Has Anyone Tried Unsteady Cavitation ? The collapse of cavitation bubbles and their nature require Compressibility Effects..However Fluent is highly unstable.

I tried to work with unsteady CAvitation with less success..though the same model/geometry works for steady cavitation.

any Suggestions ?

CFDtoy

Hi;

You can get some informations about unsteady cavitation modelling from cav2003 conf. "Reboud" & "coutier-delgosha" have said some suugestions to get unsteady behavior of cavitation. These suggestions are reffered to turbulence properties of flow field. Mateus has modeled cloud cavitation with fluent6.1.18 without compressibility assumption of cavity according to Reboud suggestion and his paper is in cav2003 conf. you should modify turbulence viscosity in two phase region of flow field according to Reboud suggestion or use k-w turbulence modelling(Wilcox 19980 with compressible cavity assumption.

To model cavitation with vapor compressibility effect, you should start with a non-cavitating condition. Befor formation of any cluster of cavity, you should set the density of vapor to a constant value as well as density of water. In the time of inception of cavitation you can set the density of vapor to "ideal gas". If you do this, you do not get solution unstability.

Can you get converged solution with steady modeling? To model a steady cavitation like sheet cavitation, I work with unsteady solution to reach a steady state.

What version of Fluent do you work with? I work with fluent6.1.22 and I never get correct results with fluent6.0 or 5.

Hope these help you

ROOZBEH

Hello Roozbeh

thanks for your reply. Cavitation models used in Fluent do not have Surface tension , Vapor pressure as a function of Temperature..which is not realistic !!

Also, the problems are such that once you say that the Vapor is compressible which is true...you start solving Energy Equation and stability is not easily achieved.

I acheived steady solution and started working with Unsteady Cavitation..however, trying all the relaxation parameters and discretization only gives me little stability....in the long run the code blows up.

What is your suggestion, I do not know how Fluent works with Cavity Implosions !!..looks like the Rayleigh Plessets eqn is being solved with a variety of bounding constraints..not truly representing the physics.

Thanks for the Discussion. Keep them coming.

CFDtoy