Usage of composite materials above Tg

[CompoMan]

Has anyone has any experience in using composite material above its Tg. If one could characterise its mechanical properties (at temperature >Tg and ignoring non linear failure behaviour), can one use them for FE analysis i.e. does the FE result applicable or appropriate??
n/b: the component will not be reused once it is being subjected to the temperature

 

[Compositepro]

Many plastics are used above their Tg where flexibility and toughness are desired. The plastic in garden hose is used above it's Tg. When it get cold enough to be below its Tg the hose is stiff. In structural applications you do not want materials to creep and permanently deform so in aerospace the rough rule of thumb on allowable service temperature is 50C below Tg. Above Tg the modulus (stiffness) of the resin is greatly reduced. This is reverseable if the resin is chemically stable at these temperatures.

Many thermoset composites can be slightly thermoformed at temperatures above Tg. Shape-memory composites use this effect.

 

[MikeHalloran]

I've seen a pair of marine mufflers made of the best quality high temperature epoxy and fiberglass. When their water supply became inadequate, the material got soft and distorted under its own weight, like an overheated thermoplastic would. Upon cooling, they regained a lot of strength, but they were no longer the intended shape.

I suggest that if you 'get away with' using a composite above Tg, you're working in someone else's safety factor, which doesn't sound like a real good idea to me.



Mike Halloran
Pembroke Pines, FL, USA

 

[Compositepro]

That rule of thumb is 50F rather than 50C, I believe.

 

[CompoMan]

Thanks for the feedback gents, but what if I have to use the composite above its Tg, does the FE analysis result which generated from non linear test data is still appropriate?? may be I should look into FE analysis on elastomeric materials??

 

[wes616]

If you are using the material above it's Tg, you should perform extensive testing of the material at the expected service temperature. My experience is that the behaivior is incredibly plastic and unpredictable. Any FEA analysis IMHO would not be valid until the characteristics of the material is studied.

Wes C.
------------------------------
No trees were killed in the sending of this message, but a large number of electrons were terribly inconvenienced.

 

[Demon3]

FEA will not work because the behaviour of the material above Tg is bit time-dependent and viscous rather than elastic. You would have to do a lot of work to characterize the time dependent behaviour. It can be done.

On a more practical note and assuming your question is regarding an application and not just theoretical, I would recommend you look for a material that is below its Tg and therefore capable of bearing load.

As an example, there are special epoxies that retain almost full modulus above Tg and all the way up to 300C.

Chris DeArmitt

"Knowledge has no value except that which can be gained from its application toward some worthwhile end."
Think and Grow Rich - Napoleon Hill

 

[RPstress]

Although thermosets (like epoxy) above their Tg are often described as "rubbery," they do not spring back to shape. They deform permanently rather easily.

The properties drop off rapidly within a few degrees of the nominal dry Tg; 10 or 20 deg C (depends on how the Tg is measured) sees the modulus well below 10%, and heading for 1%, and any significant deformation is likely to be permanent. A laminate whose matrix is above its Tg is quite likely to deform under its own weight, as MikeHalloran describes, and its deformation is time dependent as Demon3 says.

If you must try to predict its behaviour then you need to do some fairly serious material-nonlinear work, and you're quite likely to get it wrong anyway.

If it's a one-time thing as part of manufacture, then try to arrange for tooling to support the part. If it's a rare service case then that's tricky... What will it cost/break if it's bad?

You can, of course, suck it and see with a test. That might be cheaper than analysis. If it's a service case than the laminate will probably need to be wet, which lowers the Tg a lot, and complicates testing even more.

 

[CompoMan]

Thanks for all your contributions. I guess the general opinions are analysis is difficult and inaccurate and best is to carry out appropriate testing.
Thanks again

 

[MikeHalloran]

I think the consensus is that you should use metal.



Mike Halloran
Pembroke Pines, FL, USA