It is reported there are significant parasitic flow besides the interface in the simulation of two-phase flow using CSF or CSS model

Is there any good way to reduce it?

Thanks a lot

Parasitic flows near the interface are created due to the errors accumulated in the interface curvature estimation. There are several ways to reduce it. recently, an energy method to reduce the spurious current was reported in JCP by Kothe (I guess). Doesnt really matter CSF/CSS just a few differences exist in the methodology. Both require curvature estimation and thats where the problem begins !

CFDtoy

Well, yes, but the balance is very important. If you are doing capillary flows and the parasitic velicyt is 100 m/s, you can throw your result away. Therefore, writing a decent surface capturing algorithms is still critical.

Hrv

The magnitude of parasitic velocity Up ~ ratio of the viscosity to surface tension(the capillary number). One way to check the balance is to compute, dP = sigma*curvature variation in the neighborhood. Since the VOf methods use a single flow field, any variation inthe volume frac gradient reflects in the local force effects.

Smoothing, say using a laplacian in a unstrc mesh could reduce the spurious force effects (note CSF etc require smoothened VOF scalar in the field).

Still, the need for good curvature evaluation is critical !

CFDtoy