Raushan Kumar Please share CFD domain with boundary conditions

Dear Sir / Madam The following is my attempt to help you:

• https://www.afs.enea.it/fluent/Public/Fluent-Doc/PDF/chp20.pdf
• https://support.ansys.com/staticassets/ANSYS/Conference/Confidence/Phoenix/Downloads/solution-every-multiphase-challenge.pdf
• https://core.ac.uk/download/pdf/30667473.pdf

With respect

Dušan

There is a term named "spurious currents in VOF". Probably it could be helpful to Google after this term. Numerically determining gradients on a curved two-fluid interface on an - maybe not sufficiently - fine grid is a tricky thing. In the consequence the numerical method might lead to some numerically existing and non-zero gradient, while in the physics thee is absolutely no gradient present. This usually becomes visible only in some mathematical verification benchmarks, since in a real flow usually you study something which is in strong motion and than the physical gradients are comparable large in comparison to the (hopefully) very small numerical errors in gradient prediction.

Usually I tell students, that a flow solver is not very well suited to study something, which is not flowing but is at rest. Of course this is a bit ironic comment. But there is some truth in it, because in the numerical methods of fluid dynamics solvers there are quite a bit of stabilizing measures implemented, which rely on the existence of non-zero fluxes. If there is no flux (of mass, momentum, energy) than a CFD code can very quickly behave rather unstable or physically not 100% meaningful. It takes incredible care to implement a CFD solver in a way, that even the most pathological CFD benchmark can be successfully be solved with it.

Best regards,

Dr. Th. Frank.

The answer by Dr. Thomas Frank demonstrates the issue you are facing. I would suggest you to employ OpenFOAM to simulate your droplet coalescence phenomenon. You will be able to obtain a sharp interface using the interFOAM solver under appropriate settings.