Spurious results in thermal analysis with SimSolid

Mike_20941

I have a large model of a complex structure. "Automatic connections" found 3000 connections among a large set of bolts, nuts, washers and thin structural parts. Prescribing two different temperatures T_min and T_max to different parts of the structure and nothing else, i.e. leaving most parts without any boundary condition, I get spurious and unreasonable temperature distribution results. For instance, the maximum resulting temperature is as large as 2*T_max although no heat source is present.

Increasing "Connection resolution level" to "high" did not help. Neither did choosing "Solution settings" with checked "Adapt to features" and "Adapt to thin solids" as well as "high" "Refinement level" and 8 adaptive solutions.

The problem seems to worsen as seam welds are applied between metal plates.

1. Are there any guidelines as for how to set up the analysis properly? I have followed the documentation and the tutorials.
2. What happens to parts with no prescribed boundary condition? Is is the same as in the FEM, i.e. zero Neuman conditions?

Chayan

Hi Mike, it is difficult to say without knowing more about the model and the analysis you are trying to do. But from what you have described, it seems like you might be having a trivial solution.

My suggestion would be to go thru a few debug steps:

1. Remove unnecessary parts from your model.
2. Ground/constrain the model
3. Run a normal modes analysis.

You should only see structural modes in your model (non zero frequencies). If you see rigid body modes, then you need to look at those regions, and define them properly.

Hope this helps.