[MichaelPage]

Hi,

I have to do a simulation of a enclosure with hot air inside of it. The enclosure is outside at a cold temperature with wind. I must get de temperature of the air in the enclosure after 10 minutes. If I consider the outside air, the timestep is to small and the simulation will run many weeks. I thought to a way of simplication for this case and I want to have some opinion and advice.

So, I'll do the simulation with all domains first to get a stationnary profile of the outside air. After, I will get the wall heat transfert coefficient (I'll use average value) on the wall of the inside domain (the enclosure was not modeled because it is thin geometry and I specified a thin resistance in domain interface). In the next step, I'll remove the outside air and I'll specify the wall heat transfert coefficient where I had the domain interface. Then, I'll run this simplified case.

Do you think that this procedure is conservative? Do you have others solution of simplication to suggest?

Thank you.

[ghorrocks]

The approach you suggest sounds sensible. It is a good approach providing there is a large separation in time scales and there is little coupling between the inside and outside flows.

Note it is not much harder to extract the full heat transfer coefficient distribution on your enclosure and use that rather than an averaged value. Then you can capture the effects of different HTC on different areas of the enclosure.

[MichaelPage]

Thank you ghorrocks for your answer.

Also, can you tell me the procedure to extract the heat transfert coefficient distribution to use it like boundary condition?