Dvautier From United States of America, joined Oct 2006, 35 posts, RR: 3 Posted (5 years 3 months 3 weeks 6 days 11 hours ago) and read 18268 times:
I still don’t get it. Airbus says that the wire bundles did not match up because the Hamburg plant was using a different version of Catia (V4). This does not make any sense at all because the plane fitted together even though some parts were designed using the Catia V4 wireframe formulas in Germany and the other parts were made in France using volume formulas (Catia V5). This means that they did get the parts to fit because the planes are obviously flying. It also means that the wire raceways are probably also correctly sized. So why didn’t the wires fit? Changing from copper to aluminum should not be a big showstopper. Catia V4 and V5 have nothing to do with this issue so let’s stop bringing it up as a reason for the A380 delay. I am quite familiar with both Catia products. Neither version handles wiring. You need additional software tools to do this and I don’t know if it even exists let alone works with Catia. Something else is going on here that we just don’t know about.
There is another consideration that most of you seem to miss, and this is the number of connectors between bundles. When you design wiring you try to make the bundles manageable by locating the wiring panels at convenient and serviceable locations. This helps in fault isolation and replacement. But the more panels and connectors, the higher the resistance in the wire and it also generates a lot more weight and heat. If airbus decided to run vary long bundles that go across sections and not to panels they may be getting into some serious design and servicability problems. It might also explain the present mess. I suspect that they are actually redesigning some stuff.
If we could just get hold of the schematics and wire diagrams for the A380 a lot of these questions could be answered. But I suspect that these documents do not yet exist.
NYC777 From United States of America, joined Jun 2004, 4634 posts, RR: 51 Reply 1, posted (5 years 3 months 3 weeks 6 days 11 hours ago) and read 18274 times:
I think it might have been the design of the IFE wiring harnesses that were only affected but since I'm not there, it's only my guess.
N328KF From United States of America, joined May 2004, 6197 posts, RR: 4 Reply 2, posted (5 years 3 months 3 weeks 6 days 11 hours ago) and read 18231 times:
Quoting NYC777 (Reply 1): I think it might have been the design of the IFE wiring harnesses that were only affected but since I'm not there, it's only my guess.
If it were just for IFE, then the cargo variant wouldn't have been affected. I believe the flying examples have some juryrigged wiring that allows them to prove the airframe and systems, but this wiring is not the final configuration.
When they call the roll in the Senate, the Senators do not know whether to answer 'Present' or 'Not guilty.' T.Roosevelt
DAYflyer From United States of America, joined Sep 2004, 3807 posts, RR: 4 Reply 3, posted (5 years 3 months 3 weeks 6 days 11 hours ago) and read 18168 times:
Quoting N328KF (Reply 2): If it were just for IFE, then the cargo variant wouldn't have been affected. I believe the flying examples have some juryrigged wiring that allows them to prove the airframe and systems, but this wiring is not the final configuration.
I was going to ask this exact question...how can they fly the planes now with this wiring issue?
787engineer From United States of America, joined Dec 2005, 572 posts, RR: 17 Reply 5, posted (5 years 3 months 3 weeks 6 days 11 hours ago) and read 18141 times:
In my opinion it was originally a a wiring issue. Probably not a big deal hence the shorter delays at the beginning. I suspect that now it's a configuration control and configuration management issue more than it is a wiring issue.
TeamAmerica From United States of America, joined Sep 2006, 1760 posts, RR: 24 Reply 6, posted (5 years 3 months 3 weeks 6 days 10 hours ago) and read 18050 times:
Quoting Dvautier (Thread starter): This does not make any sense at all because the plane fitted together even though some parts were designed using the Catia V4 wireframe formulas in Germany and the other parts were made in France using volume formulas (Catia V5). This means that they did get the parts to fit because the planes are obviously flying. It also means that the wire raceways are probably also correctly sized.
Catia was used to model the physical arrangement of the aircraft, and hence the relative location of penetrations and raceways. My interpretation of events is that the wiring harness was designed with minimal slack (as it should be) but this allows for no deviation in the expected physical arrangement. If a raceway is not exactly where expected, connectors, branches, drip loops etc. all wind up in unintended locations. Unitended locations means unservicable at best, but more likely will not meet specification. And this is just the simplest of problems.
I believe the size of penetrations and/or raceways may not have been as expected by the electrical engineers. This could be due to the conversion from Cu to Al, or simply some sort of conversion flaw in moving between Catia versions. Whatever the case, if a hole is too small you may find that the connectors won't pass. If a raceway is too small either the bundle doesn't fit, or you are stuffing power and signal wires too close together - in either case you don't meet specification. Bottom line is that the root cause truly can be traced to the incompatible versions of Catia. I tend to take Airbus at their word so far as this is concerned.
Quoting Dvautier (Thread starter): If we could just get hold of the schematics and wire diagrams for the A380 a lot of these questions could be answered. But I suspect that these documents do not yet exist.
I agree! I don't think that Airbus has a complete wiring scheme at all. Note that Airbus is holding SQ's first aircraft for a full year. I interpret this to mean that they are fitting that aircraft by trial and error until they get a certifiable design, which they will then need to retain in order to document and replicate the successful installation. If I am correct, it is an indication of just how utterly out of control the current situation really is.
Dvautier From United States of America, joined Oct 2006, 35 posts, RR: 3 Reply 7, posted (5 years 3 months 3 weeks 6 days 10 hours ago) and read 17972 times:
So Airbus is using it as a full scale mockup. Daaaa...Good idea.
Baron95 From United States of America, joined May 2006, 1335 posts, RR: 10 Reply 8, posted (5 years 3 months 3 weeks 6 days 10 hours ago) and read 17915 times:
Quoting Dvautier (Thread starter): Changing from copper to aluminum should not be a big showstopper.
I would not trivialize this issue. Switching from Cu to Al probably most likely meant increase in wire diameter, increase in connector size and complexity, reduction in the number of allowed connections per cable run, etc.
How many holes and raceways had to be changed because of that? How many individual airframe parts had holes and raceways going through them? When EACH one of these parts got redesigned, did it cause any other cascade redesign? Did they lose configuration control with various versions of these parts making their way into the assembly of MSN001-013?
In my mind (and I don't have any sources for this), the biggest failure was Airbus continuing to build planes and trying to certify them while all this was going on. If they had stopped earlier, dedicated one airframe to be completely fitted with wiring (IFE and all), redesigned all parts, built another frame and tested the wire fitting, they'd be out of the woods now.
Instead, they are faced with reworking 13 airframes, each possibly different than the other, then chosing ONE go forward solution to certify and produce. It is a massive undertaking, IMHO.
Dvautier From United States of America, joined Oct 2006, 35 posts, RR: 3 Reply 9, posted (5 years 3 months 3 weeks 6 days 9 hours ago) and read 17827 times:
I still feel that Catia is not the issue. In the CAD/CAM world you design something in 3D, then you extract what’s called cutter-line data which is then converted into machine tool commands by adding all sorts of postprocessor instructions for the specific machine tool. A part is produced out of foam and sent back to the engineer for check out. This process goes on and on. Catia is a design tool which is used way way upstream. I am certain that German and French engineers were shipping parts back and forth for a long time and they all fit together and there was no problem in using the two versions of Catia. That’s because there is so much manipulation and checking done to the data after the cutter-line gets extracted.
Quoting Baron95 (Reply 8): I would not trivialize this issue. Switching from Cu to Al probably most likely meant increase in wire diameter, increase in connector size and complexity, reduction in the number of allowed connections per cable run, etc.
excellent point baron. looks like we will be seeing yet another delay.
787engineer From United States of America, joined Dec 2005, 572 posts, RR: 17 Reply 10, posted (5 years 3 months 3 weeks 6 days 8 hours ago) and read 17732 times:
Quoting Baron95 (Reply 8):
I would not trivialize this issue. Switching from Cu to Al probably most likely meant increase in wire diameter, increase in connector size and complexity, reduction in the number of allowed connections per cable run, etc.
Personally I think Airbus had plenty of time to solve any Cu to Al issues. The issues aren't trivial, but Airbus knows how to design airplanes. I'm sure all the tradeoffs between Cu and Al were considered.
Quoting Baron95 (Reply 8): Did they lose configuration control with various versions of these parts making their way into the assembly of MSN001-013?
IMO loss of configuration control is really the only problem that can explain the long delays we're seeing. If Airbus can't prove configuration control, they won't be allowed to sell the plane; it's that simple. You have to prove that you know everything that goes on each plane and where it is (including the rivet that you upsized 1/32" because you drilled a bad hole).
AirSpare From United States of America, joined Jun 2006, 589 posts, RR: 6 Reply 11, posted (5 years 3 months 3 weeks 6 days 7 hours ago) and read 17666 times:
Quoting 787engineer (Reply 5): In my opinion it was originally a a wiring issue. Probably not a big deal hence the shorter delays at the beginning. I suspect that now it's a configuration control and configuration management issue more than it is a wiring issue.
Exactly. The WSJ report said that there were diferent final designs in use between different A offices. I said it before and I'll say it again, how can you design any subsystem to fit another when the actual designs being used were diferent? In all practicality, the aircraft was being built without a "frozen" design. Airbus probably needed to audit at what design was used for what wiring bundle, at what pahse, then redesign it.
I was flamed by an A employee for stating this opinion, but hey, 2 years later?
TeamAmerica From United States of America, joined Sep 2006, 1760 posts, RR: 24 Reply 12, posted (5 years 3 months 3 weeks 6 days 7 hours ago) and read 17625 times:
Quoting Dvautier (Reply 9): I still feel that Catia is not the issue
Ok, then what is? I don't think anyone is blaming Catia as such; it's more a case of misuse of software related to Catia. Airbus used an in-house tool to move data between the two version of Catia, and that misguided approach apparently is the root of the problem. So if you're saying not to blame Catia itself, I agree...but if you are saying that the A380 delay is not due to "wiring" issues I very much disagree.
Given the extreme scrutiny that Airbus is under, to lie about any aspect of the situation would be literally unbelievable IMHO. EADS is a publicly traded corporation. People go to jail for misrepresentation of facts that relate to the value of a stock.
CHIFLYGUY From United States of America, joined Aug 2006, 140 posts, RR: 0 Reply 13, posted (5 years 3 months 3 weeks 6 days 6 hours ago) and read 17557 times:
A Boeing (or ex-Boeing) person posted earlier that when they switched to CAD design for planes, they discovered that the CAD software on the market was inadequate for wiring. I wonder if this, at its core, is not the real problem.
Vref5 From , joined Dec 1969, posts, RR: Reply 14, posted (5 years 3 months 3 weeks 6 days 6 hours ago) and read 17498 times:
Quoting CHIFLYGUY (Reply 13): A Boeing (or ex-Boeing) person posted earlier that when they switched to CAD design for planes, they discovered that the CAD software on the market was inadequate for wiring. I wonder if this, at its core, is not the real problem.
No... that's not quite it.
Basically, Airbus in France used one version of the CATIA software, and Airbus in Germany used a different version of the software.
This was a major issue because the software versions are NOT compatible with each other, since they represents objects in very different ways.
Airbus initially tried to resolve the problem by getting someone to write software to convert data files from one version to another. Didn't work out the way they had hoped it would.
So now they are in this big issue... and to convert from one version to another would likely take closer to a year to retrain engineers and somehow convert the existing datafiles to the new version.
CATIA is very good software, mind you. It was used by Boeing for at least the 777 project (that I know of). It's not some POS thing. It's very complex, though.
Well, Airbus doesn't have a lot of options if their two major plants are using incompatible software. They are currently doing wiring by hand because of this, which makes it very difficult to mass-produce as well as set up something that will pass regulatory approval. This is the source of delay.
Unfortunately, Airbus does not have any real easy way out of this issue. And so, they continue to work the problem.
Baron95 From United States of America, joined May 2006, 1335 posts, RR: 10 Reply 15, posted (5 years 3 months 3 weeks 6 days 5 hours ago) and read 17455 times:
Quoting 787engineer (Reply 10): Personally I think Airbus had plenty of time to solve any Cu to Al issues.
787engineer, I thought that the switch from Cu to Al happened as part of the weight savings program that Airbus started late in the project. I'd venture to say that they decided to continue to take delivery of parts and buid planes even as they were making the changes and thus lost configuration control.
Think about how many parts were already built and going into final assembly as this change over ocurred. Perhaps some planes like MSN001-003 had the original parts + post-assembly hand re-work. MSN004-005 had some parts reworked prior to assembly and some parts manually re-worked after assembly. MSN006-013 had most parts re-worked prior to assembly plus some hand rework for IFE, and so on. Could be a nightmare to get under control.
Ken777 From United States of America, joined Mar 2004, 5887 posts, RR: 4 Reply 16, posted (5 years 3 months 3 weeks 6 days 3 hours ago) and read 17357 times:
Ramblings
First, can I assume that the manager/executive that decided against having both France and Germany on the same version of the software is no longer with the company?
On CU to Al, it's now probably known at Airbus that it would have been cheaper to pay the airlines for a little overweight 380 than the on-goign delay payments.
Would moving back to Cu significantly reduce the wiring problem, in terms of the size of the wires? I recall that the 380 was beating expectations on fuel consumption - is there room for a change back to Cu?
Remcor From United States of America, joined Feb 2006, 347 posts, RR: 0 Reply 17, posted (5 years 3 months 3 weeks 6 days 3 hours ago) and read 17312 times:
Quoting Ken777 (Reply 16): First, can I assume that the manager/executive that decided against having both France and Germany on the same version of the software is no longer with the company?
I don't think you can pin the blame on some manager because he didn't decide on a common software package. Well you could do that, but it neglects the fundamental problem that Airbus is having right now, namely the lack of a unified leadership.
WingedMigrator From United States of America, joined Oct 2005, 1913 posts, RR: 56 Reply 18, posted (5 years 3 months 3 weeks 6 days 1 hour ago) and read 17262 times:
Quoting 787engineer (Reply 5): I suspect that now it's a configuration control and configuration management issue more than it is a wiring issue.
Agree wholeheartedly. They must be drowning in a sea of badly handled Change Requests. Major loss of baseline control!
Quoting Baron95 (Reply 8): I would not trivialize this issue. Switching from Cu to Al probably most likely meant increase in wire diameter, increase in connector size and complexity, reduction in the number of allowed connections per cable run, etc.
We keep hearing this over and over... when did the switch actually happen? From my understanding it was early in the design phase, before metal was cut. I haven't seen the Al to Cu issue mentioned as a significant contributor to the delays in any trade publications (Flight International or Aviation Week), or in Airbus press releases. Is this another a.net urban legend?
Pygmalion From United States of America, joined Jun 2006, 871 posts, RR: 41 Reply 19, posted (5 years 3 months 3 weeks 6 days 1 hour ago) and read 17207 times:
Its not CU vs AL. Others have already discussed this on many threads... Airbus flat missed the boat on doing proper design integration across its many suppliers and more importantly, did not do a full digital mockup of the combined designs to be sure those designs were fully integrated. When the IFE and cabin bundles built in Hamburg showed up in Toulouse... they didn't fit. Not only that but they didn't know why they didn't fit.
Quote:
Experts familiar with Airbus design operations say that French factories, where most of the fuselage was made, used the latest version of a sophisticated design tool called CATIA, made by France's Dassault Systèmes, an independent software spinoff from airplane maker Dassault Aviation.
But the Hamburg factory's design center used an earlier version of CATIA dating from the 1980s. When the wiring arrived from Hamburg, it didn't fit correctly into the fuselage. Production ground to a halt as workers tried to pull the bundles apart and re-thread them through the plane.
Manufacturers frequently retain such legacy software alongside newer versions, says Robert Weigl, the Munich-based director of professional services for Proficiency, a Waltham (Mass.) company that specializes in helping companies integrate such systems. Airbus almost certainly could have avoided the wiring mess if top managers had stayed on top of the situation, Weigl says. "It's not only a technical problem, it's a question of project management, of seeing potential problems," he says.
MOVING THE MOUNTAIN. Compounding the problem, Airbus had been slow to move to new software that would allow it to create a full digital mockup of the plane, a three-dimensional computer model incorporating all the specifications and subsequent modification. Boeing is using such a system for the 787. Dassault Aviation is even building corporate jets from a digital mockup. Yet Airbus only signed its first major contracts for Dassault Systèmes' digital-mockup software within the past year, according to people with knowledge of the deal.
Airbus just plain blew it. It is managements fault for not having the correct controls and requirements to integrate the designs and it will cost them billions of Euros.
Its not CATIA. Boeing uses both V4 AND V5 in house and with suppliers. Boeing just happens to do full digital mockup and integration with their own proprietary software and have done so on every program and derivative since the 777. You have to do the integration work. From the beginning stages through the full final design process.... or over the two year delay.
Baron95 From United States of America, joined May 2006, 1335 posts, RR: 10 Reply 20, posted (5 years 3 months 3 weeks 6 days 1 hour ago) and read 17207 times:
Quoting WingedMigrator (Reply 18): I haven't seen the Al to Cu issue mentioned as a significant contributor to the delays in any trade publications
That is a good point - lack of specifics here on a.net. I'll try to summarize my understanding of the issue with some backup data. (I'm sorry this will be long)
First, here is what IS true (factual) about the A380 wiring and how it differs from other Airbus airplanes:
Quote: Some 300 of the 500 kilometers of wiring use aluminum instead of the conventional copper as the current conductor. Thus, 20 percent of the conventional weight has been saved. Aluminum itself is 50 percent lighter than copper.
The aluminum wiring technology has been used on all types of Airbus aircraft for big sections, that is wiring or cable where the cross-sectional dimension is 5 square millimeters (sq mm) or more. The A380 is the first aircraft to use the technology for small sections, under 5 sq mm. and for the connections. There are about 100,000 electrical links on the A380 compared with 60,000 on the A340.
The challenge was protection against corrosion. The new wires are made of nickel-plated, copper-clad aluminum strands. A special protection, based on composites material, has been developed against corrosion. A new specification was worked out for the aluminum connectors. The inside of the insulation material is made of hydrolysis improved polyimide tape. The outer insulaton is made of improved PTFE tape. Airbus also highlights the insulator's resistance to arcing.
Quote: "One issue was that the CATIA computer tool used in the airplane's digital design was not sufficiently accurate when it came to designing electrical systems," said senior Airbus executive Tom Williams. "The problem was made worse by Airbus' switch to aluminum wiring when the model was designed for copper wiring, which has very different physical properties."
"Airbus officials say poor design tools were a main cause of A380 delays. Problems with the Catia digital mock-up led to inaccuracies. [Airbus executive vice president for Programs Tom] Williams says the computer tool didn't represent the wiring harnesses well and, because there wasn't always good configuration control in the database, some design changes were not properly reflected and remained undiscovered until problems during wiring installation were encountered."
Regarding the timing of the switch over to Al, it was not an all or nothing proposition. As part of the weight savings program, more wiring was switched from Cu to Al. I have no idea when it stopped, but I know that at rollout in July 2004 the A380 was overweight by 4 tons and a weight loss program was being agressively pursued. At least one source is here: http://seattlepi.nwsource.com/business/182471_airbusproblem17.html
Cloudy From , joined Dec 1969, posts, RR: Reply 21, posted (5 years 3 months 3 weeks 5 days 13 hours ago) and read 16970 times:
When there is a delay in any large project, more often than not a software issue seems to be behind it. We just don't know how to do large software projects yet. Yes there are some people who are good at it. There are some practices we know are good and others we know are bad. But there is not an accepted and widely practiced set of techniques that consistently give good results - as there is in nearly every other form of engineering. the industry has not evolved to the point where it can consistently deliver results on time and on budget. A large proportion of software projects fail to ever deliver satisfactory results at any budget or timeline. If buildings were built like software, half of the Chicago skyline would consist of partly completed, abandoned shells.
To paraphrase a common quote - Software development is the only legal industry that calls its customers "users" .
WingedMigrator From United States of America, joined Oct 2005, 1913 posts, RR: 56 Reply 23, posted (5 years 3 months 3 weeks 5 days 1 hour ago) and read 16838 times:
Quoting Baron95 (Reply 20): I'll try to summarize my understanding of the issue with some backup data.
Thanks for the links, I appreciate that you took the time to post them. So it appears that a switch to Al wiring was one of several contributing factors to what is principally a configuration management problem.
This is very much unlike the insinuations that I've read elsewhere on a.net, that Airbus was having technical problems with the aluminum wiring itself (with the obligatory references to fire-prone aluminum house wiring of the 1970s). That story remains unsubstatiated.
Quoting Vref5 (Reply 14): CATIA is very good software, mind you.
Saintsman From United Kingdom, joined Mar 2002, 2065 posts, RR: 2 Reply 24, posted (5 years 3 months 3 weeks 4 days 23 hours ago) and read 16823 times:
Hypothetically, there is a problem when modeling electrical harnesses. Whilst they can be produced on screen and dimensions made it is a different matter when manufacturing them. Cable lengths are cut to size iaw the drawing dimensions. If they are cut without any tolerance there is a potential problem when it comes to making the wiring loom. If for example a loom has a 90 degree bend in it, the wires on the outside of the bend could be half an inch (12mm) longer than those on the inside. Put several bends in and you may find a wire that is too short to reach the connector.
I'm not sure that is the case with the A380 though.
Normally when designing a new harness, on the first installation the wires are made over length and the amount cut off is fed back to design who amend the drawings accordingly. This is why the first prototypes did not have the problem that the production aircraft have.
25 Nitrohelper: Thanks for your post , this is the kind info that helps us engineering folks understand what could have really caused the delays. I am surprised that
26 Baroque: Great post, almost enough to make me think I half understand!
27 ElGreco: For your information, A380 was designed from the origin with Cooper/Aluminium cable (each wire of cable is aluminium with 10µ of cooper and 2µ of ni
28 Zvezda: I thought that was due to the general shortage of copper during the war, as much as for weight.
29 Starlionblue: Have fun paying for the hardware and software to run the diagrams. Indeed. Software is just another tool. If you use it incorrectly, it'll come back
30 ElGreco: You absolutly right, but it's look like they were even interested for weight saving (Germans Friends from Airbus Hambourg told me that).
31 Baroque: A look through Janes WWII edition shows that the Germans had some pretty formidable planes in development at war's end, especially in terms of range
32 Starlionblue: I have this fascinating book about German WWII aircraft projects: http://www.amazon.com/Luftwaffe-Secr...02-1710107-1621766?ie=UTF8&s=books It's amaz