Hi, I want to simulate the airflow for window openings under different outside wind speeds. So I divided the simulation space into two parts, one is the outside zone (for outside wind flow), and the other is the interior zone (for the air flow inside the room), and set the windows in two directions. As it is difficult to know the flow direction at these two openings, their boundary types are assumed to be pressure-outlet. But the Fluent told me the openings have two adjacent cell zones, and the grid check was failed. Could anyone give me some advice on setting the boundary conditions for openings, which connect the interior zone and the outside zone. Thanks very much.

In truth the only boundary between an the air in a room and the air outside a room is imaginary if the window is open.

I suggest you try creating the mesh without any boundary type for the open window, connecting the two fluid reagions through the window, as they would be in the "real" world.

You can then observe the flow in and out of the window using either monitors whilst itterating, or place an iso-surface in the flow to examine more closely the flow at this point after convergence.

If I knew which grid creation software you were using I might be able to give you some more specific advice.

Hope this helps,

Simon

Thank you Simon. I want to set the velocity loss coefficient for the window opening, so it is necessary to set boundary conditions. I use Gambit2.0 to make the model and creat the mesh, and I am looking forward to your further advice. Thanks.

Hia Lei,

Not really sure what you mean by a velocity loss coefficient! (Sorry).

If you are constructing the grid in GAMBIT then you can create two faces or volumes which share an edge or face. This face can be the window opening. You can set the fluid region to be all air in both volumes (or faces) by going to the Zones-continum pannel and selecting one fluid for the two zones. No further clarification of the boundary between the inside and outside is needed.

This will allow FLUENT to calculate the flow from inside to outside.

However, if you really need to specify some values at this interface, you could try using the "interior" bounday layer. However, such a value would be related to the opening geometery or the flow properties inside and outside the room. In which case, the "loss coefficient" could be incoperated into other aspects of the CFD solution, similifying your problem!

Hope this helps,

Simon