[lethu]

Hello everyone !

I am working on OpenFOAM 4.1 and I am using interFoam to model a lateral spillway with two exits. See the attached image.

https://drive.google.com/open?id=0B7...TdiekNuNjB1emM

I have tried many set of BC for my outlets. Please see the excel chart. Some words are in French but it is quite explicit.

https://drive.google.com/open?id=0B7...lJiNXc4MmlLSjQ

I am struggling with the fact that :
- either the simulation is doing well but one outlet is not emptying, or the two outlets are have very, very small outflow (such as 0.02 m3/s)
- either the chamber seems to empty in both outlets but it encounters some diverging problem. I am monitoring U and P with fieldMinMax in controlDict and it shows exploding values !

checkMesh does not seem to report anything wrong.

I have tried so many combinations of BC for the outlets that I am starting to wonder if it doesn't come from my mesh.

Please, can anyone help ? I have been working on this case for three months now.

[piu58]

I recommend setting zero gradient for U and a fixed pressure for p at the outlets. I did not find this combination in your sheet. But it may be I missed something.

For the inlet a fixed U (which you want to simulate) and zero gradient for the pressure.

That means:
The outflow is "free", without any constraints. The pressure level has to set somewhere, and the outlet is fine for this.
The inflow is also free: It is not (additional) pressure driven.

[lethu]

Thanks for your reply Uwe.

But isn't it the case for the first try "0" I made ?

From the .H file, I understood that pressureInletOutletVelocity means that zeroGradient is applied for all components on U when there is an outflow.


I am going to try your solution anyway. For the outlet, should I use totalPressure with p0 = 0 ? Or fixedValue = 0 ?

I hope this time, it will work !

Plus, I am also wondering if it comes from my initial conditions.
I used for the last cases, solution that seems correct for the initial conditions using mapFields. The fact that it is not emptying or diverging cannot come from that, can it ?
I don't know if should use a simple setFields and initialize with a simple box of water on the top with a small velocity. I suppose this would take more time to converge than using mapFields.

[piu58]

> pressureInletOutletVelocity means that zeroGradient

I don't work with pressureInletOutletVelocity , but always use zeroGradient.


> use totalPressure with p0 = 0 ? Or fixedValue = 0 ?

I always use fixedValue.