Re: Simulation of Thermal Expansion. [Please help...I am so frustrated]
Lawrence,
I took a look at your image, and it appears that you could apply symmetry, using just 1/8 of the model (if it's not symmetrical, then this won't work). You have identified simulation with shells. Could you not use volume regions for the copper fills and assign the appropriate material properties and run the model as a solid?
The below images is not your model, but the concept works when applying symmetry to three sides. You will not get singularities, because the three sides that have symmetry are not locked, but they slide along the plane of symmetry. The 1/8 aluminum cube has a 500 degree F thermal load, and you can see the global magnitude displacement for the model, as shown below. You could plot the displacement in X, Y, Z alone. For example, the Y displacement is shown below. So the total displacement in Y would be double of what you see 2(0.032") = 0.064".
If your model is not proprietary, could you upload it? This is an interesting problem to work on.

Figure 1: Symmetry Constraints

Figure 2: Global Displacement

Figure 3: Y Displacement
