cyclone
 

HI!!!!

I'm working with a cyclone. I have some problem with convergence.

I' using RSM model.

My BC is " mass flow inlet" and "Pressure outlet":

mass flow rate = 0.02 kg/s

Operative pressure = 101325 Pa

Pressure outlet (gauge) = 0 Pa (Static pressure)

During the iteration I have reverse flux on outlet, then turbolent viscosity limited to viscosity ratio and then an error for invalid number.

Is there someone that can help me?

Re: cyclone
 

You may want to solve the flow and turbulence with the k-eps model first and then switch to RSM for the final convergence. Also, change to second order for calculation of parameters.

Also, you could specify the outlets as mass flow outlet and estimate the fractions at the outlets. (One outlet will likely have almost all of the flow).

If the viscosity ratio error bothers you, you could just change the limit. The limit of 1x10^5 is just arbitrary.

Good luck.

Re: cyclone
 

Allan is right about running the solution using a less complex turbulence model first but just to make sure you need to use the K-e RNG model with the swirl dominated flow radio button on and adjust the swirl parameter to your model.

When running the RSM on second order you may find that the solution will not converge to the limits specified thats if the solution doesn't diverge. If that is a problem first solve the steady state solution then switch to the transient solver as this will allow the solution to converge easier with an appropriatly small enough time step. Also don't worry about reversed flow on the outlet if you are operating in a high swirl regime (S>0.6) as this physically occurs in cyclones within the SBR region of the PVC.

Hope this helps