CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   STAR-CCM+ (https://www.cfd-online.com/Forums/star-ccm/)
-   -   vof=0.5 in dfbi (https://www.cfd-online.com/Forums/star-ccm/63430-vof-0-5-dfbi.html)

vof=0.5 in dfbi

In the tutorial, why do they choose 0.5 to visualize the isosurface of VOF? What does VOF=0.5 mean? Is it a surface where it is 50% air and 50% water? Shouldn't we use VOF=1.0 to show the actual surface of water? Can I get explanation from the DFBI experts in here? :p

Actually any fractional VOF value (i.e. 0<VOF<1) means that gas-liquid interface is within this cell. And because isosurface creation tool doesn't use any interface reconstruction technic, any isosurface, for any VOF value, will not coincide exactly with actual interface.

Thanks Anton! So, that means we should be showing the VOF=1.0 if we really want to show the actual surface of water? Why does tutorial choose 0.5? When I played with VOF values, almost nothing appears when the 0.9<VOF<=1.0. And between 0.5 and 0.8 the surface looks radically different so the question is which VOF value truly represents the surface for my customers to understand? Your info on interface construction technique is good. I think CD-Adapco should update their adapco-online website with this kind of basic general information.

No, the situation is more tricky. There's no way to get CORRECT representation of REAL free surface with isosurface of VOF. The best way is just to select all the cells for which 0<VOF<1 (note that the inequality is strict!). This will give you all the cells which are crossed by the interface. This will not look nice, I guess, but at least it will be physically correct. And you can then compare isosurfaces with different VOF values with this cell set and choose the best match. Probably VOF = 0.5 will be the best, but you'll have to try.

If you want something nice looking AND correct at the same time, then you'll probably need some advanced postprocessing. You'll have to use interface reconstruction techcnics, which are usually built in into VOF algorithm. I'm not a VOF expert, but I think you can find some in VOF papers. And implementing them will require pretty heavy programming.

 ping April 16, 2009 05:57