head loss K on free outlet

[mattkoch]

So, I am trying to run a relatively simple case where I specify a duct with a wall, an inlet with an imposed volume flow rate and an outlet that represents a throttle in real live. Obviously then, the outlet cannot be just a simple "outlet" to reference pressure. Instead, I think I will have to use a "free outlet" and give it a K value. The CFD domain should then (slightly) pressurize until the pressure drives enough flow through the throttle to match the incoming flow. In fact, if the wall and remainder of the flow contribute little pressure loss, then I would think the pressure rise should be on the order of dp = K*1/2*rho*v^2, where v is the velocity at the outlet.

The theory manual discusses this for a "free inlet" but not a "free outlet". In the "free inlet" discussion, K is positive. If the exact same equation discussed there were to apply to the "free outlet", then I believe K should be negative for a "free outlet"? Can someone confirm or deny this?

Also, my calculation, which seems to at least run through for a regular "outlet" keeps blowing up when I change to such a "free outlet". The "listing" file says this is due to "pressure divergence"? Any thoughts on what might be going wrong?

[Yvan Fournier]

Hello,

Is your mesh orthogonal or does it use tetrahedra at the outlet ? Could you start with small k values ?

Regards,

Yvan