Hi RPstress,

I don't quite understand why you would knock down the ILSS allowable between a 0deg and a multiaxial laminate... This property is linked to the resin properties. i.e. the resin properties don't change whether your fibres are at an angle to the load or not.

Now, on the other hand, the interlaminar shear stress in a laminate will be different for a given load between a UD laminate and a multiaxial. Say for example you have 2 laminates (1 UD and 1 multi) that you pull in tension. In the UD case, since all the fibres are in the same direction, they will take the same load and strain and therefore will have the same stress... this results in theory in no interlaminar shear stress.

In the case of the multiaxial one, fibres that are off-axis will take less load that the 0deg fibres. Each orientation will take a different load. Therefore, the strain being equal for all plies, the stress will be different. All this means that the will be a certain degree of interlaminar shear stress between adjacent layers which have a different orientation.

From this you can then understand that the bigger the difference in orientation between adjacent plies, the bigger the interlaminar stress. Generally it is recommended to minimise as much as possible this angle. For example if you use standard 0deg/+-45deg/90deg type of laminates, you should allow only 45deg difference between 2 plies. Obviously this is not always possible.

May I ask what you want to use the ILSS prop for? What type of calculation?

Anybody can confirm my line of thought?

Anyway, I hope this helps!

Cheers,
SkyD.

There are several reasons for the difference in interlaminar shear results between unidirectional and multidirectional laminates. First, unidirectional laminates do not have nearly as much stored interlaminar residual stresses due to the differences in shrinkage and CTE between the the zero and 90 degree direction within each ply. These stresses are proportional to ply thickness and resin modulus. Second, the fiber plies in unidirectional laminates can intermingle creating some fiber reinforcement in this property. This is much less likely to be a factor in cross-plied laminates. Third, none of the test methods are perfect in creating only pure shear stresses. Different methods give different values. This is why new test methods are still being developed. I belive this third reason is what SkyD was refering to.

Thanks guys.

I hadn't thought about the built-in residual stresses.

I have anecdotal evidence that you can see fibre-bridging in an all-0 degree UD ILSS test (not seen one done myself), so it seems quite likely that most of the difference comes from these two effects (intermingling and residual thermal stuff). Our quasi-isotropic layup seems to have ILSS of about 60% that of an all-0 degree laminate. We felt it was a bit of a large difference. On the other hand we've got woven and NCF materials with ILSS's similar to the QI.

SkyD: the ILSS doesn't often size anything, but we always include a check on interlaminar shear stresses coming out of FE analysis. It can also be an issue where point loads are being introduced to a laminate over a very small area. Finally, it is sometimes used as a way to calculate fastener pull-through (though in this role it is usually used only to extrapolate pull-through test data to laminate thicknesses or fastener sizes different from those tested).

-RP.

RPStress,

I have one piece of test data that says uni SBS specimens produced 17.8 ksi and a [06,45,-45,0]s (that's 6 plys of 0's on the outside) produced 15.0 ksi.  

I believe free edge effects may start a shear failure.  In a transverse loading condition I try to keep the off axis plys away from the midplane because this is where the max shear stress is.  This may not be applicable to your loading condition, but it is certianly applicable to a SBS test.

Interesting. Not as big a reduction as we've (maybe) been seeing. On the other hand, you had a couple of 0s on the centerline. It looks like we had two 90s on the centerline. (We had QI ([+45/0/-45/90]n)S.)

Also, that strength is about 20% higher than the all-0 degree results we've been given. Was that a fairly normal epoxy? (Our resin is a low density one for out-of-autoclave cure.)

Does anyone know of fiber volume effects on this property? (We're not expecting any at our Vf of around mid-fifties to low sixties.)

It's Hexcel 8552.  The fiber/sizing is important too, and in today's marketplace, getting what you want can be challenging to say the least.

I'm not sure about the effect of fiber volume, but void content has a big effect.  For best results, degas every 1/2 ply :).