A380 Wing Comes Off
 

This has just popped up on ATI:

"The wing of the Airbus A380 static test specimen has suffered a structural failure below the ultimate load target during trials in Toulouse, but the European manufacturer is confident it will not need to modify production aircraft."

Isn't that title a BIT misleading...?
More details here
...

Just like the vertical fin separating from an American Airlines Airbus departing from JFK in November 2001, the French will rule that the test bench operator was to rough on the equipment, like the pilot of the airbus that crashed.

Long live the manufacturer of disposable aircraft!

How would Mr Boeing play this though?

Let's hope that disposable passengers don't become in vogue!

I seem to remember the C17 wing broke early too. Sometimes it happens, usually because the test loading of the wing is not adequately representative of the in-service loading.

The MD-90 split in two during testing hard landings too.

The bottom line is that the majority of problems are found during testing rather than during passenger operaations.

Hmmm.

The rig "proves" the Finite Element Model (with some error), so redesign of the wing based on the FEM is OK.

Hmmm.

:hmm:

It's called designed obcelesence: the basic rule of American car design for at least thirty years :}

*j*

Is it just me, or has anyone else noticed how often posts are made by individuals how are willing for the A380, in fact Airbus in general to fail.

I fly the 737 and work for an all-Boeing airline. I can see their strong points and their weak points, and I'm quite happy to comment on them, however without any experience of Airbus products, save a few times paxing on them, I don't really feel qualified to comment on their merit.

I wish others would do likewise.

PS I hope the A380 is a big technical and commercial success. Success improves the breed in general.

One would HOPE so!!!!

I am just curious how much time Airbus marketing has spent before they came out with this.:}

As for EASA (and FAA consequently) it would be interesting what change is Airbus proposing in order to justify this.:confused:

Catch is that usually local CAA allows some test flights based on limit loads only and then the rest of the stuff based on the ultimate loads (1.5 x limit loads).

If you go to break the wing that means all the other test went smoothly. No it is unclear if they have finished ultimate loads tests or no. Also there is no word for what loadcase did the structure failed.

Still this is embarrassing and the only thing they could say is that "FEM was correct". Too bad.

Stating they will "use calibration of FEM" is, ahm, questionable. One can always do this with bad results so...

As someone said once in the end Airbus=EASA, Boeing=FAA and thus the way things will resolve themselves.

Better to find it now, if true, than too late like the Comet.

That is what testing is all about. I think it is what is called 'no news'.

I find your remark quite facetious and insulting to all involved in that tragedy..

Well, If you can't meet the standard, change the standard. Problem solved.

Testing is there to prove analysis, any data whatsoever that came out of this test is useful. They can only test for a very limited amount of loadcases and represent them as best they can. The tests are not completely accurate.
The results will be used to parallel the more comprehensive analysis and if they correlate the assumption that the rest of the analysis is good will be made.
Every test I have been involved in has had the conclusion of failure, you rarely get the opportunity to see how it breaks and test articles are not cheap.
I agree as above, this is no news. All part of aircraft development.

I suspect this could be handled by a 3% reduction in MTOW for existing hardware, and a bit of beefup in subsequent airframes...

And if they planned the test well, the test specimen might have been a bit undersize, in which case the analysis could possibly show that production standard hardware needs no modification. :8

Got an issue... get a tissue.:{ :{ :{ :{ :{ :{ :ugh: :ugh: :ugh: :ugh:

That's why American refuses to by any additional Airbus products even though regardless of the financial advantages offered by Airbus... which reminds me of an autmobile offering it's customers cash insentives.

I'm not sure some of you have read this correctly

“This static test airframe has the first set of wings built, and we have refined the structural design for subsequent aircraft due to increased weights etc. We will use this calibration of the FEM to prove the adequacy of the structure on production aircraft.”

Bascially what they are saying is that the current one failed close to the point predicted in the FE analysis. The wing has been redesigned anyway, and now they have calibrated the FE model and it is accurate they are happy to take the FE analysis of the new wing as read.
Whats the problem? Standard practice, calibrate an FE model at a few test points and then use the model. If you don't like that process then don't get in an aircraft, or a boat, or a car, or .....

Umm, sorry but this is no big deal. What it shows is that the Finite Element Modelling and the testing rig produced failure within 3% of where it was supposed to be - which is nothing.

I haven't found out the Airbus G limits but the failure load was 1.45 times the +G limit so it has no effect on aircraft performance or weights.

Furthermore, both Airbus and Boeing use failsafe design techniques so that the failure of one structural element does not cause the rest of the structure to fail (absent the tail attachments).

This testing will go on forever, with cracks appearing and being fixed etc. etc.

It never ceases to amaze me how cynical Ppruners can be. The same people who lamented the technological backward step when Concorde was retired are hell bent on deriding Airbus for trying to push the aviation boundaries once more. All I can say is that I'm glad Pprune wasn't around when the Wright Brothers were doing their thing. We'd still be travelling by horse if some Ppruners had their way.

Gentlemen, this aircraft SHOULD be something the aviaiton world supports and welcomes. Yes, there will be problems along the way. Yes, those problems should be discussed openly and constructively. But why the hell are we trying to condemn one of the biggest aviation developments in 30 years to the graveyard before it has even finished its' testing phase? :hmm:

Well,
I guess, as I said, they won't tell for which loadcase that happened. So they will fix it, adjust the loading limit for that specific maneuver, for example.


Sunfish:

I would somewhat disagree with you. In my opinion this is sort of a big deal since they made it public. Maybe info leaked outside of Airbus so, in order to prevent rumors, they went public which is OK. Failure during test is a bit embarrassing, after all, since you do assume that the structure was designed with some overshot. Going on the zero margin will save weight but it will bring you too close to the edge.

One more slight correction: from what I am aware last 30 years all aircraft are not fail-safe anymore but "damage tolerant". That means they can sustain some damage that can be then followed by proper inspections etc. etc.

Cheers

It's no big deal. If it is not strong enough they just add more structure to compensate. Adds weight but then it will make the requirements.

:)

During testing the Boeing 777 wing failed at 154% of the design load. The Boeing engineers were hoping the B777 wing would fail at 157%.

The A380 wing failed between 145% and 150% of the design limit. For all intents and purposes, it sounds like the A380 wing failed at the expected limit of approximately 150%.

False capture:

That is right as long as you're on the other side of 1.5 mark. If you go for zero margin assumption you're exposed to either this or flight test results that are actually requiring a cert report update.

In Boeing's case there was no need for changes.

Still it seems to me we do not have enough info. Are loadcases comparable (your examples)?

Are these the most critical load cases? Is it the same part of structure?

Too many unknowns...anyway nothing is unsolvable.

Just a little bit of pre-emptive PR.;)

Cheers

Grunf,
For certification purposes they're allowed a 3% margin of error/spread. As a result the failure was within limits for certification and almost 50% more than what you refer to as the "zero margin".

Using typical remarks, they are saying everything is beautiful and nothing is wrong here.

Isn't the 1.5 load factor a certification requirement ?

If so, how will Airbus address this structural failure below the set limit ?

EDIT - Answered my question FC ! - you must have posted when I was :)

Maybe they should snap 9 more and see what the average is.

slight topic drift, but still on topic, during these structural tests, are flight control movements tested at the same time? eg, the wing is stressed to 125% of its design limit, are all flight controls then tested through their full travel to ensure full control if the wing sees such loading in flight?

eg during unforseen severe turbulence, it would be a bad thing if the wing is stressed to 100% or slightly higher, and the aileron control jams or deflects causing temporary loss of control...

As an ex-airbus engineer I can only re-iterate that this is a non-event the whole idea of the FEM is precisely to catch this problem and as AB say the production models have been re-engineered due to weight change requirements if the wing then failed now that would be a problem and a possible redesign.

You are confusing "factor of safety" with "margin"
Factor of safety is the relation between ultimate load and limit load, generally a 1.5 (or 150%) is required here. Some special factors may apply.
Margin is between rupture load and ultimate load.
Zero margin means rupture at ultimate load, or at 150% limit load.

No.
Everything tested above limit load is not "seen in flight" (at least it should not).
The rule requires flight testing of the control system up to 80% limit load and ground testing (or analysis) of the control system up to limit maneurvre load.
Nothing above limit load is tested in combination with other loadcases or other systems.

If in doubt, read the rule. It´s available on the web.

i.e. CS-25

CS 25.305 Strength and deformation
(a) The structure must be able to support limit
loads without detrimental permanent deformation.
At any load up to limit loads, the deformation may
not interfere with safe operation.
(b) The structure must be able to support
ultimate loads without failure for at least 3 seconds.
However, when proof of strength is shown by
dynamic tests simulating actual load conditions, the
3-second limit does not apply. Static tests conducted
to ultimate load must include the ultimate deflections
and ultimate deformation induced by the loading.
When analytical methods are used to show
compliance with the ultimate load strength
requirements, it must be shown that –
(1) The effects of deformation are not
significant;
2) The deformations involved are fully
accounted for in the analysis; or
(3) The methods and assumptions used
are sufficient to cover the effects of these
deformations.


CS 25.683 Operation tests
(a) It must be shown by operation tests that
when portions of the control system subject to pilot
effort loads are loaded to 80% of the limit load
specified for the system and the powered portions
of the control system are loaded to the maximum
load expected in normal operation, the system is
free from –
(1) Jamming;
(2) Excessive friction; and
(3) Excessive deflection.

(b) It must be shown by analysis and, where
necessary, by tests that in the presence of
deflections of the aeroplane structure due to the
separate application of pitch, roll and yaw limit
manoeuvre loads, the control system, when loaded
to obtain these limit loads and operated within its
operational range of deflections can be exercised
about all control axes and remain free from-
(1) Jamming;
(2) Excessive friction;
(3) Disconnection, and
(4) Any form of permanent damage.

Can I state that I am not an aircraft engineer and I do want the A380 to work (as long as it doesn't mean I am stuck in the back row of a high density A380 with fat passengers!).

But could this wing have broken after being given a lot of other stresses, and thus been weakened? Or should the wing always be able to take that much stress through cycles (i.e. 20,000 flight cycles at maximum loading).

There's a misconception that "beefing things up" will always give increased failure load.

The opposite effect has sometimes been the unexpected outcome (well, unexpected until you consider, for example, the redistribution of loads caused by "beefing up"). :uhoh:

Volume thanks for the explanation.

To all of you thinking this is a non-event:

OK, so if that is really the fact then why did Airbus came public with that info?

It should stay within their realm, right?

non event.
 

Why this news is "made public", and why does this thread exist? Because the destructuve tests are far more news worthy than the non-destructive tests.

Is it normal practice to validate an FEA model based on empirical data? Yes, absolutely.

Is 3% a significant error margin? Not really, what is significant is what is done with the result afterwards.

Should we be more interested in details such as the crack propogation rate that was caused by the non destructive cyclic testing (that represents real world loading scenarios)? Yes, but this is dull so won't get the same tabloid attention.

What I would like to know is if the component(s) that failed in these tests were the same component(s) that first exceeded their elastic limit in the FEA modelling.

Guys - A few points
1. Airbus has more than one structural test wing set. 150% will be achieved on a set built closer to the final build standard (see below).
2. Luoto - the final destructive test is only completed at the end of the life cycle tests. So the wing set in question would have completed as many cycles as any aircraft achieves in service...then failed at 147%
3. 150% may not sound like much but if you check out the pictures you will see the wing tips at around 90 degrees to the horizontal before the final failure. Hands up who has ever experienced anything like this in flight.
4. Throughout the 18 months of extreme fatigue testing minor failures are induced (this happens on every aircraft type). The repairs weaken the wing prior to the final destructive test. See coments on stress allowances below.
5. At day one development wings are weaker than the production wings for two reasons.
First - The first seven aircraft each have hundreds of miles of Flight Test Instrumentation cable passing through scores of extra holes drilled in the rib / spar wingbox structure.
Secondly - Despite the fact that Airbus assembly personnel have massively superior training than their American opposite numbers...EVERY first off aircraft ever built has hundreds of design "concessions" on each wing. These are , for example, slightly oversize holes...drill runs...tooling causing loss of land on hole edges.
These errors are a fact of life in the E Fatigue and E Static wings and Flight Test only aircraft... but are iorned out before the production batch. Stress engineers will theoretically factor these weaknesses into the final stress calculations. If you have ever spoken to a stress engineer though you will know that they are the most conservative people on the planet. Built into their calcs are huge comfort factors so a critical concession which could cause a 10% reduction in rib 2 strength will either be assumed to have no effect or, they will begrudgingly allow between 1 and 2% to the fatigue testing people.
Bottom line wing set probably achieved closer to 160% rather than 150%. It really is not a big deal.

Could Airbus be disclosing this to avoid any legal problems in the future "just in case" and to show open ness?

RMC:

Being one of those "conservative people" :E I would still say that this news needs far more explanation then there is room in FI or anywhere else in public.


It can be open to discussion and there is no valid info out there (and it shouldn't be).

Rumors yes, but valid news no.

As for the personnel training I hope you were employed at both sides ( Airbus and Boeing or Northrop or Lockheed) in order to comment on their skills.

They all do make mistakes, more or less, the same way.

As for the stress and testing I am sure that public picture will be right in the end (as will the official EASA side as well). It is hard to explain all that in plain English to general public assuming that this is the FI's target audience.

Slight problem might occur if the whole thing was designed for a 0% margin (I've seen these type of requirements) although it is still avoidable with "conservatively" mentioned above.

I am happy to hear they can do the fatigue test in 18 mths - a very fast schedule comparing to others, at the same task.

As tallsandwich I would also like to hear the crack propagation scenarios. This would tell us more how good this structure really is.

No it doesn't and the wing didn't come off either. A component simply broke.

Those pesky French !
 

Good post ocnus, thanks starting this thread and for letting us know this important information.

It's getting to the stage where it just won't be safe to be outside for fear of getting hit by Airbus fins or wings and things dropping out of the sky.

Wouldn't happen to a Boeing you know. No way! Just try to imagine a fin falling of a Boeing. Impossible! Although some clown once tried to tell me that a fin dropped off a 747 in Japan. What rubbish! Only happens to Airbus.