[hamed.majeed]

I am doing an air-water VOF analysis. The domain has just one open condition-an open bc. The water level in the domain rises drastically in the simulation.

Any views on this discrepancy!

[hamed.majeed]

I used CFD-post to find the volume of the two phases and I get this.


Total volume of domain = 0.0229118 m^3

Time step ; Volume of water ; Volume of air; Total volume.
-----------; ------------------ ; ---------------- ; ----------------.
0; 0.00089976 m^3; 0.022911776 m^3 ; 0.022911776m^3.

50; 0.00124594 m^3; 0.0216658 m^3; 0.02291174 m^3.

116; 0.0089267 m^3; 0.0139851 m^3 ; 0.0229118 m^3.


Note: Air is considered incompressible!

[ashokac7]

Originally Posted by hamed.majeed

I am doing an air-water VOF analysis. The domain has just one open condition-an open bc. The water level in the domain rises drastically in the simulation.

Any views on this discrepancy!

This is the problem with VOF. i.e. error accumulation over the time. I know Eulerian multi-phase may not be the practical. But you can play with physics set up to reduce the error like reducing time step etc. You can also try setting initial conditions as Vof VC's.

[hamed.majeed]

Originally Posted by ashokac7

This is the problem with VOF. i.e. error accumulation over the time. I know Eulerian multi-phase may not be the practical. But you can play with physics set up to reduce the error like reducing time step etc. You can also try setting initial conditions as Vof VC's.

Thank you Ashok for your response.

I actually figured out the exact problem. The issue was with the global courant number. I did not realize that in VOF there are two courant numbers to consider (one for the volume fraction and another related to fluid flow). Moreover, the explicit scheme with Geo Reconstruction seems more accurate than the implicit (earlier I was using implicit).

Now, I find my numbers to add up perfectly. You are right, due to error accumulation especially with high global courant number I had issues.

Thank you.