Composite Fretting 3

[ICbrokethings]

I'm conducting an MSG-3 Analysis on a largly composite aircraft but I'm not too sure if carbon/fibre composites suffer from a type of fretting corrosion or not.
Anyone?

 

[SWComposites]

Composites are susceptible to lots of damage types, but corrosion is generally not one of them.

 

[Compositepro]

Composites are sensitive to abrasion due to fretting. Carbon fiber composite can suffer from galvanic corrosion when in contact with metals. Its actually the metal that corrodes. Some resins used in composites can also degrade or "corrode" due to caustic (low pH) corrosion products of aluminum. Polyimides, cyanates and BMI resins do this.

 

[ICbrokethings]

Thanks for the replies. I appreciate that 'corrosion' is a misleading term when applied to composites. What I'm really asking is, would you expect to see a 3000 FH (or so) check for degredation in similar locations to those you would expect to see fretting on metallic components?

 

[SWComposites]

Maybe. What is the specific location, configuration and materials of interest? As mentioned above, composites are susceptible to wear in joints which have movement.

 

[tbuelna]

ICbrokethings,

Fretting damage at joint contact interfaces is generally a difficult condition to design against, since it is not totally understood. The commonly accepted theory of fretting damage is that it is due to the repeated diffusion bonding and shearing of the local micro-level surface asperity points that contact under load in a fastened metallic joint. This repeated bonding/shearing action results in material transfer from one surface to the other, and thus produces the classic surface pitting failures typical of fretting damage.

In order to produce fretting, the faying joint surfaces must have some relative motion. But since fretting occurs at the micro level, the relative motion required to produce fretting may only need to be a few microinches or less.

Most composite-to-composite joints are adhesively bonded, so I don't believe it would be possible for fretting to occur in them, since the bonded joint faying surfaces would not be subject to relative motion.

A fastened (bolted or riveted) composite-to-composite joint would be unusual, but I would suspect that the faying surfaces would still be resistant to fretting. Since the composite's supporting resin matrix is not susceptible to the effects of diffusion bonding like metallics are, which is the root cause of the fretting mechanism.

However, if what you are more correctly concerned about is abrasion damage (as opposed to true fretting damage) between two sliding composite surfaces, the carbon fiber debris generated and trapped between the rubbing surfaces would be extremely abrasive in nature.

The best way to prevent fretting in any joint, whether metallic or even composite, is to make sure the joint is fastened or bonded adequately such that there is never any possibility of relative movement between the mating surface, under any foreseeable loading or thermal conditions. With a bolted joint, this usually means a whole bunch of fasteners.

I design aircraft mechanical and propulsion systems, and any bolted shaft joint that transmits torque is always designed to show that it is capable of transmitting that torque purely through friction at the bolted interface based on a very conservative coeff. of friction (ie. <0.15). That's the only proven way to prevent fretting.

Best of luck.
Terry

 

[SWComposites]

Ah, no, the statements "Most composite-to-composite joints are adhesively bonded," and "A fastened (bolted or riveted) composite-to-composite joint would be unusual" are not correct. There are many thousands of bolted composite-composite joints flying. Such joints exist in many locations in commercial aircraft (737, 747, 777, 787, various Airbus models) and military aircraft (F18, F22, F35). Structural joints generally do not consider friction load transfer in their design or analysis. This may be done for some machinery type joints.

 

[ICbrokethings]

Thanks guys. So I think the answer then is no, in static joints there should be no fretting wear.

 

[lillidelft]

The answer is no, there should no fretting, because fretting is caused by cyclic loads.