Hi;

Should I set the option "differential viscosity model" in RNG k-e model for a cavitating flow

Thanks

ROOZBEH

Generally speaking, you enable it if your mesh is designed for integrating the governing equations down to the wall. In contrast, if it is too coarse close to the boundary, you should use a wall function approach. In latter case, the differential viscosity model should be disabled. Check e.g. the y+ values to get a guess what kind of near-wall treatment you need.

I don't know for which combination (cavitation model + fine near wall grid + low-Re model or cavitation model + coarse near wall grid + "standard" turbulence model) you will get improved accuracy as I have not been using it very much. However, I could imagine that a setup with fine grid + low-Re model is better suited since bubble creation and growth appears in general close to the boundary. Accordingly, a "standardised" formulation might not be the proper choice for modelling strong diparities.

Philipp Beierer