Leeham News and Analysis
There's more to real news than a news release.
Leeham News and Analysis
- Boeing unlikely to meet FAA’s 90-day deadline for new safety program April 18, 2024
- Focus on quality not slowing innovation, says GKN April 18, 2024
- Boeing defends 787, 777 against whistleblower charges April 17, 2024
- Dissecting Boeing CEO’s statement next new airplane will cost $50bn April 15, 2024
- Bjorn’s Corner: New engine development. Part 3. Propulsive efficiency April 12, 2024
Bjorn’s Corner: Communicate or not?
By Bjorn Fehrm
19 February 2016, ©. Leeham Co: Mitsubishi Aircraft Corporation (MAC) announced on Christmas Day that they delayed the entry into service of the MRJ90 regional airliner by over a year. At the same time, they also announced that they had to reinforce the aircraft’s wing and fuselage.
The market’s reaction to the news was with disappointment. A further delay to a new aircraft from a new player in the market was not good news, but it was not that surprising. Bringing new aircraft to market on time is tough for the established players. Other new entrants, COMAC/AVIC and Irkut, are also running late with their programs, and Sukhoi was late with its SSJ100.
What worried many more was that the aircraft needed reinforcements, directly after its first flights. After only three flights, the aircraft was grounded and was scheduled for changes to its airframe. That was really bad news. “It’s going to be heavier.” How could MAC miss to gravely was the common reaction.
Having worked in a 50% Japanese company for many years, I wondered what was behind this all. Was the aircraft really in dire straits or did we witeness a cultural mismatch I’ve seen many times?
What to say and how to say it
To find out, I set up a meeting end January with MAC’s Marketing Director, Hideyuki Kamiya, when we both visited a conference in Dublin. We talked about the strong reaction from the market and I asked him to explain what was behind the structural reinforcements.
Kamiya showed me a drawing of the join between outer wing and the center wingbox. There were small L shaped reinforcements to be riveted to the stiffening flanges of the center wingbox ribs, Figure 1. There were also similar profiles to be added to the stub frames that went from the center wingbox to the side of the fuselage (the Americans call this the side-of-body join, which is totally misleading. The join is between the outer wingbox and the center one and has little to do with the body (i.e. fuselage)).
Figure 1. Slide showing reinforcements of MRJ after static testing. Source: Mitsubishi Aircraft Corporation.
Kamiya told me about 5kg of material was added. For serial production, it would be less, as MAC only needs to mill away less material for the flanges than needed reinforcement. These changes had been detected in static testing and were needed to withstand ultimate load (150% of the highest load the aircraft will ever be subjected to).
Now I knew this was all about national cultures. The reinforcement was nothing and I would not have told a soul about it. Such minor modifications happen all the time during aircraft development programs and flight testing. In this case, it came out of structural testing and was a non-issue during flight testing. No big deal.
Later that day I talked to a senior aircraft OEM person about the background to MAC’s announcement of structural problems. He said, “we would not even have mentioned it, not when it’s such a small change.” Given the industry practice to not worry the market “with details,” this was clearly MAC being too honest and up-front.
Customers got really anxious, and no wonder. When OEMs communicate they have such problems, what they reveal is often the tip of an iceberg. Here it was the inverse. The Japanese culture of being totally honest was not in tune with what was the industry practice.
What is the right thing to do?
The incident is an interesting one as it highlights the communication culture that should prevail in the aircraft industry. Should one also reveal smaller problems or should the market be spared such details?
I would say one should communicate things that impact the customers. The delay in flight testing and its more thorough execution going forward makes the aircraft come to market later and it should be told.
But that one needs to mill away 5kg less material on some ribs and stub spars? On a 25 tonne aircraft? Questionable, such details are not really important to the customers and can be kept inside the project.
As an enthusiast, I like the transparency. One positive, in my estimation, is that the openness can slow the rumor-mill. I believe Airbus would have had a lot more rumors and harmful speculation on the A320neo if they had not communicated the issues with the engines: combustor-seal thermal wear, and shaft bending. And since a new entrant likely is particularly vulnerable to rumors, the culture of Mitsubishi may potentially be beneficial down the road.
Yep, the irony when companies try to cover up problems is that the cover up becomes a bigger problem than the original problem ever would have been.
And a related question: to fix before first flight or after. I think the former, in which the programme could hopefully then run and be seen to run more smoothly, is better for reputation building. But what about the engineering side? As far as the engineering goes, do you think they made the right decision to fly and then ground or would you have gone with fix first Bjorn?
I don’t think they knew in time to fix it before first flight, it only becomes obvious when you are approaching Ultimate load in the static test and that is late. For flight test you only have to prove Limit load with a margin (100% of max load in operational use). You approach Ultimate load very carefully as once something start deforming your test is over and you want to understand everything that happens up until then.
Once you align the organization to go for first test flight you keep it unless things deform/break close to limit load in the static test aircraft. Also, with first test flights done you could correct things you saw there in the downtime, if there was anything major. I understand the flight test sorties went as expected.
Unfortunately other Japanese companies, Takata and Toyota, had no problem hiding, lying and denying a quite deadly defects for years.
Mitsubishi Aircraft Corporation should be lauded for their refreshing honesty in this day of high profile cover ups such as Takata, Toyota or deliberate malfeasance, Volkswagen (at least no one died). from countries that most people expect higher ethical behavior from.
I hope the MRJ90 makes it.
Hello Bjorn
If I recall recent japanese programs
P-x, C-2, US-2 all delayed at some point, after first flight, with structural problems as an explanation.
Maybe also some internal cultural problems ?
Bonne journée
My mum always said that it’s better to just fess up, the lies just get bigger and everyone finds out in the end. Honour is very important to the Japanese,just look at the mess Olympus got into.
The error that MAC made was the initial statement of a long delay with very vague explanations. The truth is unlikely to be worse than people’s imagination in such situations.
Thanks Björn !
I’m just wondering why such a large new delay, one year …
For a not so dramatic flaw, detected before first flight, well understanded and promptly corrected !
Another Japanese cultural particularity ??
Or may be some additional unveiled problems undercover ?
The delay of over another year was to very little extent due to the strengthening of the first test aircraft, less than three months as the others get strengthened before they fly. The major cause of the delay was a change in test flight procedures. MAC had dimensioned for a classical test flight program. The people that came in to the US engineering and test center told MAC, this is no longer the way you do it. You don’t ready the aircraft and fly, then evaluate. You jack up the aircraft (to make it think it is flying) and fly the complete test on the ground first, then you fix the faults you see, ground fly again and finally you verify your test aim in the air. Then you evaluate and correct and then the whole thing starts again.
This takes longer time but you get a much better result and a better integrated aircraft. You make sure the first customers are not part of the debugging of the aircraft to the same extent. This more time consuming flight test procedure was fully accepted and the re-planning showed it would take another year.
You can read too much into cultural stereotypes, but to the extent they exist I would say Japanese culture is quick to own up to problems but slow to confront them. It can mean that systemic issues aren’t addressed. I have no idea whether any of this applies to Mitsubishi, or whether the MRJ90 is more a 787/ARJ21 cluster-muckup or an A350, that is delayed, as all complex airplane development programs are, but the issues are manageable and managed. I guess we won’t know for another year or so, but I can see it is frustrating if you are in charge of communications.
Hmm. To tell or not to tell? Well, there is another company in aerospace that seems to have lost the art of communication. Everyone is watching the PW1000G program like a hawk with the result that the effect of a problem (eg. uncontained engine failure, rotor tip contact or cockpit message error reporting) is more talked about than the cause. Weirdly, the cause can be less dramatic than the effect and yet secrecy prevails. Does anyone know what hardware is being changed in the PW1100G-JM before Qatar accepts its aircraft? Rumour can be more powerful than it needs to be. Transparency CAN be your freind…
Is the wingbox part of the body? Is the wingbox pressurized? Is the pressurized compartment the body?
In my world it is not. I think the side of body expression is very old and comes from the days when you thought you mated the wing halves to the body. Today you mate the wing pieces to the center wingbox to form a complete wingbox, which in turn interfaces the body.
Makes sense to me. From a static structural point of view, the wingbox is the center piece. Two wings with upward shear force, and two body halves, (one with a tailplane) with downward shear force.
Bjorn: thank you for clearing that up. I never could wrap my mind around it when presented. It made no sense.
Does that match the idea that the centre box is integral to the fuselage and built together , ie centre barrel and wing box, the wings of course are added at the end.
The fuselage centre barrel- wing box seems to be fairly new as the 747 was assembled first by joining wings to centre box and then to fuselage ( may have changed ono 747-8?) and the DC-8 seems to have had its wings ( 3 spar)made in two complete halves and joined on centre line and fuselage added on top.
But the key is that the wings (the outer wing halves) attach to the center wingbox, not the fuselage sides, which the expression side-of-body-join leads one to believe.
I am fully on board with Bjorn on the notification.
I think Mitsubishi over spoke and by a lot.
The big delay per Bjorn is the testing procedure and that’s the one that should have been conveyed.
In our day in day out work, we all run into problem and sometimes mistake we make or made. If it has not impact other than it needs to be corrected and is corrected without any impact, then there is no need to notify anyone, its part of the work life.
Keep in mind A380 had a similar issue, it was only public because the wing failed at less than 150% (147%)
It failed where they thought it would, just sooner
All it took was a bit of reinforcement (which was probably just thickened a bit in the subsequent production)
they did not have to run the tests again as the authorities were satisfied with the modeling and the modeling of what the fix would do.
A bit unusual but 787 had that issue as well. Probably due to the refinements of hitting the goal with the minimum structure, models will miss it from time to time.
By miss I mean not spot on, close, which means the model is very good, just not perfect and that is why we do tests.
Keeping in mind, sometimes a fix just causes the stress to go someplace else and break so the modeling has to be up to snuff.
I have no reason to think it was not
The A380 production wing was already beefed up for heavier loads than that of the very early test wing.
The 787 had much more than a ‘couple of kilos of reinforcement’ ( or production structure milled slightly less) as they were 10 tonnes over weight
I don’t quite get it?
If you are comparing the two they have nothing in common.
Boeing over build except the wing box area, removing weight took that down below specs
That said, they overbuilt the rest and have taken weight out to the point its now below promised weight.
Observers learn the company, Scott and Bjorn are translating Boeing’s and Airbus’s statements from what they say to what they mean every week, and if one says something it might not mean the same as if the other said the same thing. Mitsubishi as a new company is still an unknown language. If everybody knew it there would have been less reaction.
Now that you guys have explained it, the wing problem doesn’t sound too bad at all. Maybe they should have explained it a bit better. On the other hand now that you’ve explained the certification issue it seems to be a bit of a blunder. Perhaps understandable but the inescapable conclusion has to be that they didn’t really know what they were doing and should of brought in outside help earlier.
Given the time this, (how many years now since launch?) as well as other new, programs have taken to get from conception to service it could also be a case of changing goalpost in the certification process. Established OEMs are awake to this (Airbus’s sudden change from Li-I to Ni-C batteries is an example) but a new entrant isn’t so awake.
Yes they came right out and told everybody about the issue – on Christmas Day!?! and it was vague. I would not take the explanation of the marketing director.
Christmas day is an ordinary working day in Japan. Their winter holiday is New Year’s day.
It’s also very easy for the import of a statement to get lost in translation!
Well, they were open but should have given details.
The need to reinforce is a risk of not doing static tests before flight. While they have experience such as 787 it may have been wise to time the program differently.
(Of course Boeing is an example of how to really muff development, the 787.
Failures do occur, the 767 aft fuselage failed in the static test, I don’t recall the impact of the fix nor even if the airplane had yet flown. It was an interesting test event, watching from a balcony IIRC, everyone focused on the wings then a bang we couldn’t see the source of.)