[dom]

Dear all, I am trying to obtain the velocity streamline for a ventilation system and I have generated a simple 5m by 5m room with a protruded squared surface that represents the air-conditioning system. The general idea is to observe the streamline pattern as it fills the room and exits through the window. However, the flow profile doesn't behave as expected as it 'creeps' through the walls rather than filling the empty room spaces. Please advice!


Uploaded with ImageShack.us

[doctor evil]

I think your simulation didn't reach convergence (see legend of the left top, do u think the velocity in your problem will be higher than 1e2m/s)

[dom]

Many thanks for the swift reply doctor evil. No, of course it wouldn't be that high. (that said, I don't know how it got that value, my inlet velocity was set to 5 m/s) Also, How do I ensure convergence has occured?

[Daniel Tanner]

Are you sure that the size of your room is what you think it is? If it is a cm to meter difference this might explain why the velocities are high. Check the domain size in the scale dialogue.

[Sixkillers]

Moreover what boundary condition do you use for outlet? What turbulence model do you use? Is it isothermal case? Also some information about mesh would help us (Mesh -> Info -> Quality, Mesh -> Info -> Size).

Also, How do I ensure convergence has occured?

At least values of residuals for continuity and velocity should be below 10^-3. For energy it is 10^-6. However more important is to monitor quantities, which have physical meaning. For example, for your problem it could be mass flow of air on inlet and outlet. Difference between them shouldn't be higher then several percents.