Yves
Posts: 29
|
 Posted: Wed Oct 05, 2011 9:57 am Post subject: Simulating internal louvres |
 |
|
Hello,
I have a quite complicated configuration that I would like to model using CFD. Here it goes:
- a sloped, totally glazed roof
- underneath the roof, louvres composed of large horizontal opaque blades, arranged like a kind of staircase
- in each space between two blades there is a pane of glass (so that the volume above the louvres and the room below them are separated).
To sum it up, it is a kind of double skin sloped roof, where the outer skin is a simple tilted glazing, and the inner skin is like steps with opaque horizontal surfaces and transparent vertical surfaces.
The aim would be, in summer, to protect from sunlight and evacuate the hot air to the outside (by chemney effect, or mechanical ventilation), and in winter to take the hot air between the two envelopes and feed it back to the room, like in a trombe wall.
In order to do this, I encounter several problems.
First I wanted to use E+ and import the results to the CFD, but I don't know how to do it accurately, because from what I understood, the solar heat gains in E+ will be distributed in the room, instead of heating just the surface of the louvres (which are described as component blocks). Furthermore, the temperature of the blades should depend itself from the air stream above them, so a constant temperature boundary doesn't seem to be quite realistic.
A second option would be to use a flux boundary condition on the blades, that I could calculate "by hand" using the incident solar radiation. Would that be more correct? This option implies is also not very realistic because obviously if the air above the blade is hotter the heat flux will be smaller. That is, the total radiative + convective heat flux should be equal to the incident solar radiation at equilibrium, but it is hard to tell what proportion of the heat will be sent as radiation to the outside in which proportion will heat the air above the blade.
Another issue: in the first place I will consider a forced ventilation, place air flux boundary conditions at the bottom and top extremities and try to calculate the temperature difference. But ideally I would like to model the natural ventilation by chimney effect. Is that possible?
I'm aware it's not an easy design to model, but if anyone has an idea of how to handle it (even with a lot of simplifications) it would be great!
Thanks
Yves |
|
FAQ
Search
Log in to check your private messages
Usergroups
