<<Planning what happens beyond that is less interesting to me, Although it would be good to know what approaching failure and failure look like.>>

  You might want to plan on a failure where the foam goes first, which might give you some warning when that boat is stressed way past its design goals.

  For reinforced concrete, you want the steel to fail, because it will stretch in tension and the structure will deform.  If you put in so much steel that the concrete matrix fails in compression, the failure is instantaneous and explosive.

  The key is a "safe" failure that gives you time to react.

  So it might make sense to do a 3D FEI analysis based upon the materials being used.

  You'd want a failure that's close to the glass, so you're not wasting material, but goes before the glass, so there's at least a tiny warning envelope.

  It wouldn't be as dramatic and clear-cut as with concrete, give nature of the skinned composite, but the goal would be to be able to back off when you hear a sickening noise, look at a deformed component, and then limp home. 

        - Mike

 

| One interesting point is the modulus of elasticity. That is something I was missing in the simple calculations of sandwich panels which I made.
| The core thickness and shear modulus has to be somehow linked to the compressive/tensile modulus and thickness of the skins, right?

I do not see why. This condition is relevant only in overload, right?

| I think that during bending, ideally, the elongation of the skins at failure should align with the shear deformation of the core at failure. If not, either of the materials will not be fully utilized.

You are talking overload conditions, right? We need to remember the composite, I think, should be fine under expected loads, plus some safety factor.

Planning what happens beyond that is less interesting to me, Although it would be good to know what approaching failure and failure look like. Like, cracking gelcoat was interesting. I wonder where house paint would crack as we climb the safety factor to yield.

I feel less interested because I think we should optimize composites around cheap, light, and low labour, not around failure, and failing 'properly'. I don't care if the skin or the core fails first. I care it does not happen until I am nearly past my safety factor. Perhaps I care that I expect one to fail first, and when failure happens like I expect, I feel smart and that I understand composites well. But then, I imagine, that might be the least important thing on my mind in that moment my boat makes a new noise.

The hexcel sheet, as I see it, allows an understanding of composite panels to some first order. I think that is enough for me. I can move on to building a boat cheaper...

Or did I misunderstand you? Why should I care how core thickness and shear modulus are linked to the compressive/tensile modulus and thickness of the skins?