2phase w mass transfer and k-eqn

Niklas Wikstrom

Hi,

I try to�@model cavitating flows and use a variable density, but incompressible approach, LES for sgs turbulence modeling and a VOF type handling of the "two phase continuum". Then, IF using a 1eqn sgs model, with turbulent kinetic energy governed by e.g.

ddt(rho k)+div(rho k u) = sourcesnstuff,

(and here is the question), what to�@do to cater for the possibly large ddt(rho) (emanating from mass transfer i.e. vaporisation) in the flow problem? I.e. if rho decreases k would increase unphysically?

I have seen "cavitating solutions" from e.g. Fluent, using RANS, that refuses to go�@unstationary (the reality sure is) MABY because of a massive production of K at the cavity interface, where vaporization occurs, cooking syrup in the whole cavitating area.

Someone with experience in this? I suppose it applies to many multiphase flow problems...

Niklas