CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM (http://www.cfd-online.com/Forums/openfoam/)

 kaifu April 12, 2011 09:44

about bounded alpha equation for bubbleFoam and twoPhaseEulerFoam

Hi Foamers,
This time I will deal with subcooled boiling flows modeling which includes evaporation and condensation models. Actually the task is the continuous work of Edy, see http://www.cfd-online.com/Forums/ope...ated-wall.html

I start my solver from bubbleFoam and twoPhaseEulerFoam, which use Weller's strategy to solver UEqn, pEqn, and alphaEqn. In Weller's approach, one of the most important things is that he found a way to bound alpha in [0,1]. Here I need to point out that the reason why alpha is bounded at one is due to the "conservative" feature of volume mixed flux, that is <eq>\nabla \cdot \bar{U}=0</eq>.

However in my case, since there is a phase change term (due to evaporation and condensation) in alpha equation, which means alpha could be still bounded at one but not zero any more by Weller's approach, the program crashed. I noticed that the divergence may come from pEqn.H, which should promise the "conservative" feature of volume mixed flux. However it reports as,

time step continuity errors : sum local = 4.22651e+09, global = 5.82749e+07, cumulative = 5.82749e+07

And then the program crashed...

So anybody could give me suggestions? I have actually considered to start my solver from interPhaseChangeFoam, but I do not need to track the interface...

Thanks a lot.

 kaifu May 5, 2011 11:00

Some updates: this problem comes from pressure equation. In bubbleFoam since there is no phase change, the volume flux is conservative. When dealing with phase change, we have to use the mass flux instead of volume flux as in bubbleFoam.

 sabin.ceuca May 6, 2011 05:33

Hi Kaifu,
I’m currently working on direct contact condensation with VoF and if I understood your problem correctly then interPhaseChangeFoam is not the right source for you. That solver considers phase change due to cavitation and has a one fluid approach. The only information gain is that it shows you where you’ll have to add source terms in the alphaEqn.

 All times are GMT -4. The time now is 08:05.