Test methods & laminate definitions to measure CTE of composite lamina

[tapirg]

I’m trying to standardize a method to characterise CTE values for the composite materials we currently use. (to lamina level longitudinal and Transverse)
I’ve look into some literature, forums but I haven’t found much information about it.


Basically I’m looking to test a wide range of fabrics combined with two or three different resin systems. (I.e. UD carbon, weave 2x2Twill carbon, Biaxial carbon ±45º, Biaxial Glass and a 0º/±45 Triaxial glass)

I’ve identify two possible methods: Interferometer (ASTM E289) or Thermo mechanical analysis (ASTM 831); I’m also considering to use a non-contact video extensometer
,for conduction my coupon testing but I’m struggling to come with an adequate laminate definition to perform the test. (I.e. Thickness of the specimens, stack up sequence) and obtain meaningful information from them. (Lamina Longitudinal and Transverse CTE)

I think that for the UD it’s clear that the laminate has to consist of a Nr of layers all orientated at 0º and the same for the Carbon Weave [0º/90º]ns. And again similar for the Biaxial carbon [±45]ns (as we have both fabric configurations plus/ minus and minus/plus)

The real issue comes whit the laminate sequence for the Biax Glass ±45 and especially for the Triaxial Glass 0º/±45 as I don’t have a clue about how to stack up the plies without introducing out of plane bending due to the asymmetry of the laminate.

Hope all this makes sense.
Any Help would be greatly appreciated.

 

[RPstress]

Still no reply after four days.

I've never done, or had done, any CTE work personally. Also, I don't have any familiarity to speak of with triaxial woven material (we have used a bit in a braided element of a large shell, but never had any thermal data for it). However, I can give some commonsensical advice on layups that won't warp much. Hope it's not too basic; a little difficult to tell from your post what you really want to know.

For the woven material, even if it is twill or plain weave, it is best to reflect the side of the cloth you put down either side of the laminatge centreline. I.e., weft-side down below the CL, warp-side down above. Also have an even number of plies. For harness satin weaves this is mandatory; a ply of HS has cloth many similarities with a combination of ply of two plies of UD, one at 0 and one at 90.

Also make sure for woven that 0 and 90 and +/-45 don't get confused. The differences between warp and weft may be small but they are usually there in small degree. A ply of +45/-45 woven material weft side down is sometimes interpreted as a ply of -45/+45 material warp side down.

For the triaxial the same advice goes. Reflect the side laid down about CL, and if you must have plies at both 0 and 90 make sure of consistency.

The total thickness is not too important unless through-thickness CTE is needed. A decently symmetric laminate should be thick enough.

You seem to want some sort of basic material axis properties. You may well find that laminates more similar to the actual proportions of 0/90/+-45 in your layups of interest give better results. Using classical laminate theory for predicting a laminate's CTEs using basic material CTEs is a start of sorts, but is not likely to be as good as measurements for a more representative layup.

The usual assumption is that the basic material CTEs and indeed stiffnesses are orthotropic; that is, a ply of basic material has a CTE-12 (shear) of zero. This should be so. If it is not, something pretty funny is going on. However, especially for carbon, the measurements are so small that apparently anomolous readings can result.

Sorry not to be more help. Maybe someone with more relevant experience just hasn't read your post yet.

 

[CompositeGirl]

As someone who has done a large handful of composite testing, I definitely prefer ASTM E831 over E289.

Two main reasons:
- Less prone to surface conditions as for interferometers use light.
- laboratory equipment is more readily available as you could use the same equipment used to measure Tg

It's easy enough to have all the plies oriented in the 0° for unitape, but as you have stated, it's not that easy for woven fabric.

I have tested varying lay-up, thicknesses in both 0°, 90° and Z-direction on quite a bit of different fabric/resin combinations. Our stress guys seem to only really be worried about thru-thickness CTE (Z-dir).

In fact, our Z-direction CTE is significantly larger than any in-plane CTE.

My suggestions:
- Use ASTM E831
- Do limited testing to verify (and convince to yourself) that Z-direction is the bigger issue and that there is little sensitivity to CTE differences inplane regardless of lay-up.