cant get convection setup to work

[bptsj]

Hi All,
I am having difficulty getting thermal convection setup to run in the GUI. Currently my model has 18 bodies and 1 heat equation plus a boundary condition (Dirichlet) on one surface. This is working as expected with a 30C rise. I want to add a heat sink to the surface that now has the Dirichlet. In that Boundary Condition setup I removed the Dirichlet condition and entered a Heat Transfer coefficient and an External temperature for the surface in question. Where before the BC was 300K I now have an external Temperature of 250C. The problem does converge but the whole body is about 10000K. I expect a 20-40C rise above the external temperature.

I have searched the forum and read the documents trying what was suggested but I'm obviously missing something. I've attached the sif file and run log.

Thanks for your help,
Brian

[bptsj]

My results might not be unexpected I am using a heat transfer coefficient of 12 W/m^2-K. If I do a simple calculation of the applied 3W transferring through the small area defined I get ~8000K.

Question, is the heat transfer coefficient with external temperature the correct setup to use for a solid conducting heat to a heat-sink surface?

Thanks,
Brian

[mzenker]

Hi,

if you want to model heat transfer to ambient air, then this is certainly the simplest approach. To find the suitable heat transfer coefficient, you might have to consult engineering tables for heat transfer.
If you want to model your special setup in detail, you might try to model convection with the Navier-Stokes solver (see models manual). I don't have any experience here, but there are a certain number of posts in this forum on that subject.

HTH,

Matthias

[campo]

I also think is correct to give temperature and heat transfer coefficient, but if you get so high temperature the coefficient must be wrong because it is dependent on temperature and must be higher at that level. You can get simple coefficients on http://www.thermal-wizard.com/tmwiz/default.htm.