freefall |
November 18, 2013 09:12 |
Natural convection from heatsink
Hello,
I'm currently simulation natural convection in 3-D from a plate-finned heatsink with multiple heatgenerators.
I use pressure-based solver, with gravity.
- k-epsilon turbulence, with scalable wall fc and full bouancy effects.
- P1, with emissivity:0,85 on most solids.
i use Simplec with discretizations:
- standard pressure
- Second order on the others
as materials there is only air and aluminium
Questions regarding air properties:
- How do i best define density for air here? I have used mostly boussinesq approx. Here i wonder what sould i define operating temerature as? (in operating conditions) Should it be my ambient air temperature, or should it be the same temperature as the boussinesq density is defined (in materials) I have also tried incompressible ideal gas with a defined density (in operating conditions) With this it seems to get alot hotter, about 10 degrees...
- An is it possible/meaningful to use piecewise polynomial to define thermal conductivity, Cp and viscosity and still get convergence? Or should i define them as constants?
|