I am currently looking at the "Heat Flux" possibility on T variable for a wall boundary condition. Is there any documentation and/or tutorial that shows how it is working? Is there a specific model that should be used?
Heat Flux requires a definition of W / m2 (equivalent to kg / s3) at the wall.
As with all temperature calculations you will need a relatively fine mesh against walls.
Can you be more specific? E.g., application, flow regime?
I have a box which receive a fresh air using an inlet (around 1l/s).
On the opposite side, the wall is warmed, between 1 to 10 W/cm2 (J/s/cm2).
Steady state regime.
How fine the mesh should be? Does Accuracy->High is enough or should it be more spécifique?
I presume you are using wall functions (not low-Re turbulence model), so you will want to check your Field Scalars->y+ and aim for the range 30-300 (for low-Re you will want y+~1). This may require that you set an explicit Accuracy:Custom->Resolution. High is equivalent to Resolution = 20. For more details see "Accuracy Tool".
You will need to run the simulation to generate the y+ values, from there you can get an ideal of how much to increase/decrease your resolution. y+ scales according to the local Reynolds Number, i.e., it depends on length scale, velocity and kinematic viscosity.
Many thanks, I will look at it.