets a month and no issues.
Boeing should be aiming to do the same.
BA aiming to have $10b FCF by 2025/26-ish.
‘Should they be slow forever? ‘
They should be slow until they have proven that the aircraft going out the door to customers are safe. Have they done that yet?
As I recall, BA had 90 days to tell the FAA how it was going to change things to make exactly that happen (see new Leeham article).
No. We’ve got some deck chairs being arranged in the c-suite, but that’s about it. But here’s the Chief Operations Officer talking about speeding up deliveries.
Just what investors want to hear…(same old, same old)
“Should they be slow forever? ”
No they should be slow until they can build planes according to standardized processes and consistently have time left over for coffee between planes.
The recent comments out of the Boeing executive suite are baffling.
I read the comment below that CFO Brian West made at a recent Bank of America conference.
“For years, we prioritized the movement of the airplane through the factory over getting it done right, and that’s got to change.”
Perhaps Mr. West would like to expand on what he meant by this at the upcoming Congressional hearings. I imagine the lawyers representing passengers on Alaska Airlines flight 1282 look forward to deposing him.
Question #1 should be:
Given how after the Max crashes you promised to focus and quality and safety, how does that square with your statement.
I wasn’t reassured by Boeing’s statements in its defense. The issue raised was not “fatigue.” I already knew that composites and aluminum age differently. Delaminations don’t grow the way fatigue cracks do. So, using measures appropriate for aluminum do not address concerns about composites.
If Boeing has a requirement for .005 inch gaps, then I would like to see some analysis or prediction of the implications in service from having those gaps. For instance, porosity or water intrusion or UV damage or whatever, followed by freeze-thaw cycles that degrade the bonding between layers or the bonding of matrix from resin.
What actually goes wrong with composites, and what did the data show about those characteristics?
Thanks for this comment.
If two sections are not mated properly, with intervening gaps here and there at the joint, then forces transmitting across the joint will be concentrated at some points more than others. In effect, parts of the joint will suffer dynamic loads that exceed design spec. Given long enough, this could cause effects such as cracking, buckling, de-lamination, etc. I’m not at all convinced that Boeing has (adequately) modeled such effects.
This certainly takes me back to research work I did on fasteners in composite joints back in the early days of the 787 and A350. I remember at the time making some comments here or on A.net about my concerns for the long term…
Any idea how Airbus handles surface roughness and unevenness in the join interfaces?
( the panels are “outer mold”. I seem to remember that longitudinal joins are overlap but the section joins are done with an inner circumference strip?
IMU the frames fixed via brackets solution copes with inner panel unevenness)
@Uwe – I’ve never really thought much about surface roughness since the outer surface is epoxy against the mould and will have a consistently smooth finish. The inner surface will also be quite smooth since it’s vacuum bagged and (I assume) will have additional tooling applied to ensure a good surface at interfaces. The edges of components will be designed to avoid any ply drop-offs, joggles etc. in the area of the join. Some roughness is good where adhesive or other surface treatments are required.
As regards gaps, I believe “liquid shim” is used when absolutely necessary – a paste which hardens after being pressed between the parts – but interfaces will be designed to allow parts to match smoothly and avoid any gaps. The kind of fit problems the 787 barrels experienced are the only time you would expect to have significant gaps appear, and obviously this is something to be limited as much as possible.
Build some trial fuselages with as f’ing close to perfect mating conditions as possible and some with typical gaps and some with excessive gaps. drill them all. Clean the debris out of some of them and leave it in others. Cycle them all until at least half of them fail. Look at how many cycles it takes and which groups fail sooner and which later.
This isn’t a question of the safety of the design, or the simulation of the design. It’s a question of whether leaving debris in a joint where a shim is supposed to be shortens the cycle life of the fuselage.
Do the “inspections” include opening the joints to see if 1) there are shims inside it? 2) if there’s debris from drilling in it?
Why is it affordable to inspect and offer questionable assurances and not affordable to open, clean, and shim (as needed) the joints?
“We can cross our fingers at zero cost”
“Why is it affordable to inspect and offer questionable assurances and not affordable to open, clean, and shim (as needed) the joints?”
hehe!
You can see in the quarterlies that it is not affordable to do inspections post production and convert findings into expensive remedial action.
It would have been much more cost effective getting to grips with the issue up front ( in the initial production years )
But someone with deciding powers thought they could get away with laxness ( guess what whopping number of shares they could buy back instead of spending some money on process control! )
Hindsight is perfect but inaccessible to the MBA troupe 🙂
Thanks! Just looking to see if my thoughts on the subject made sense in the outside world!
I’d really like to know more about the specific fuselage joints in question though. Are composite fuselage sections bolted through the composite material or do they have embedded metal (rings?) to provide for the bolt-through joints? I have some comprehension of stress distributions in a bolted joint in metal but none if the parent material is in fact composite. In any event, the presence of debris from drilling bolt holes (is this in fact how the bolt holes are created? through composite material? at the time of assembly?) can’t be good and might be disastrous in terms of fatigue life OF THE BOLTED JOINT.
@Mike: “Are composite fuselage sections bolted through the composite material or do they have embedded metal (rings?) to provide for the bolt-through joints?”
In most cases the composite is simply a drilled hole with a lock-bolt, or larger tension bolts where required. Depending on application, the head and nut-washer in skin panels and other thin parts may be wider than in aluminium to avoid “pull-through”. Sleeves in the holes have been considered but as far as I know are not normally used. Expansion sleeves/collars and oversized fasteners are not used since these damage the composite. Glass washers were considered for electrical insulation purposes but I don’t recall if these are applied in the end.
“the presence of debris from drilling bolt holes (is this in fact how the bolt holes are created? through composite material? at the time of assembly?) can’t be good”
Yep, just drilled through. I know they use different drill bits and I expect there are other differences to mitigate delamination and ensure a clean cut through the fibres, but otherwise it’s no different.
Thanks Someone! Are the joints in composite fuselage sections more like slip fit pipe with hub connections (radial bolts) or more like flanged connections (longitudinal bolts)?
Also, it sounds like the discussion is conflating “gaps” in aluminum fuselage sections and “gaps” in composite sections. Not only are the materials different but the structures, and stresses, are likely to be considerably different as well. Is Boeing having problems with both?
@Mike – trying to visualise what you meant I think it’s what you described as slip fit – the vast majority (especially in the wing and fuselage skin panels) are “single-lap joints”, where the edge of one skin panel is laid over the edge of the next one and fasteners are bolted through holes drilled across the two panels.
The illustration in the article visualises a circumfurential joint like you’d have between barrels. It also highlights my concern about the fasteners – the joint is designed for them to take load running along the skin from one panel to the next (at 90 degrees to the bolt shaft). Closing the gaps will add fit-up loads at 0 degrees to the fastener shaft, therefore unplanned stress on the head and nut.
The gaps would be the same for aluminium or composite. This isn’t about local differences between imperfect surfaces (I disagree with Uwe here) – we’re talking about two huge pipes which are supposed to have “exactly the same circumference” – but of course manufacturing at that scale will have inevitable tolerance problems leading to an overall gap, an inner which is too big for the outer or bits which don’t line up because neither side is a perfect circle.
@Someoneintoluse: Then a common hardware analogy would be fitting stove pipe? Male end “crimped” to make its effective diameter small enough to slip in an un-crimped equal diameter? Then shoot some self-tappers through to hold them together? Or an exhaust pipe? Goop the small end up with sealer, slip it into the big end, and apply a muffler clamp (instead of screws)?
How deep is the insertion of the small fuselage section into the larger (length of the “lap” joint)? How thick are the composite walls? How many fasteners to go all the way around the barrel? (or the “pitch” distance between adjacent fasteners) Are there multiple rows of fasteners? Is there goop like the muffler sealer? Or is that where the shims are supposed to be slipped in?
(Maybe I should un-retire and work at Boeing Everett right down the road instead of bugging the nice folks here with my unbearable curiosity…)
@Mike: ha ha! Crimping would be a VERY extreme analogy of the problem being discussed in the first place. You *definitely* don’t fly anything that gets that bad…
Also no self-tapping screws here – it’s all precision drilling/reaming for flush-fitting countersunk heads on the outside with washers and collars for small joints (nuts for barrels) on the inside.
By the way, I should have been a bit clearer about something – the skin needs a smooth surface transition from panel to panel for aero and stress reasons – so it’s not so much one tube of skin fitting inside another as two tubes being joined by fitting over an internal strap and structural item such as a circumferential frame. In the case of a barrel, I would expect one side to be prepared with these parts pre-assembled as the “male” side, while the next barrel would have those parts inserted during the join – so *effectively* you’re joining tubes with one side larger than the other.
Sealant “goop” and shims are applied during these joins – which is the subject of the article.
Depending on the beefiness of the joint, the skin thickness is in the order of a few millimetres. The overlap maybe 10-20 centimetres. A minimum of two or three fastener rows each side of the join (probably more for barrels) and fastener pitch a couple of centimetres both between rows and around the circumference. There are a *lot* of fasteners in a barrel joint…
@SomeoneinToluse Thanks! So there is a ring to which both tubes attach from either side (to reinforce a butt joint)? And the ring, it is metal (aluminum?, titanium?) Could the ring (in theory) be attached to either tube first?
(In VSM (variation simulation modeling) one of the hard things is convincing designers that parts don’t simply un-explode to each other. They assemble in a sequence chosen by someone, and it’s best if the sequence is 1)consistent and 2)chosen by engineers for specific reasons 3)not just left to a “Hey, let’s put the ring in this end today” kind of process.)
So given 3 rows (per side) of fasteners (6 rows per joint), the rings are 32+ cm wide minimum? (More to set the first row a bit further back from the edge of the composite?) Is there a torque or a tension spec on the (rivets, hucks, bolts) fasteners? (I’m reaching for a visualization of the stressed volume under each fastener head and do said volumes overlap?)
And good grief! Simulating the loss of integrity of composite parts when a machining process cuts, trims, or drills them, pulling fiber away from the matrix. Special drill bits indeed! Are the holes drilled through both composite skin and metal ring with the same drill bit? Speeds and feeds? Wow! Incredible work!
@Mike: you’re asking specifics about the 787 which, as someone who doesn’t work for Boeing, I can only speculate.
A butt joint is what I was describing, although they could also have integrated it as a thickness increase and joggle at the male side. To be honest, I don’t know for sure. As for your other speculation… well it all sounds about right to me – but then I’m in the same boat as you! 🙂
To get an idea of what it looks like, at least:
https://c8.alamy.com/comp/PAJDK6/a-boeing-787-centre-fuselage-one-piece-barrel-in-production-line-at-the-alenia-grottaglie-factory-PAJDK6.jpg
https://image2.slideserve.com/3991637/completed-test-barrel-l.jpg
https://simanaitissays.files.wordpress.com/2012/12/boeing787fuselage.jpg?w=1024
Great pictures! Thanks!!
regardins the improper conflation of 787 composite fuselage join issues (shims) and 777 aluminum fuselage join issues ( shims ) and tolerances.
The ONLY commonality between 787 and 777 re the fuselage is that they are similar shapes and join in similar places ( lenghts ) by similar methods ( fasteners ). They do not fatigue or ‘ corrode ‘ in the same way. Stress ‘ cracks ‘ in aluminum typically proceed to the next hole and usually stop. Laminates transfer loads much different and as Stan S said issueis usually delamination.(failure of the ‘ epoxy’ glue the fibeers are embedded in )
So the fatigue and service time issues already shown by boeing on the 787 should be sufficient to answer the questions for 787. But 777 is a much differrent story both in assembly forces required, and effects of fatigue and corrosion during service life. AFIK assembly processes of early 777 body join were later changed- dates and units numbers notg availabe to this old injun ear who worked on both 777, 767, ( aluminum ) and B2 ( composite ) programs.
Boeing IS still bloviating re its records, accountabily, and trying to improve its posterior covers. Had they NOT stonewalled the whistleblower early on simply to make cost and schedule- there would be little to discuss.
Just my .00000001 worth.
Nothing boring would’ve said would assuage your concern. Let’s just be honest here
The issue raised was indeed fatigue. The whistleblower claimed a 787 could break apart in midair due to fatigue. So far there’s no proof
Everything is unprecedented until the first time that it happens…
So no progress re fatigue and corrosion on fuselage sections and design has been made ???????
Aloha Airlines Flight 243 (IATA: AQ243, ICAO: AAH243) was a scheduled Aloha Airlines flight between Hilo and Honolulu in Hawaii. On April 28, 1988, a Boeing 737-297 serving the flight suffered extensive damage after an explosive decompression in flight, caused by part of the fuselage breaking due to poor maintenance and metal fatigue….
Note the plane stayed together despite major loss of fuselage skin, has to do with circumferential and longitudial stringers taking basic loads. Are we to believe no such parts or design considerations exist in 787 OR 777.
I know they are there in 777- but do NOT know if similar in 787.
Since the engineering requirement has remained at .005 rather than the observed value of .008, this would indicate that the .008 cannot be substantiated by conventional stress analysis. So they went to plan B, which is testing, inspecting, monitoring.
I’m not sure what to make of that. As has been noted, durability of composites is a whole different animal from ductile metals in which tiny flaws slowly grow to detectable cracks. Do they have robust means to detect fatigue damage in composites before it reaches criticality.
“Do they have robust means to detect fatigue damage in composites before it reaches criticality.”
Using a “waterjet “and ultrasonic inspection of composites to detect flaws- bonding isssues in composites has been around for at least 40 years .
Sort of like ultrasound on your body.
Can you “ring” composite sections like a bell (or a dinner plate or glass?) Assuming it possible to avoid damping by contact with supporting structures.
Or kind of like sonar but in solid material? Flaws create weird echoes?
think echo sounder.
you inject ultrasound “pulse” into the accessible surface.
discontinuities ( lamination issues , interface to air, … ) produce echoes. Time to echo is the depth of the discontinuity under the surface.
different method: you can “ring the bell” and analyse the sound produced. ( frequency, damping ) rings “right” 🙂
If it’s outside the drawing requirements then a concession should record the deviation. This concession has to be signed by design & stress normally before the aircraft can be delivered.
This seems like a perfect opportunity for pushing the schedule to override proper documenting of a deviation. “Looks like 0.004999 to me… Send it.”
0.004999 is good, >0.005 is bad.
I always wonder where specs like that come from. In my experience many are arbitrary. Sometimes though analysis or experimentation has in fact been done to show units (whatever they be) exceeding the selected value rather quickly approach some danger zone. Sometimes it is a two-tailed analysis: too much to one side (say big gaps) and failure becomes more likely, too much to the other side (tiny gaps) and (scrap and rework) costs increase. The increase in probability of failure and the increase in costs are often steep curves resulting in some sort of “sweet spot” to target and establish limits around.
After this, and I wouldn’t trust a congressperson to understand it, we still have the question of there being a means to mitigate the probability of failure curve by some special means (adding shims, or filling gaps with adhesive or putty). It’s possible that gaps considerably out of spec could be “fixed” by special means. Has Boeing ever done such studies? Would congress understand them?
Then there’s the obvious: It is going to be far cheaper, and faster, to omit the special means claiming it is an unnecessary restriction on build rate. At the same time, the presence of the special means procedure may in fact lead to worse performance to the initial spec. (Why rework or repair? They can shim or putty it into spec.)
Again, we can cross our fingers for free.
I think we ought to stand back and have logical discussions about this, not driven by fear or emotion. It would be good if Boeing could say something like “we really appreciate Sam’s thoughts about this because it is important to bring up concerns. Our plan is to work with him to assure our processes are safe and robust, make changes where necessary, and describe to him what work has already been done to address those issues.”
As a former PR person, I applaud the creativity in coming up with “the gap management journey”, but not when “we have not closed all of those items” is used in the next sentence.
Also can we note this whistleblower is not a stress engineer or analyst and has no experience in stress testing.
We can.
But we can also note that Boeing hasn’t provided us with sufficient detail of the stress modeling that it performed in this out-of-spec situation. Until such time as it does, we can conclude that jumping on precision composite panels to force them to fit is generally not conducive to their subsequent structural performance.
There are some of us that know when those in another field of expertise should be consulted. Whistleblowers generally fall into this category. Bean counters often reject the observations of the whistleblower rather than consulting the expert. Especially if they no longer have as many experts on staff to call on. The cost of a consultant might be too much to bear.
quite often a whistleblower is the expert that was ignored “in House” 🙂
Also, I’m curious about the Steve Chisholm comments. In the Everett factory in the tour balcony of the tower between the 40-21 and 40-22 bays there is or was a cross section from section 42 of a JAL 747 that the company bought back and did an extensive fatigue test series on to proactively discover any unknown issues that might require and adjustment to design, assembly or maintenance processes. It was a thorough destructive testing exercise. Is he claiming to have done that with a 787 that has been in service for an extended period?
They have from 2011 to 2015. There are pics of the frame
747 is aluminum- 787 is composite. JAL 747 crashed- which 787 has been taken out of service to allow the same invasive test and exam ?
Independent of different materials the issue is
a persisting one in Boeing’s production system.
The 777 AL sections need excessive force for joining
and the CFRP 787 sections also have (excessive) tolerance based issues.
i.e. Boeing has been and still is unable or unwilling to sufficiently upgrade the production ( system ).
This is in stark contrast to the projected image as “best of breed” company in civil aircraft design and production.
Additionally and IMHO we have to revisit the historic Boeing image. Were they really in the lead or like Microsoft the perfect utilizer of the existing combined with perfect public presentation.
competitiveness methods used against Airbus are very similar to how Microsoft tried to hold down OS competitors.
( promising “super” features for the upcoming product “right next year …” )
uwe said ” The 777 AL sections need excessive force for joining”- excessive ??? source ??? design forces ?? how much excessive ?? 1 .3 times deasign or allowable ??
My point is where do you get the number- definition of ‘excessive’ compared to allowable ??
Do a bit of research on your own.
You will find some notes on 777 as well as 787 join issues.
I would be inclined to believe Boeing if it were not for their appalling behaviour after the second MCAS crash. They tried to call in a favour from the president and would have sold their own grannies in order to avoid the MAX being grounded.
How far would they go to avoid 1000 787s being grounded?
on the other side of the pond
Flight Global
Comac assembles expert group for C929 research meetings
“At the meeting, six key themes were discussed: aerodynamics, aerostructures, composite structures, avionics, mechanical systems, and propulsion.”
“According to Comac, the meeting brought together more than 280 academics from 21 universities in China, along with experts from seven local aviation firms and 43 other state-owned firms.”
Great news for Boeing,Rolls Royce and GKN!China is wasting all its money employing experts and skilled personnel,they’ll never make a profit like that
+1
Step by step..
For every ARJ21, C919 and in the future C929 delivered is one less aircraft Boeing will get Remind me again, when was the last time China ordered from Boeing? 2017?
Have you noticed from production lists that the mean time between successive C919 deliveries is steadily decreasing?
It won’t be long now before the rate hits one per month…then two…then three…
One wonders at what juncture the White House will stop LEAP-1C export licences…
+1
AW
Comac Expected To Deliver At Least Seven C919s In 2024
“SINGAPORE—China’s Comac is set to produce at least seven C919 airliners in 2024, according to aircraft delivery timelines laid out by local carriers China Eastern Airlines and Suparna Airlines. This would mark a significant ramp up by the OEM, which produced four aircraft in 2023”
As for your LEAP comment…it could happen…if so Boeing will not get a new order for a long long time….
Remind me again, how many Chinese 737 and 787 are in Boeing storage “yard”
Despite a Leeham opinion article from 2022 suggesting that “China needs Boeing” (see below), I’m of the opinion that the Chinese are decoupling from the US more rapidly, extensively and successfully than vice versa. I’d be surprised if China ever orders another Boeing aircraft again. And if the LEAP-1C gets banned, that will just serve to accelerate further refinement of the Chinese CJ-1000A engine.
https://leehamnews.com/2022/10/31/china-still-needs-boeing-as-much-as-boeing-needs-china-despite-interminable-limbo/
Biden calls for higher tariffs on Chinese steel
but “China was the seventh-largest exporter of steel to the U.S. in 2023, with shipments of 598,000 net tons, down 8.2% from 2022”
“Although China is a minor contributor of imported steel to the U.S. directly, its ability to furnish finished steel to some markets and semi finished steel to rolling mills elsewhere can affect the demand for both American-made steel and ferrous scrap shipped from the U.S”
Is China steel imports to the US a real problem or is US steel industry obsolete because of the lack of investment and innovation!
Lots of good comments here.
Some have questioned the validity of the conclusions of the 787 whistleblower based on his observations & bona fide. Some have questioned Boeing’s response for roughly the same reasons.
I want to address a concurrent issue with another whistleblower at Wednesday’s hearing.
The witness was Ed Pierson, the executive director of The Foundation for Aviation Safety.
What is The Foundation for Aviation Safety?
Who works there? What is their work product?
Who funds it?
Mr Pierson made some pretty dramatic assertions. Does his organization have the bona fides to warrant tracking them down?
Some info on the Foundation:
https://www.foundationforaviationsafety.org/podcast
“Boeing hit with 32 whistleblower claims, as dead worker’s case reviewed”
Boeing has been the subject of 32 whistleblower complaints with the workplace safety regulator in the United States during the past three years, newly obtained documents reveal, amid mounting scrutiny of standards at the beleaguered aircraft maker.
“The figures shed light on the extent of alleged retaliation by Boeing against whistleblowers as the Virginia-based company is facing mounting questions over its safety record and standards.”
https://www.aljazeera.com/economy/2024/4/19/boeing-subject-of-32-whistleblower-complaints-documents-reveal
One of the things I’m confused by is, why aren’t they using air pressure lubrication of something that’s more like an interference fit, rather than the static “perfect” (0.005″) fit they’re currently using? Are the fastener holes pre-drilled?
If the joint was such that there was an initial mating ramp, one could then use internal air pressure (handily, they’re making a pressure vessel) to force a gap between the outer and inner components of the joint. That would apply an even and fixed amount of force between and – more importantly – an even amount of deformation of the two components, and you’re confident of how much deformation one is applying to them. It also provides an air cushion to keep them separate whilst you push the joint closed, even if the goal is an interference fit, and it would also blow dirt out of the joint at the same time.
Once the joint is closed and the air pressure removed, the slightly too small outer would pull tight to the slightly too large inner, probably with very even forces all around the joint. If one makes sure that the residual deformation leaves plenty of elastic margin, you have a durable joint.
The amount of airflow needed to ensure zero friction whilst assembling the joint would tell you an awful lot about the quality of the fit. An experienced worker able to feel where there’s more airflow leaking out than there should be, indicating that the joint needs to be pulled back apart and a shim put in. Similarly, a lack of airflow indicates that things are a bit too tight. But if the joint goes together smoothly, with the right amount of airflow overall and felt all around, you’d have a lot of confidence in it.
May be I should patent that idea, if no one else has had it.
If you pressurize the fuselage both segments expand.
Additionally pressure moves the pieces apart. ( force is over pressure times cross section area, nothing to sneer at )
The trick is to produce to desired tolerances and not applying some trick to get around ones “pfusch”.
Hmmm, well I think one could engineer the joint so that an air cushion did form. If the outer part were ever so slightly more flexible than the inner part, that’d do it.
Of course the two parts would have to be kept in place, but I don’t see why that’d be a problem.
With CF, the goal is to achieve a tight fit, with minimum distortion. The way they’re trying to do that at the moment is with shims, but I think that it’s probably better to play the material’s properties; have the outer part in slight (and well controlled) tension around the inner part, fitted using air pressure to stretch the outer over the inner for assembly.
The use of shims would then be subtly different. Instead of using them to achieve a specific gap fit, they’d be being used to achieve a specific tension.
You can see in the first image in the article that they’re applying fit-up forces to judge the shim thicknesses required, and then the fastener forces do the rest. What they actually talking about is squashing the outer around the inner, and keeping the localised circumferential compressive forces (which is really bad for CF) the fasteners apply to a minimum.
I’m simply suggesting eliminating them altogether, replacing them with circumferential tension forces that transfer the compressive forces to the inner part, only much more evenly spread (thanks to the air cushion). The other advantage is that the inner part can actually be structured more appropriately to take those compressive forces (far more so than they can the outer part), which would also mean that the inner part expands less with the assembly air-pressure than the outer, referring back to your first sentence.
Confusion cleared up. I asked someone that knows.
What I outlined above won’t work, with CF. The issue is that CF does not deform elastically as such; the fibres will break before they stretch. So, one could not use air pressure to stretch an outer half of the join over the inner half of the join.
Curses – I though I’d had a good idea.
However, it does illustrate what Boeing are up against, trying to assemble such large pieces with sufficient precision that one can trust the joint for, what, 30 years. In terms of engineering with composites, the concept is definitely way off down at the “risky” end of the spectrum. I suspect Airbus in choosing a less ambitious size of CF parts to assemble bought itself some production line insurance; it’s perhaps more work to assemble, but it is easier to ensure that it is assembled correctly.
There’s also an interesting comparison between what both Airbus and Boeing are doing (assembling separate pieces into a completed structure), and what McLaren are doing with their car chassis. They’re made in a single operation, using a process that they’ve guarded quite carefully, apparently something to do with “inflation”. They not having to assemble individual pieces, they lay up the whole chassis (which is hollow) and then cure it, and it comes out of the autoclave complete. I wonder if this process could ever be scaled up to something the size of an airliner. The result for McLaren is that, instead of taking 4,000 hours to lay up a McLaren F1, it takes 4 hours to do their current chassis. That’s quite a step improvement.
“I suspect Airbus in choosing a less ambitious size of CF parts to assemble bought itself some production line insurance; it’s perhaps more work to assemble, but it is easier to ensure that it is assembled correctly.”
You are so gracious 🙂
Airbus position in that domain (which I share) is that one piece barrel sections will always lag in precision over the panel solution.
Also note that inserting the frames and further structure into a cylindrical section is a demanding task.
the “cylindrical barrels” are a “duh, simple …” solution that is more or less unsuitable. ( note that Boeing hasn’t got to grips with the problem in 2 decades of production. I don’t get this infatuation with excessive shimming either. )
It works rather well for conical sections though.
The Airbus panels are similar size to the Boeing barrels in terms of amount of skin + support structure – just a different shape.
The McLaren process doesn’t actually sound all that innovative to me – aircraft parts are usually made from skin sections of varying layups and thicknesses mated to stiffeners, frame sections etc. then vacuum-bagged and baked in one shot. They might have some novel tooling to position everything before entering the autoclave, I suppose.
This is interesting:
“A former employee of Boeing claims to have handed documentation to the FBI detailing the work done on the panel of Alaska Airlines’ 737 MAX 9. The American manufacturer had declared the non-existence of these documents.”
“The statements of Ed Pierson have been taken by a Senate subcommittee last Wednesday. Pierson alleges having handed federal investigators documentation confirming the work done by Boeing on the plane involved in a serious incident last January 5th.”
“Pierson stated: “There are records detailing the hectic work done on the Alaska Airlines plane, and Boeing’s corporate leaders are aware of it. I know this documentation of the Alaska plane exists, because I personally handed it over to the FBI.””
““In my opinion, this is a criminal cover-up,” he added during the hearing, alluding to the fact that an internal whistleblower at Boeing presented the documents and have been in the possession of the National Transportation Safety Board (NTSB) for months.”
https://www.aviacionline.com/2024/04/former-boeing-employee-confirms-existence-of-alaska-airlines-737-max-9-work-records/
Now that would be a massively big deal, if it’s true. TFTL.
Adding: I have a sinking feeling that nothing at all will come of this, though. Hope I’m wrong.
interesting article
Boeing and the Dark Age of American Manufacturing
Somewhere along the line, the plane maker lost interest in making its own planes. Can it rediscover its engineering soul?
https://www.theatlantic.com/ideas/archive/2024/04/boeing-corporate-america-manufacturing/678137/
CNN just ran the 2 part series again re the Shuttle Columbia disaster 2003-2004
The second section comments by various executives as to why /how the piece of foam that came off during launch caused the damage to left wing which resulted in ‘ burn thru ‘ and the posterior covering and the NASA culture of NOT reporting concerns sounded more like a current ‘splanation’ of Boeing quality and safety issues – deja view all over again or historic preview.
Pushing cost, schedule, posterior covers over safety seems to be a common thread
I have a theory that these things come about
via lawyer perceptions.
From an engineers perspective laws of physics ( as in observed behaviour ) are hard as a punch in the nose.
From a lawyers view ( coming from a field of “laws are written down consensus” ) such laws are things you can talk your way around. Just use more contrived words. Unpleasant and negative synergies derive from linking to the MBA hordes.
An understanding of probability helps too. I just flipped 37 heads in a row. What’s the probability the next toss comes up heads?
MIS-understanding of probability …
in physics “37 heads” in a row does not produce precedent 🙂
in the lawyer and MBA sphere it does though.
I thought about that. I couldn’t figure out if a lawyer (or Congressperson) would be more likely to favor heads (Because it’s always been heads) or tails (Because it’s been heads for so long it has to come up tails) (;-)
The odds are the same as they were before the 37 heads-up flips.
But does a congressperson realize that?
Is anyone surprised that Calhoun did not make it to Washington to testify at the whistle-blower hearing?
Truth is, anything he would say is “non value-added”, means nothing simply because he knows nothing. The posters here have a better idea of what’s actually happening on the shop floor and in the engineering groups than he does. Totally out of touch.
It becomes unfortunate that “leadership” is often the people who are designated to become the talking heads. When they don’t show up there should be some mechanism for an engineer to take their place.
Someday maybe Congress will get sick of being told “We’re taking care of that.”