[Bodo1993]

Hi,
I am using twoLiquidMixingFoam solver.
I have a single inlet and two outlets in a symmetrical geometry.
At the inlet, I specify a uniform velocity and a zeroGradient Prgh.
At the outlet, I specify Prgh = 0 and pressureInletOutletVelocity.
I get a back flow at the outlets. I tried to use over velocity boundary conditions at the outlet such as the zeroGradient, inletOutlet and flowRateOutletVelocity, but the issue persists. I have also extended the geometry, but it did not resolve the problem.
I would appreciate any assistance.
Please let me know if any further details are required.
Thanks.

[jherb]

Does using type inletOutlet with inletValue 0.0 not help? This should do, what you want.

[Bodo1993]

Hi, Thanks for the reply.

Please find attached document that has the details and snapshots of the results.

I look forward to hearing from you.

Kindly, let me know if further information is required.

Thanks CFD_Online_compressed.pdf

[Bodo1993]

Hi jherb,
I would appreciate your assistance.
Thanks

[jherb]

For Outlet1 and Outlet2 try the following boundary condition for U:

"Outlet(1|2)"
{
    type inletOutlet;
    value uniform (0.0 0.0 0.0); // just use for initialization
    inletValue uniform (0.0 0.0 0.0); // this should prevent inflow
}

see e.g. https://github.com/OpenFOAM/OpenFOAM.../TJunction/0/U

[Bodo1993]

Dear jherb,
Thanks for the reply.
I have used that already; however, the problem persists.

[jherb]

Are there walls (vertically in your picture) between the inlet and the outlet? Are you sure, it is not physically correct, that there is inflow from the outlets if you have a pressure boundary?



If you use inletOutlet, have you checked the values on the faces of the patch? So explicitly loading the patches in paraview? They might be zero, even if the cell center values show inflow.

[Bodo1993]

Hi,

Yes there are physical walls between the inlet and outlets. I believe that it is not physically correct to have a back flow. What I simulate is a fraction of a very long geometry in real life.

The backflow gets worse when the inlet channel is shifted from the center line. In that case, the flow gets back to the domain and messes up the results (not just only at the outlet boundaries). For example, I tried shifting the injector to the extreme right, leaving a narrow gap for the flow to escape to the outlet. In this case, I always get terrible backflow, as if the flow leaves one outlet and enters the second outlet.

I would appreciate your thoughts on the above. Thanks.