[Abhya]

Hi

I'm dealing with a solid to solid heat transfer problem in which solids of varying thickness (range of thickness = 0.009m-0.4m)

I have two questions regarding the same:
1) How many minimum number of cells should fit between the least thick solid
- this in-turn governs the size of element throughout the model

2) When is it advisable to opt for Thermal shells - is it possible to use them even when there are two different solids on either side of the solid being modeled as a thermal shell

Thanks in advance

[LuckyTran]

1) If the temperature distribution across the thickness is linear (which implies negligible lateral conduction), you can get away with as few as 4 cells, but if it is parabolic or other complex profile then you'll need more. Try this as a starting estimate.

2) The shell conduction model is highly appropriate where the thickness of the components is small compared to other parts of the domain (or the cell sizes in the thin solid are much smaller than adjacent cells in other regions). The shell conduction model still solves the heat diffusion equation using virtual cells rather than explicit cells.

edit* I want to add that the shell conduction model really shines when it is not feasible to mesh the thin regions directly. The virtual cells can be made arbitrarily small by setting increasing the number of layers.