Boeing 787 dreamliner : troubles

[misako]

Voids of less than 1% are the desired goal in hand laid up parts. However, depends upon manufacturing method, with filament wound parts having 3% or more voids. I suspect (no knowledge) with tape laying (a la Boeing tape laying) that 1.0- 2.0% is accepted level. If higher than that you run into fatigue, flex strength and compressive strength issues. It is not a dimensional issue, but one impacting design allowables in fatigue and static strength regimes plus creating moisture ingress issues.

[Avro]

Thanks for your input misako.

It is not a dimensional issue, but one impacting design allowables in fatigue and static strength regimes plus creating moisture ingress issues.

What I ment with dimensioning was for example that you could add more layers to increase static strength. Won't you gain some static strength by adding layers on cf/epoxy ? As for fatigue cracks, doesn't cf/epoxy behave with a fibre bridging mechanism as with hybrid materials ? If yes adding layers would also solve the problem no? Any help is appreciated because I don't know a lot about cf/epoxy composites.

Chris

[Stepha380]

CFRP layers have to be pressed regularly either/both manually or/and by vacuum during the manufacturing.
A lot of problem may occur during the curing in autoclave especially if you have a loss of vaccum. The pressure is applied by compressed nitrogen inside the autoclave.

Basically the vacuum is created by laying up a plastic film fixed by sealant on the working tool on which all CFRP and honeycomb have been laid up and the air is pump between the tool and the film through a very thick textile. Sometimes the sealants may unstick the tool and there is a loss of pressure. The pressure is mesured inside the autoclave during the curing and you may exactly know at which moment it happened but you can't do anything but throwing it to the bin and doing another one. It happens rarely for big pieces because of the ratio leek/surface but never say never

The cost of 1 square meter of CFRP with aeronautical specs was around $60 3 years ago but it depends strongly on the specs and volume. Consider hundred of square meters and layers up to twenty (don't do math it will be useless, because every piece of every layer of CFRP is cut to the match the stress that the piece will face)

Hope it will help people to understand
Molding
http://www.netcomposites.com/education.asp?sequence=56
Curing
http://www.netcomposites.com/education.asp?sequence=61

[misako]

Chris,
Thank you for yours. In terms of voids, you need these as low as practical as stated earlier, otherwise you have design allowables problems as well as moisture ingress issues. For a decent and acceptable composite you need a final cohesive structure per spec without large number of voids or disbonds. If your process results in such, (can be broken vacuum bag or other problems such as poor tooling allowing air ingress, mandrel collapse or defects), you have a non-acceptable part from both QA and engineering standpoints.
Boeing had mandrel tooling problem in their case whose shifting or partial collapse resulted in a voidy fuselage barrel. Clearly this was unacceptable and Boeing did completely the right thing in scrapping the part and in reverting to their earlier successful steel mandrels. Composites are totally in-process dependent and you must achieve a cohesive, repeatable and dimensionally correct part free of significant flaws or else you end up with rejected parts in most cases. Building to the process specification is key in addition to original materials meeting all aspects of material specification. Adding plies to hopefully compensate for bad materials or processing defects does not fix the problem, but exacerbates it. The composite material is governed by two key specs, the material spec for inital material acceptance and the process spec for manufacturing. Both must work well and be adherred to or scrapped parts result. There is also a Janus problem with composites as you want to build large monolithic parts for cost efficiency and because composites don't like joints. Conversely large parts cost a lot to scrap. If you have cured the part and have high voids for example, it is scrap, no decent repair is possible. It is an accept/reject issue in end which depends strongly on having good specifications and following them rigourously. Hope this helps and doesn't confuse matters. It is totally and always will be process dependent issue for composites, at least in my experience.

[Avro]

Many thanks for your detailed answer misako. I understand your point very well now - and agree with you.

Chris

[smokejumper]

There has been much discussion regarding Boeing's 787 fuselage containing voids. "Randy's Journal" of June 30 says that they've made 8 fuselage sections without defects and used a new lightweight mandrel (experimental) for the 9th - this is the one that caused the issue.

Boeing is not out of the woods with the 787 and many unknowns will come up over the next 2 years. They will have problems - let's see how well they address them!

[Avro]

Boeing is not out of the woods with the 787 and many unknowns will come up over the next 2 years. They will have problems - let's see how well they address them!

Very well said!! And that’s just the beautiful thing about the rivalry between Airbus and Boeing. Both continuously need to push themselves beyond their limits in order to compete with the other. This will only be beneficial to the customer who in the end will get better aircrafts throughout the years.

[beaucaire]

I appreciate the post's of Misako,who give a great insight into the complexities of new materials.
Could you elaborate some lines on how potential fuselage -dammages will be repaired on the 787 ? (typically the service-truck running into the lower part of the aircraft body..)
It is my understanding it is not that simple ,in that the multi-layer structure of the 787 body requires very dedicated ways of laying layer -upon-layer in a very special manner.

[misako]

Thank you for the kind words, Beaucaire, they are much appreciated. In this area both Boeing and Airbus have similar problems re ground impact, sorry Chris, but it is true.
My feelings are that both aircraft are susceptable to ground damage and results of ground truck hit would be similar in either aircraft. In either case, either CF or GLARE, there probably would be disbonding and internal delamination as well as potentially cracks, holes and other damage. Extensive QA via ultrasonic, crack detection and tap tests and other inspections would be required to determine extent of damage, bearing in mind, in composites, that just because it is not visible damage does not mean that damage does not exist. Repair schemes can be successfully used, but some big caveats apply; namely total extent of damage must be determined, suitable repair schemes and procedures in place and tested, and , finally, and this is sometimes the biggest headache, competent, trained and qualified on-site airline personnel exist and are available on site. I mention this as a repair scheme can well exist, but if improperly performed, renders the repair dangerous and likely to reccur. Some of us or perhaps many of us in aerospace, harbour doubts regarding competence in airlines on a worldwide basis and other in-service aspects such as aging, fatigue and the 20-30 years in-service life expectancy for civil aircraft only compounds this concern. This is a far cry from military aircraft in-service requirements. Note also, I have not mentioned certificating agencies in this mix.
In addition, repair schemes must recognise that the materials involved are anisotropic and not homogeneous as in metals. Care, great care, is needed to ensure that original stiffness is matched and often this is missed. Load follows stiffness and this should be borne in mind in all repair schemes. Generally stepped joints are used with offset on a layer by layer basis to ensure good repair strength and good laminate performance while scarfed joints have also been employed.
Finally as well as competency or potential lack of it in airline customers, there is question of high cost of repairs in hybrids and composites. It can be done and can be done well, but costs money and time and demands high skills from repair personnel as well as qualified and certified procedures of all those involved as well as OEM properly developed repair specifications. Airlines don't like high cost repairs and rightly so.
In summary, a big challenge to all for both B-787 and A-380. Plenty of military experience, but commercial field is far different to military in many respects. Hope that this summary helps and best regards to all.

[Avro]

In this area both Boeing and Airbus have similar problems re ground impact, sorry Chris, but it is true.

No need to be sorry misako I don't say that Airbus has no troubles with Glare I just want to point out some specific advantages of a hybrid such as Glare compared to a full composite such as CF reinforced epoxy.

In either case, either CF or GLARE, there probably would be disbonding and internal delamination as well as potentially cracks, holes and other damage.

The damage done during impact depends on: how the stresses build up, how they compare with the various dynamic material strengths, how they are redistributed as the fracture progresses, and the fracture energies and the energy available to drive the fracture process. These in turn depend on the material properties and geometry and energy and momentum of the projectile.
Impact at low energies can result in surface damage and delamination. Impact at higher energies can produce cracks through the whole thickness or complete perforation.

Extensive QA via ultrasonic, crack detection and tap tests and other inspections would be required to determine extent of damage, bearing in mind, in composites, that just because it is not visible damage does not mean that damage does not exist.

Indeed accurate non-destructive testing methods are needed to detect damage in composites. The aircraft should be able to sustain detectable damage for a certain period before its detection. According to the US Airforce the flaws and damages that are assumed to be non-detectable for composites are scratches (a surface scratch that is 100 mm long and 0.50 mm deep), delamination (delamination that has an area equivalent to a 50 mm diameter circle) and impact damage (impact of a 25 mm diameter hemispherical impactor with 135 J kinetic energy, or with that kinetic energy required to cause a dent 2.5 mm deep).

For Glare it is difficult to C-Scan the specimens because not only the delamination but also the plastically deformed dent in the aluminium layers will reflect the ultrasounds since it is in general not perpendicular to the signal. However, Glare has a visible dent after impact and the damage size is always smaller than the dent size. This is one of the big advantages of Glare with respect to impact damage. This means thatone doesn't need to invest in very costly materials and highly skilled people in order to detect some impact damage on the field. Since the dent size is always bigger than the internal damage the total damage to the stucture is easily identified. This however is not the case with CF IIRC, which is the tricky part for CF.

The picture below shows what I just explained. The picture shows the result drawn from several tests performed on Glare specimens.



As can be seen the internal damage due to debonding of layers is inferior as the dent which is visible to the human eye. r.

To conclude Glare is stronger than aluminium at higher impact velocities due to the strain rate dependent behaviour of the fibres. The dent depth after impact is comparable to aluminium with the same thickness. The residual properties for Glare are comparable in compressive strength, and better than monolithic aluminium in fatigue. The inspection of a Glare structure is not more difficult as for aluminium, since Glare has the same indentation and because the damage size due to the impact is smaller than the visible dent size.

Repair schemes can be successfully used, but some big caveats apply; namely total extent of damage must be determined, suitable repair schemes and procedures in place and tested, and , finally, and this is sometimes the biggest headache, competent, trained and qualified on-site airline personnel exist and are available on site.

As explained above for on site repair conventional methods can be used to identify the damaged area (Again, I'm talking about Glare).

Finally as well as competency or potential lack of it in airline customers, there is question of high cost of repairs in hybrids and composites. It can be done and can be done well, but costs money and time and demands high skills from repair personnel as well as qualified and certified procedures of all those involved as well as OEM properly developed repair specifications. Airlines don't like high cost repairs and rightly so.

Here again for Glare it has been shown that the damage can be repaired by applying a normal Aluminium riveted patch. This means that we don't need specific tooling and personal for on field repairs. Further more the Al riveted repairs on the Glare structures show a fatigue life which is considerably better compared to Al structures with Al patches. Of course bonded repairs could be applied as well. However this option would cost a lot more and require more qualified personal to perform the task.

I fully agree with your post Misako and don't deny that there are a lot of problems which might occur with CF structures and wih hybrid ones such as Glare, especially on the operational side (impact damage). But with my post I just want to point out that the hybrid materials such as Glare show some non negligable advantages compared to CF. This of course doesn't mean that it's perfect

Some of us or perhaps many of us in aerospace, harbour doubts regarding competence in airlines on a worldwide basis and other in-service aspects such as aging, fatigue and the 20-30 years in-service life expectancy for civil aircraft only compounds this concern.

Indeed this will be a major issue. If Airlines need to invest a lot in tooling and highly qualified personal in order to repair their composite aircraft some issues will occur sooner or later.

Hope that this summary helps and best regards to all.

Very nice summary. I don't want to sound contradictory with my post or whatever, I just wanted to complement your point of view with some additional info on hybrid materials. Feel free to critisize

Best Regards
Chris

[misako]

Chris, thank you for sagacious and thoughtful response and I like glass fibre as was involved heavily in Corvette leaf spring FEA work a few eons back. Nevertheless, one must be careful re glass and in particular, composite/ metallic hybrids as I will seek to illustrate. My experience regarding hybrid metallic/composite structures started a while back like 60's or 70's in early days of "advanced composites". NASA had a big research contract with Lockheed Georgia to reinforce the spar caps of C130 aircraft with boron/epoxy strips bonded to existing aluminum spar caps. NASA's idea was to have aluminum work to limit load and then boron to carry remainder of load up to ultimate load. I was not directly involved in this program, which occurred at as time when early C-130's were having wing fatigue troubles in service, but was an interested observer and investigator. NASA was also playing with CF/ metallic hybrids, but again CTE aspects defeated them. The Achilles heel of the C-130 program as with the NASA CF/aluminum programs was the differential CTE (coefficient of thermal expansion), being high for aluminum and very low for boron, as it is for glass, I must mention in passing. depending upon units used boron was in 2 or so range whilst aluminum was in 13-14 range if memory serves me correctly without bothering to refence my text books. Clearly on 130 foot wing span this was a killer due to the differential expansion ratios as aircraft had to operate from -55 degrees F to + 180 degrees F, just as does the A-380. NASA tried lower cure temperatures, low modulus/high shear adhesives and ending up with doubtful capability RT cures, but physics won the day and the program quietly died.
Now why do I bring this point up? Well good old glass is down around CTE of 2-3 and in GLARE is bonded to good old high expansion aluminum again I am assuming. Similarly, glass is not a solid, but rather a supercooled liquid and hence has a low tensile stress rupture point of around 30 KSI vs a short term tensile strength over, let us say north of 350KSI. I am no GLARE expert and don't know size of GLARE panels on A-380, but would hope lots of thermal cycling tests of large monolithic structures in lengths of say 100 meters or so or at least 1.5 X the longest length of monolithic GLARE used on A-380 would have been and is a good idea. Now these tests may have been done, but I am wondering whether delta CTE problems could raise its ugly head again re GLARE as it has in other hybrids. In case of ARALL worse problems exist as Aramid fibers have negative CTE. I don't know adhesive used in GLARE and its cure temerature, but physics still exist.
I raise this point, Chris, just to get your kind attention and expert response and also, if CTE thermal cycling on large scale structures not yet done, it is something to review the need for and to mull over. Hybrids have advantages, but have disadvantages also. It will not show up on small specimens, but is there on large scale structures with a vengance if history is any guide.
All best regards as ever and if I am wrong, biff me.

[beaucaire]

Misako,Chris -thanks for the excellent responses which do elaborate quite perfectly the issues related to field-repair and the complexities involved.
A quite refreshing difference to some a.net posts....
Lets's hope both aircraft - A380 and 787- will see a bright future.. and the due announcement of the A370 (?) will bring some serenity to the people in Toulouse,and Hamburg !

[Avro]

The issue you mention Misako is a very interesting one. I've been taught about the differences in CTE's concerning the interaction between the repair patch and the structure. However never about the possible differences inside hybrid materials.

The Achilles heel of the C-130 program as with the NASA CF/aluminum programs was the differential CTE (coefficient of thermal expansion), being high for aluminum and very low for boron, as it is for glass, I must mention in passing. depending upon units used boron was in 2 or so range whilst aluminum was in 13-14 range if memory serves me correctly without bothering to refence my text books.

I just looked up in my notes and you seem to remember quite well

CTE Al: 23.6
CTE boron/epoxy: 4.5
CTE Glare: 16.3

Due to the fact that Glare is given a high CTE we were always told that this can be an advantage when applying Glare patches instead of boron/epoxy patches to Al structures. However Boron is on some cases preferred since it's stiffer IIRC.

Your post made me think however. Glare is given a CTE of 16.3 but this must be some sort of average since as you said there must be a big difference between the Al and composite layers. This will eventually result in some thermal residual stresses in the structure itself. And ultimately I could think of delamination problems if the stress becomes too high... Since I don't know the real answer to this I've send an e-mail to a teacher in the hope that he will give me an answer to that subject.

I am no GLARE expert and don't know size of GLARE panels on A-380, but would hope lots of thermal cycling tests of large monolithic structures in lengths of say 100 meters or so or at least 1.5 X the longest length of monolithic GLARE used on A-380 would have been and is a good idea.

I don't have the numbers right here but the biggest panels should be 11.8m long and 3.2m high. The size limit being the size of the autoclave.

Now these tests may have been done, but I am wondering whether delta CTE problems could raise its ugly head again re GLARE as it has in other hybrids. In case of ARALL worse problems exist as Aramid fibers have negative CTE. I don't know adhesive used in GLARE and its cure temerature, but physics still exist.

Delta CTE problems could certainly arise for Glare. I hope I'll get an answer soon so that I can let you know. The cure temperature for Glare is 120°C , I don't know the exact type of adhesive which is being used tough.

if CTE thermal cycling on large scale structures not yet done, it is something to review the need for and to mull over. Hybrids have advantages, but have disadvantages also. It will not show up on small specimens, but is there on large scale structures with a vengance if history is any guide.

Hybrids have certainly disadvantages as well. And as you said in an earlier post Airbus, Boeing and the airlines will face some challenges with those "new" materials. We certainly don't know everything about them yet. And ageing will be of a big concern.

What happens if an airplane flies in very aggressive athmosphere (salt air) ? How does the adhesives behave to such exposures in the long term etc... A lot of questions which still need to be anwered in my opinion (even if it is said that hybrid and CF have a good behaviour).

I will do my best to question the teacher so that I can bring you with answers.

Misako,Chris -thanks for the excellent responses which do elaborate quite perfectly the issues related to field-repair and the complexities involved.
A quite refreshing difference to some a.net posts....

Thank you very much for your kind words Beaucaire.

. and the due announcement of the A370 (?) will bring some serenity to the people in Toulouse,and Hamburg !

I'm really curious to see which type of material Airbus will go for. The worst option IMHO would be the AL-LI fuselage which is nothing else than a lighter Al fuselage....

Best Regards
Chris

[Stepha380]

An article dealing with potential risk of 787 delays, value of Boeing shares, supply chain issues...

http://articles.moneycentral.msn.com/In ... heSky.aspx

If the link doesn't work, leave a pm.

First class article :rock:

[beaucaire]

unfortunately the link does not work-would you happen to have a soft-copy of the article?
thanks in advance

[misako]

Thank you , Chris , for excellent and objective set of responses and glass is definately in 3.0 to 3.5 range, so GLARE CTE claimed doesn't compute or glass and aluminum are involved in a big fight just as physics predict. Good luck on finding more detailed answers re GLARE and all thanks for your many detailed and kind responses. Would be fairly easy to do FEA bond line analysis and determine shear stresses at -55 degrees F or better at -65 degrees, fatigue strength of bond would be in 35% range of static or less for adhesive and watch for any glass stresses for extended period of over 30 KSI. I would worry if more than 300-350PSI in bond at -65 degrees F. Sorry, but you can tell how ancient I am from my use of degrees F.
Remember you have locked in stresses starting at around 250 degrees F, so temperature range involved is around 310-315 degrees re CTE and bond line stresses. Good project, Chris, for a dissertation!!
I did very detailed study of CF/epoxy versus Al-Li back in 80's and Al-Li came out very badly then, so totally agree with you re worst material for A-370. Thank you again for excellent and detailed thoughts and responses on this thread.

[Avro]

Good project, Chris, for a dissertation!!

It would indeed be a nice project for a dissertation But I've already chosen my project ..

Thank you again for excellent and detailed thoughts and responses on this thread.

I thank you for your expertise and the very nice and detailed explanations.

I did very detailed study of CF/epoxy versus Al-Li back in 80's and Al-Li came out very badly then, so totally agree with you re worst material for A-370.

Indeed, that's why I was surprised when I first heard that Airbus wanted to use Al-Li in the A350. But then I also heard that France had some stakes in the Li industry or something similar. France would thus push Airbus to use AL-Li in order to boost their Li market. That's what I heard as the reason behind the Al-Li choise. However I'm not 100% sure whether it's true or not.

I'll keep you in touch once I get the responses from the teacher.

Cheers
Chris

[misako]

Dear Chris,
Well, it was a thought re dissertation and your kind words are greatly appreciated. Maybe we can get your teacher to investigate and be a hero too. And, yes, re Al-Li was always a French industry producer push, but I think very wise if EADS don't use it based upon my earlier study results, but their choice.
All best regards as ever

[CX]

i wonder what everyone's reaction will be if Airbus STILL goes for the Al-Li that was planned to be used on the A350, i will surely be gutted! but honestly with so little time for planning, i think the fact is that there won't be much surprises... Just about 2 weeks left until Airbus announces the A370!

and Airbus has not updated the A350 stuff on their website...

[DFW]

In the book Twentieth Century Jet and subsequent documentary(which was mostly a silly piece of marketing), Alan Mullaly went to great lengths to explain why Boeing didn't use Al-Li on certain parts of the B777. His reasoning was that surface cracks, which are tolerable, would none-the-less make customers uneasy.

Are there other disadvantages to Al-Li?