How to simulate windows?

[duanqa]

Hello,

I have a question about window simulation. I would like to model a office with a single pane window.

I built two modes :

1)  At first, I set window as "air" fluid material and its exterior surface film coefficient is 5.6 W/m2-K to simulate a single pane window; I used steady analysis and obtain the global temperature is 24.3 Celsius degree, heat transfer from wall to fluid is -27.06W.

2)Then, I set windows as "glass" solid material;  I revised its resistivity is 3.94 m-K/W, its thickness is 0.0254 m, and its exterior surface film coefficient is 12.5 W/m2-K, in this way the R-value (or U-factor) of the window is equivalent to 0.18 m2-K/W, it is the same as the first situation. I used steady analysis too, but obtained the global temperature is 24.5 Celsius degree, heat transfer from wall to fluid is 50.337W.

In these two models, the other conditions including geometry, boundary conditions, and mesh etc are same.

I think theoretical the results of these two models should be same. Why they are different?  Especially why the "heat transfer from wall to fluid" is different even though the window has equivalent R-value (or U-factor)?

Thank you!

[Jon.Wilde]

Your first approach sounds a bit weird to me 🙂

Might you be able to share a model (CFZ) with both scenarios included so we could have a quick look?

[duanqa]

Hi Jon,

Thank you for your reply. Please see my models in the attachment.

For the first situation, I follow this exercise to simulate the window as air with film coefficient: 

For the second situation, I simulated the window as glass and set its material and film coefficient to make its U-factor equivalent to that of the first situation, but the results of both situation are totally different. So I want to know what is the correct way to simulate windows

[duanqa]

This is my second situation.

[Jon.Wilde]

Are you aware you can have multiple designs and scenarios within the same project? Often this really helps when comparing results as you can do it directly in the interface 🙂

Why not use the solar loading? That way you actually have energy passing through the window as it would in reality?

I'm not sure you can compare in the way you are. The energy radiated from the panel would change with the new material, which would need flow and thermal on to capture the internal effects.

If you need to model a window, I would just use a window material. If you plan to keep the model as you do now, you would need a better mesh on the window part, at least 2 elements through the thickness. Use a manual mesh to achieve this throughout the volume.

Hope that helps

Jon

[duanqa]

Yes, I had multiple designs and scenarios within the same project, but they are too large to be uploaded to the forum when I saved it as cfz file, so I saved as different file for each situation.

Thanks for your suggestion that I will manual mesh to get better mesh on the window part. For these two situations, I just want to compare which one is the better way to simulate the window. The first situation followed the exercise

 http://auworkshop.autodesk.com/library/cfd-aec/exercise-boundary-conditions-simulation-cfd?language=...

The second situation used a window material (glass). I would like to set the overall U-factor of the window be equivalent in the two situations, so I can compare the computer time or results precision between them. I did not consider solar radiation in this model so I did not use solar window.

Actually, I do use solar window in another model when I consider solar radiation. But I am confused about the solar heating dialog, what is the weather data behind this? Can I export the weather data from Autodesk CFD? Because I would like to know the solar irradiance data, temperature data etc. when I chose a specific time from the solar heating dialog.