Thermal steady state

GermanGerman
I think something is wrong with Termal steady state calculations.
Look at this very simple model. It is a cube with heat going in through a face of constant high temperature, and going out by the center of the oposite face, at cold constant temperature.
The rest of the surfaces are isolated.
Surface normal integral of heat flux on the hot surface should be exactly the oposit of surface normal integral of heat flux on the cold surface, but it is not.
Surface normal integral of heat flux on the isolated faces should be zero, but it is not.
Surface normal integral of heat flux on all faces should be zero, but it is not.
Surface integrals of heat flux Z on In and Out surfaces give the same results as Surface normal integrals of heat flux.


Where is the problem?

Comments

  • disladisla
    Hi German,

    Thanks for all this years of Manual's translation.
    Could you post that simple model. I can't figure what's the problem. ?

    Regards
  • VictorVictor
    Hello @German I think I've reproduced the problem you described. It looks like discretization error. Even on the isolated faces, there is often a component of heat flux normal to the surface which skews the solution with a coarse mesh but reduces with mesh refinement.

    I think the effect is worsened by the hard boundary of the cold patch so that most of the heat exits around the perimeter where edge effects are stronger.

    Let me know if this doesn't address the problem you found.



  • disladisla
    edited October 16
    It helps to extrude a thin layer of the exit window. Thermal reaction flux is better extracted. T_BC should be applied to the new outside surface.


  • GermanGerman
    edited October 16
    Thank you both for your answers.
    The difference extruding cold faces is awsome.
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In Register