[Kelevraa]

Hey all,

I'm about to do some simulations regarding thermal comfort in buildings.
I want to include radiators and solar radiation in general. As far as it concerns the transparent or semi-transparent (air/glass) mediums it looks like the radiation works somewhat properly (even though I don’t know the best mesh-histories-ratio yet).

The problem lies in the configuration of the incident solar radiation on the outer walls of the building.
I want to set the solar radiation as a directional radiation flux as a source on the boundaries. To do so, I have to allow radiation in the solid and therefore have to set the radiation properties in the used materials and I have no idea what to set there (absorption coefficient as an example) because in general I assume the solids to be opaque.

So the question is: How is it possible to apply solar radiation on the outer walls of the building? (what is the best way to do so?)
Do I have to generate a fluid outside the building?

Best regards

[ghorrocks]

I am no expert on radiation modelling but I will give it a shot....

If you want the solar radiation to enter your domain as radiation:
You do not need a fluid domain outside. You should be able to apply a directional radiation source to the walls on the outside of the solid domains. But this will only work for Monte Carlo models, you cannot do this with the discrete transfer model.

If you want the solar radiation to not enter your domain, but act as a heat source on the outside of your domain:
Then you just need to apply a heat flux to the outer walls of the domain.

[Kelevraa]

Hi glenn, thanks for the fast answer.

The problem in applying the solar radiation as a heat flux is to neglect the directional components or rather the distribution of the heat flux.
Therefore (I think) I have to apply it as a directional source. But when I do that, the first element has the temperature given by the boundary condition (Heat transfer coefficient with outer temperature) and the second element is the one which gets heated by the incident radiation. From that element on the radiation "travels" through the domain depending on the absorption coefficient set for the material.

Normally the radiation would hit the outer wall, get absorbed or reflected (but would nearly not transmit it) and the heat transfer would be based on the temperature gradient between the inside and outside of the building.
So I am looking for a way to apply the solar radiation on the outer walls in a way it considers the directional distribution. Through windows I do it with the Monte Carlo model and a directional heat flux source which works fine. But I do not see a way for the opaque solids but adding an outer fluid (what I don’t want cause of additional simulation time). Am I missing something important?

Any help is appreciated

[Martin_Sz]

You can try inner module of Ansys in Workbench - more simpler
Static Thermal or Transient thermal

[ghorrocks]

as a heat flux is to neglect the directional components or rather the distribution of the heat flux.

You are confused. Heat flux is a flow of a scalar through something. The heat has no direction (although you could say the flux does). In other words Temperature is not a vector quantity. Radiative heat flux is electromagnetic radiation and that is a vector quantity and therefore has a direction.

Yes, you are missing an important point here. Temperature is a scalar field and therefore has no directional component. You can define a temperature boundary condition as a temperature or a function of temperature - no direction. Radiative heat flux does have a direction and is a vector quantity.

[ghorrocks]

as a heat flux is to neglect the directional components or rather the distribution of the heat flux.

You are confused. Heat flux is a flow of a scalar through something. The heat has no direction (although you could say the flux does). In other words Temperature is not a vector quantity. Radiative heat flux is electromagnetic radiation and that is a vector quantity and therefore has a direction.

Yes, you are missing an important point here. Temperature is a scalar field and therefore has no directional component. You can define a temperature boundary condition as a temperature or a function of temperature - no direction. Radiative heat flux does have a direction and is a vector quantity.

[Kelevraa]

Static Thermal or Transient thermal are no options for me cause I'm primarily interested in the fluid inside the building.

Glenn, I know that the heat flux boundary has no direction and the temperature is a scalar quantity. That's why I think I have to apply the solar radiation as a directed radiation source. But there lies the problem cause I dont know how to do it exactly.

[ghorrocks]

A quote from the documentation: "Radiative Flux into the domain for the Discrete Transfer and Monte Carlo radiation models is specified via the Sources tab for the boundary condition in CFX-Pre. The Monte Carlo model allows directional and isotropic radiation sources to be specified"

[Kelevraa]

I might have phrased my last comment unfortunate. I know how to set the radiation source in general (I did it successfully for the solar radiation through the windows with Monte Carlo as example). But it doesn’t work the way I imagine it would do on the walls.
I want a directed source on an opaque solid which heats the outer wall. (I could imagine that the main issue might be the absorption coefficient which I don’t know how to set (I'm actually simulating the wall as a participating media))

[ghorrocks]

I think you better post an image of what you want. I do not understand your question.

[Kelevraa]

Here is an example. It's just a simple “building” which gets heated by the sun. The arrows show the direction of the applied solar radiation source. They are applied on the outer boundary of the window and the wall the same way.
The inner fluid (the ground in particular) gets heated through the window the way I assume it should.
But the walls work like a semi-transparent too (like I said, the second element of the wall is hotter than the first if you look from the outside) but it should be opaque.

[Opaque]

I am bit confused by your description, but here is my stab at it. Let us assume a 1-dimensional configuration

External radiation source --> || glass || air  || <-- opposite boundary

Radiation model is active for the glass and air, external radiation source is applied on the external boundary. The external radiation source must travel through the glass, cross the glass-air interface, travels through the air, and arrives to the opposite boundary. Correct ?

If the above is correct, I expect the second element on the glass to be hotter than the first since it is being heated as the radiation passes through. If the glass is opaque, the radiation will (by definition) NEVER pass through the glass (so my confusion). If you do not want to heat up as radiation passes through it, you must make the glass perfectly transparent, i.e. set the absorption coefficient to 0. However, even for a transparent glass I expect refraction of the rays as it goes from the external medium through the glass, and once more refraction when it cross the glass-air interface. If that does not happen, the calculation is meaningless and you may as well delete the glass, and apply the external radiation at the "glass-air" interface directly.

Hope the above helps,