Thermal Analysis Heat Flux Through Walls

[Rycosso]

I am modeling walls to test thermal performance, and I have found that as my model gets more complex the difference between the interior and exterior heat flux gets larger. I applied a film coefficient boundary condition on the interior and exterior of the wall, and on all the edges I put a heat flux boundary 0.

 

At first I thought it was meshing, but I still had the issue when I made the mesh as small as the software allows. The problem also goes away if I apply the same material to all layers of the system.

Why am I getting a different heat flux on the interior and exterior of my model? 

[Royce_adsk]

Try running with tight convergence with ADV5. Let us know if that helps!

-Royce

[Rycosso]

I turned the convergence to the tightest that the slider would go, and changed the Advection to ADV5, and my results did not change at all.

[apolo_vanderberg]

Rycosso,

  As your model gets more complex, be sure that the mesh quality through the thickness (of each material) remains at a similar mesh quality. It is possible that your more simplistic models had a better quality mesh through the material thicknesses such that we were able to maintain an energy balance more easily.

There is no need to assign a Heat Flux of 0W on the sides/edges, we automatically assume that an exterior wall without any thermal boundary condition will be treated as adiabatic.

Apolo

[Rycosso]

Thank you Apolo for the suggestion. I have reduced the mesh in the model to the lowest the software will allow, and the change in the result is negligible. I have also tried the adaptive mesh options, and they do not resolve the issue either.

Thank you for the Adiabatic tip, that will help speed up the model making process.

[Rycosso]

I have even tried the simple wall below to show that meshing could not be the issue. I made this super simplified wall model with 1" thick studs with the settings you described above, and I am still getting different heat fluxes on the interior and exterior.

[Royce_adsk]

That is a 3.6% error which is typically regarded as a reasonable error.

-Royce

[megmac]

Hi there.

Please see my post here: http://forums.autodesk.com/t5/Simulation-CFD/Wall-Calculator-Heat-Flux-in-AutoDesk-Simulation-CFD/m-...

I believe I am having a similar problem (i.e., interior surface fluxes do not equal boundary fluxes at steady state).

Did you ever find a resolution to this problem?

Thanks!