PDA

View Full Version : 787 Composites


blakmax
3rd Sep 2009, 10:40
Brian Abraham requested that I start a new thread after discussions about the AW139 which lost its tail taxiing in DOH (see Rotorheads forum).

As far as the 787 composite structure is concerned, I am a strong supporter of composites. However, they will fix some problems such as fatigue but create their own peculiar set of materials driven problems. The toughened resin systems used for 787 appear to have addressed the very poor impact resistance in older composite structures. But bolted joints in composites remain a real concern, and 787 uses them every where. The problem is that shear-out strength of composites is exceptionally low compared to metals. Further, the bearing strength of composites is so much lower than the bearing and bending strength of fasteners that the design is bounded purely by the properties of the composite.

My concerns relate to the Boeing proposed repairs. They are mechanically fastened so that Boeing can "dumb-down" the SRM so that Blogslovia Airways can perform repairs. The penalty on parts replacement and overall airworthiness will be paid by more competent operators with reliable bonding shops. Apart from thick composite structures, adhesive bonding still remains the most efficient way to repair damage.

There are differences in rules about edge distance, fastener row separation and fastener pitch which may trap the unwary repair designer who uses the rules applicable to metallic structure.

However, I have every confidence in Boeing's metal bonding processes (at least Boeing Commercial; Boeing's processes on C-17 are alarming!). They are at least aware of the issue of bond durability and the work of Kay Blohowiak (in conjunction with USAF Wright Patterson Lab's Jim Mazza) on the sol-gel process is world leading. Pity C-17 still uses the totally discredited Pasa-jel process.

I am unaware of proposals to repair delaminations and disbonds. If they propose injection repairs, then they need to talk to Walt Disney so that they can suspend reality. Injection repairs for disbonds can never and will never do anything other than fill the air gap so that NDI can not find the defect. There is zero chance of actually restoring any strength. With regard to injection repair for delaminations they should talk to Boeing STL. They at least admit that injection repairs do not restore pre-damage strength and at best slow down delamination propagation. Hence injection repairs can not be considered a permanent repair that restores the certification basis for the aircraft.

I'd like to extend the discussion not just to composites but to other aspects of adhesive bonded repair, especially hot-bonding technology. Is there enough interest out there in pprune land for me to spend my time doing that? Are the Moderators happy with locating such discussions in this thread? Please respond.

Regards

blakmax

Brian Abraham
3rd Sep 2009, 14:18
Here's one hand up.

leewan
3rd Sep 2009, 14:30
Here's one hand up.

Here's another. Am interested in the area of composite structural repair as sheetmetal seems to be going the way of the dodo with all new a/cs coming with composite fuselages.

Cardinal
3rd Sep 2009, 16:47
Bring it on!

muduckace
5th Sep 2009, 09:51
Man, I saw the quick patch 787 fuscelage procedure and laughed as loud as I do at FOX or CNN news. Temp/weather/surface condition? The AA A300-600 over long island got my interest. Aluminum primary structure is easy to GVI and easy to NDI if required. Composite material may look fine on the exterior but be seriously degraded inside and is not to todays standards easy to inspect.

Having said that and given the MAINT. schedules of operating aircraft. A composite airframe spending lots of time in humid areas like the D.R. and operating within limits but more than the normal surface loads is more suceptable to the internal wear and tear that I fear may fail prior to scheduled inspection intervals.

A composite airframe operating in a dry climate with low load factors on the flight surfaces may fly for centuries w/o fail in the same respect.

blakmax
5th Sep 2009, 23:47
You are partly correct Muduckace. It is well known that composite materials absorb moisture in humid environments and that moisture reduces the glass transition temperature, the temperature at which the resin system stops being a glassy, rigid material and becomes a soft rubbery material. This has the effect of reducing the strength and elastic modulus at high temperatures.

These effects are so well known that the reduced properties are factored into the design. There is also a requirement that the Tg is well above the maximum operating temperature. Further, apart from areas adjacent to engines, the temperature cycle seen by a civilian airliner will be such that this should not be an issue. The certification basis of the current FARs, together with the FAA Advisory Circular AC-20-107A should be adequate for composite structures.

However, the same can not be said for adhesive bonded structures because the current FARs and AC 20-107A do not prevent production of adhesive bonds which are resistant to hydration, and that hydration is accelerated by hot, humid environments.

Was the quick 787 repair bonded or fastened?

Regards

blakmax

Lt.Fubar
10th Sep 2009, 18:43
My knowledge of composite materials structures is somewhat limited to their use in small sized sail boats. Not really up to scale with airliners, so can someone explain to me what are the differences between the manufacturing technology between the Airbus, Boeing 787, and maybe Lockheed ACCA, and the Rutan's White Knight Two ?

muduckace
14th Sep 2009, 00:54
I hijacked it from another PPRUNE similar topic.

http://i205.photobucket.com/albums/bb123/tonyzimex/11.jpg

Vaccume, heat and paint in 59 minutes - I don't think so. Not to mention environmental conditions. They must be assuming 75degF and no rain.

blakmax
14th Sep 2009, 05:10
Firstly, lets deal with Lt.Fubar's question. The 787 is a predominantly fibre composite laminated construction, with much of the fuselage, wings, floor beams, empenage and control surfaces fabricated in that material. The Airbus A380 uses some of that same technology but also uses a large amount of GLARE a Glass Laminated Aluminium REinforcement material which is made by bonding thin foils of aluminium alloy (about 0.1 mm thick) interspersed with layers of fibre glass. This is used extensively in the crown of the fuselage.

Now to the repair outlined in that web page. Where do I start? The colour is not so good, but that looks like Flashbreaker 1 tape in the "coating removed" picture. That particular tape and a number of variants use a silicone adhesive which I have found transfers as a release contaminant onto surfaces to which you are about to bond. Bad karma.

Wot, not even a solvent clean? No abrasion to generate an active surface?

Next to the "Heat Pad". Let me assure you that of there is any substructure in the area a single heat blanket will not produce an adequate adhesive cure. Depending on where the control thermocouple is located (didn't mention thermocouples did they?) one of two things will result, either the thin skin will risk overheat damage of the repair adhesive over the thicker structure will not cure. You need to analyse the structure and use separate heater blankets for each structural region, one over thin sections and another over the thicker section. The issue of where to locate thermocouples is also of importance. They can not just be applied anywhere. It is important to find the hottest point under each heated zone to prevent overheat damage. However you also need to find the coldest point around the repair to provide assurance that the adhesive has been fully cured.

Just had a horrible thought; maybe they didn't mention thermocouples because they plan to use those dreaded heater blankets with the thermocouples built in!!!! These are useless. Let me demonstrate this. Place the heater blanket on a bench away from the repair site and turn it on. When it reaches temperature, measure the temperature of the skin in the repair site. Has the heater blanket made any difference? No, because the temperatures measured are those in the blanket, not at the repair. It is the temperature of the repair which will determine if the adhesive is cured or if the structure suffers heat damage. Therefore the temperature must be measured there, not in the blanket. This is even more important if these is any substructure.

With regard to vacuum, just hope that the damage is not through the panel and that there are no leaking fasteners in the area.

Next the 35 minute cure cycle. Unless the adhesive is a thermoplastic, I fail to see how this is achievable. If the adhesive is an epoxy then there is a required heat up rate and a soak period at cure temperature. Excessive heat up results in microvoiding, and most epoxy adhesives require a soak period of about one hour.

As for the overall time of 59 minutes, this is stuff that would make Tinkerbell blush. If everything was at hand and ready, and everything went perfectly, you would still need a whole bucket of "speed" to do this repair in anything like 59 minutes.

And that doesn't address the issue of envirnomental controls to reduce contamination.

You are right Muduckace, it would be humorous if they weren't serious.

Regards

blakmax

airsupport
14th Sep 2009, 07:01
AA A300-600 over long island got my interest.

Best not to mention that here, very fiery topic. :ugh:

Walnut
14th Sep 2009, 08:04
Blackmax

I have been rivited by your postings on Rotorheads re bonding. Having built a boat in the mid 80's using fast curing epoxies, and recently a kit plane from OZ using a much slower curing epoxy LC3600, just about to fly, I found this modern material much better to work with. The heat build up appeared much lower.

However my question is whether you have any theories on carbon fibre composites and how they would be affected by electrostatic discharges.
There was a helecopter crash about 10yrs ago where the tail rotor failed, it was made of composites. It passed the required strength tests, in fact was stronger than the original metal part, but appeared to fail after a strong lightning discharge past through it.

blakmax
14th Sep 2009, 09:09
Hi Walnut

Many carbon composites these days have metallic mesh bonded on the outer surface which is grounded to the structure through the fasteners. That seems to do the trick.

I'd also like to see the damage because if you followed my postings in Rotorheads, there is a possibility that there may have been other causes which were not common knowledge at the time.

Lt.Fubar
14th Sep 2009, 09:58
Walnut, is probably talking about the Bristow Flight 56C Puma that had tail rotor separation in flight. It was caused by atmospheric discharge, causing destructive failure in one of tail rotor blades at the boundary between carbon fiber and aluminium sections. Apparently difference in resistance between both materials was enough to produce a lot of heat, that destroyed the part.

There was an documentary on that, it is available on YouTube:

Part1 (http://www.youtube.com/watch?v=UEI6G6nkuSo)
Part2 (http://www.youtube.com/watch?v=R9pXIg_5PmA)
Part3 (http://www.youtube.com/watch?v=D87iyBSvs_o)
Part4 (http://www.youtube.com/watch?v=PMcN1B_483k)
Part5 (http://www.youtube.com/watch?v=rbyb5L6BXwM)

blakmax
14th Sep 2009, 10:23
Sorry Fubar. My steam driven internet system with an ultrafast modem took fifteen minutes to download about one tenth of the first video. I gave up. Thank you "3". As soon as my contract is up in a month or two, I'm outa there!

Lt.Fubar
14th Sep 2009, 12:06
The video is mostly reenactment of the event anyway. You can check the accident report here (http://www.aaib.gov.uk/cms_resources.cfm?file=/2-1997%20G-TIGK.pdf), though it's also quite heavy at 9.5MB, and have no pictures. Those are in other pdf file, here (http://www.aaib.gov.uk/cms_resources/2%2D1997%20G%2DTIGK%20Append%2Epdf) (11MB).

freespinner
21st Sep 2009, 15:31
Hi was recently down in toulouse and the airbus design guys think that the 787 delays are probably due to lightning strike/conduction issues with the composite fuselage.