Discussion Forum

I am solving a fairly simple heat conduction problem of a 2-D MEMS structure that is being heated, but different parts of the structure are in contact with different fluids of varying temperatures. I want to set the MEMS structure at room temperature (300 K) and solve the system where I can look at the temperature of the structure (surface temperature plot) at different times. It seems to solve just fine, but I do not see any time-varying temperature changes in the structure - at time = 0 the surface temperature plot is exactly the same as t=1, or t=10, or t=100.

I previously solved this problem with COMSOL 3.2 with good results where I see the structure heat up at different locations over the same time span, but using 3.5, I don't see any change.

I have tried using the different solvers (time-dependent SPOOLES or UFMPACK) and increased the meshing resolution and used time steps of 0.01, 0.1, and 1, and everything seems to give the same result.


Secondary to this question, In COMSOL 3.5, in the Physics>Subdomain Settings, for a given subdomain element of some material, I can specify the T_ext, T_amb, and also a T_initial (at t0). In COMSOL 3.2, I could only set a T_initial which I had set to 300 K and it seemed to get good result. For 3.5, what is the difference b/w T_ext and T_initial?