code_saturne User's Forum

Hi,

I am currently running the modelling of the air diffusion for a hotel lobby. So far the aero calculation is converging. I have also added a wall heat flux boundary condition.

The next step is to enable the radiative module and to add the proper radiative boundary condition through the large glazing wall (1000m2). This boundary condition is defined as :

radiative heat transfer through transparent media on Autodesk CFD
Semi-Transparent Wall Boundary Conditions [for the DO Model on Fluent.]

A beam direction is to be defined that corresponds to the solar angle.

Can such boundary condition be modelled in Code Saturne? Standard radiative boundary conditions in the code assume omnidirectionnal beams.

Thanks for your help.

Hello,

I don't believe this can be solved using only user subroutines. Checking files in src/rayt, raycll.f90 and raydom.f90 might be modified for this purpose.

Otherwise, as the hotel lobby is closer to a transparent medium, assuming heat is transferred to surfaces first, then diffused and convected, you might have more luck using the transparent radiative module from Syrthes.

This means you would also need to model the window and walls enclosing the domain with a solid volume "heat transer" mesh, in which you would model heat exchange, and use conjugate heat transfer (Code_Saturne/Syrthes coupling). Code_Saturne/Syrthes coupling must be done through a solid mesh (we cannot use just the radiative module of Syrthes), so this moduling is a bit more involved, but is probably a better fit for your case (the Syrthes documentation explicitly mentions solar radiation, for example)

Regards,

Yvan

Hi,

Thanks for your reply.

Even though having a semi-transparent surface boundary radiation would be a very nice feature in Code Saturne (very useful for Thermal comfort modelling), I am pretty sure I will be lost if I start to modifying the f90 files. So I will give a try to Syrthes.

I have attached a picture of the model (24 jet diffusers on top, curved glazing wall, exhaust on the back of the room).

1. Do I really have to extrude a solid volume from each boundary surface? If yes, would a thin wrap around the fluid model (2/3 cells and maybe 10cm thick) be sufficient? I have no particular interest in studying conduction effect in the building structure.

2. The radiation calculation will be done on syrthes and therefore I shall deactivate the radiation module in Saturne. Am I correct?

3. I read that Syrthes does not handle opaque surface. Therefore I will have to alter the meteo file to simulate the shading coefficient of 0.25 (meaning only 25% of the radiation heat flux is transmitted). Side question : is there a way to model the greenhouse effect?

Thx

Hello,

Here is a short answer :

1) The solution you suggest seems OK. Assuming the building materials have pretty low conductivity, and heat is transferred mainly through convection with the fluid, this seems reasonable (even with only a few cells, you can do 2 runs with significantly different conductivities and/or specific heat values in the solid to check there is not a big sensitivity to this aspect).

2) You are correct.

3) I'm not sure if Syrthes has a specific way of handling the greenhouse effect (and the Syrthes developers are out of office this week, so I can't ask). If not, I would assume the lower the specific heat you assign to the glass, the less heat can be absorbed through it, so this might approximate (in a crude manner) the greenhouse effect. Just a quick idea, so I'll let you check if this seems correct physical modeling to you.

Regards,

Yvan