[hmasenger]

hi dear friendsI am modeling a transient cavitation simulation and I need to have an initial condition(guess) before activating cavitation function in cfx.I know I should run my model without activating the cavitation parameter at first and use the primary results as an initial guess for cavitation simulation. The problem is that my model is not a steady state run with a single time step so i need several initial guesses for every time steps in my transient model .i don’t know how to define a multi time step result as an initial condition (guess)for a multi time step cavitation model.Please guide me Tanks a lot

[ghorrocks]

You need to define an initial condition at the start of a transient simulation, from then on the time steps work out the changes from the last time step. So just define a sensible starting point and the solver will work fine from there.

[hmasenger]

thanks for your reply ghorrocks

in my model i am changing inlet velocity in pipe to find out the velocity which cavitation is going to begin .in one of CFX tutorials (which was an steady state model)i found out that i should first run the simulation without activating cavitation function then and then the results must be used as an initial condition BUT the point is, in that tutorial the author knows the inlet velocity which cavitation occurs and in my model I don't ,and the right cavitating velocity is the goal of my model .do you think this won’t effect the results? Am i following the right way by just using a proper initial condition for first time step in transient run ?
by the way a formula for inlet velocity variation, to find the cavitation velocity made my model transient.

[ghorrocks]

If you "just" want the flow rate at which cavitation starts, then why not do this:
* Run a single phase model at a low flow rate (which definitely does not cavitate).
* Increase the flow rate until the minimum pressure in the region of interest equals the vapour pressure (still running single phase)

This is a much simpler approach as it does not involve cavitation modelling at all, and is a straight forward series of steady state simulations. You could also do this as a transient simulation but that is likely to be slower.

[hmasenger]

As you know cavitation usually does not accrue in fluid vapor pressure.it also depends on nuclei and dissolved bubbles in water. Well as without activating the multiphase function we can't see the volume of cavity which grows with pressure drop. And we can't figure it out where it immerses.
bye the way thanx for your reply.