CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   dambreak switched properties boundary (

story October 28, 2013 04:29

dambreak switched properties boundary
Hi everybody,

I was trying out the dambreak tutorial, but swithed the transport properties.
I wanted the air to rise out of the domain as a rising bubble. I have gone through several threads in this forum and also found the single bubble rising papers of Hysing et al. and Klostermann et al. However in these cases the rising bubble does not leave the domain.

I wanted to simulate my bubble leaving the top boundary. Therefor as an example I switched the properties in the dambreak tutorial. All sides are all wall except the top boundary. For U it is pressureInletOutletVelocity and p_rgh is the totalPressure. For alpha1 I have set to calculated. However, the simulation crashes once the air reaches the top boundary.
Does any of you have a different boundary condition?

For other boundary conditions the simulation runs fine, however I do get a free surface on top for zeroGradient, inletOutlet and outletInlet condition for the alpha. But I want my alpha1 to leave the top boundary.
Do you think I should change also the other boundary conditions for U and p_rgh?

Thank you all in advance.

nimasam October 28, 2013 06:38

your question is too vague, however choosing calculated B.C. for alpha1 is totally wrong.
you can simply use zeroGradient for U and alpha1 and fixedValue for pressure, then bubble would leave the domain.

story October 28, 2013 09:48

2 Attachment(s)
Hi Nimasam. Thanks for the reply.

You're right. My question does seems vague.
As I stated I started with the dambreak laminar interfoam (openfoam 2.2.2) tutorial as an initial test for gas rising out of domain starting with a initial shape.

I just renamed phase1 tot phase2 and phase2 to phase1 in the transportProperties file. That is the only change I make besides the boundary conditions.

But If I set the BCs as you suggested to zerogradient for U and alpha on top and p on top to a fixedValue, my phase1 still does not leave the domain. I still get a free surface on top. I attached the initial and end state images of the simulation with your suggested BCs.
Got any extra pointers?

As per your suggestion the boundaries are:

In the original tutorial files:

type zeroGradient;

type zeroGradient;


type fixedValue;
value uniform 1;

Thanks again.

Attachment 26424

Attachment 26425

nimasam October 28, 2013 10:08

consider for alpha1

type  inletOutlet;
inletValue  uniform 0;
value  uniform 0;

story October 28, 2013 10:22

1 Attachment(s)
I have tried it also. But in the end, phase1 still does not exit my domain.
I attached the whole tutorial with the minor adjustment of myself. As you can see, I have set up the tutorial with no major changes. It is still strange that for zeroGradient of alpha1, it forms a free surface on top.
Do you think the OpenFOAM version matters?

Attachment 26427

All times are GMT -4. The time now is 19:01.