Composite Shells

[vtnick]

Is there a way currently to model composite shells? By composite I just mean in general a shell representing bonded layers of differing materials. From what I can tell there seems to be 2 missing components:

1. ability to specify layers in the shell and their thicknesses
2. non-isotropic shell material properties

[GiudGiud]

@bwspenc do you know?
Maybe there is some TRISO work that's applicable?

[bwspenc]

You're correct that we don't have the capability for composite shells. We have the ability to specify the number of integration points through the thickness, which could be used as a starting point for that, but we don't have the ability to tie those points to layer thicknesses or properties that differ by layer.

We've done some work on allowing arbitrary constitutive models for shells, which would address your second point. We've been working on that but never got this finalized. Here's my branch, which is pretty out of date: https://github.com/bwspenc/moose/tree/shell_new_tm
If I remember correctly, we got it to the point where we rotated the material to the shell coordinate system, which should basically do what you need for a non-isotropic property.

[GiudGiud]

but we don't have the ability to tie those points to layer thicknesses or properties that differ by layer.

could you use subdomains to differentiate the layers?

If I remember correctly, we got it to the point where we rotated the material to the shell coordinate system

This would be nice to get in MOOSE as an example. Someone was asking for this at the MOOSE training this January

[bwspenc]

This is something that's usually handled within a single element, so you can't define the layers as subdomains, at least with standard implementations of that. Maybe it could be done with overlapping elements for each layer, but I think that has a lot of issues.

I agree that it would be good to have examples of this once we get this capability more developed.

[vtnick]

Thanks Ben.