Safety of composite airframes

According to this report from New Scientist, Boeing's composite airframe for the 787 has been showing a good bit of weakness lately.

How safe do you think the 787 will be? And do you think it's really possible to know, at this point, how well the airframe will hold up over a long period of time?

 

Log in or Sign up to hide all adverts.

Unfortunately the article doesn't quite say whether it's a problem caused by the composites themselves or the actual design (e.g. load paths or something similar)*.

Probably no less safe than any other aircraft of a similar class.

Testing isn't just (as seems to be implied by that article) done by computer simulation, most aircraft go through full-scale (actual airframe) fatigue-testing which runs through a lifetime's worth of load cycles in a relatively short period.
And safety standards are such that NDT is done fairly regularly: maybe more so on a new-design/ new materials in service aircraft than an established type.
Composites have been used before (maybe not quite so extensively - I'm not a fan of self-loading luggage extrusions, more military aviation, so I don't know how much more, if at all, composites are used on the 787 compared to combat aircraft), and the number of composite-using military types don't seem to be suffering significantly more than traditional "tin" types.

* I just read some of the further links from that article and appears that Boeing (foolishly in my opinion) relied too much on simulation for the initial work and not enough on proof-of-concept testing.
C'est la vie...

 

Log in or Sign up to hide all adverts.

There are no gremlins lurking in composite structural technology: It's mature.

In compression, strength to weight is not so revolutionary. But it is a step ahead in every other aspect- strength/weight in tension and in sheer, fatigue life, weathering/oxidation, thermal properties, production $ maintenance savings, and more. Many more advantages than surprises. I don't like really big passenger airplanes though, because of the time required to embark/disembark people. Big herds are a big routine nuisance. But much worse, getting out in dire emergency gets a little too animalistic with big confined panicked crowds. The accidents although rare will have bigger numbers for the headlines- a challenge to rational perception. The 787's efficiency gains are mostly hype- same engines on a comparable or older airframe would do the same or better.

 

Log in or Sign up to hide all adverts.

There is a catch to using composites that tends to trip people up...

We'll look at impact damage alone here...

Let's say you have a metal box, and you smack the side with a hammer and make a dent. There it is, there's a dent, it's damaged. If there's no dent, it's good to go. Structurally, it's that simple for a ding.

Now, let's take an equivalent composite box designed to do the same thing. Smack the side with a hammer. It might deflect and come back and look perfectly okay. You might've severely damaged a group of fibers within the composite and hence severely weakened it, and it can still look perfectly fine. You clear the box as just fine, then it fails when you put a weight on top of it, and everyone wonders what happened.

Composites are much more tricky structures to do damage diagnostics. At least with metal, if there are internal micro-fractures, an eddy-current test can be done to reveal those cracks. Can't do that with non-metallic composite structures, since you need metal to do that sort of a test.

Composites are great, and can be used wonderfully, but "the carbon god is a false god", as I once heard, and composites also have their inherit shortfalls. In this case, the more complex, non-homogeneous material is going to do strange things when damaged, and alot of that fatigue we don't know about till the part in question fails completely.

 

Nah, acoustical testing tells everything about composite structures that visual inspections of metal ones ever did- and then some. It's also easier to embed sensors and networks to warn of any anomaly. Next-gen composite structures may repair their own minor aging and damage.

 

The best way to deal with structural integrity would be either to integrate nanotechnology in the form of processor nodes throughout the composite structure much like a neural net array, in this arrays messages would be sent much like packets over TCP/IP with a destination in mind and a TTL and the capability of passing each notes MAC address along (If a nodes down it would identify it was unreachable and be potentially damaged) or a cheaper less effective version would be to embed fibre optic threads throughout it. The idea being that fibre optics has a shear factor, should the cable become severed it could be registered that the framework has been impaired in some way and should be grounded for maintenance.

 

Hypewaders, I'm probably mistaken, but I thought acoustical testing depended very heavily on a good understanding of the part in question. By that I mean that testing an individual composite part with a known shape would work, but testing it while it's bolted to a plane with countless other parts resonating in their own ways would add uncertainty to the problem?

I really don't know, it'd be sweet if you could fill us in. Thanks!

 

You can use a thermoplastic matrix, like polyurethane.

Using epoxy is like using an alkaline battery.

 

Being a simulation/CAE engineer by profession, you may find it strange that I agree with you.

Recently, even Toyota admitted that their recent slide in quality was related to an over-reliance on simulation.

There is no way that complex aerospace and automotive structures can be qualified strictly with simulation today - it will be decades before we approach that level of capability with structural analysis.

 

By taking acoustical scans of the entire airframe (or at least key componenets) immediately prior to delivery, that "snapshot" can be compared to later scans to note variances.