CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   how 2 freeze 1 phase flow field & start lagrangian (

KK February 12, 2008 15:38

how 2 freeze 1 phase flow field & start lagrangian
I solved single phase flow field with LES and now I want to inject particles of different sizes.

Is there any way to 'freeze' single phase flow field and track the particles of different types? By freezing I mean that I want to stop solving single phase flow field (carrier fluid) and use that flow field to track my particles. It will save my computational time.

Glenn Horrocks February 12, 2008 17:33

Re: how 2 freeze 1 phase flow field & start lagran

Yes, this can be done. Set up a simulation with 0 set as the maximum iterations (or you might need to do one iteration just to keep it happy) and do an uncoupled particle tracking on the result of that.

Alternately you can stop the solution of various solvers (eg Velocity/pressure, enthalpy, scalars, turbulence) individually using expert parameters.

Glenn Horrocks

KK February 12, 2008 18:31

One clarification please
Thank you Glenn.

In first approach, did you ask me to define maximum time steps? (I am doing LES and hence transient). Or does it mean zero (or one) iteration in steady state solver taking solution from LES as initial guess?

In short, I am confused 'iteration' Vs 'time step'. Please clarify.

Glenn Horrocks February 13, 2008 17:24

Re: One clarification please

If you just want a particle track based on a frozen flow field from an LES simulation, do a steady state simulation with zero iterations and uncoupled particle tracking. I think this should work.

Glenn Horrocks

KK February 13, 2008 17:40

Re: One clarification please
Thank you very much Glenn. I will give it a try.

By the way, I can imagine that liquid-solid particle simulation with Eulerian-Eulerian approach with LES model would be extremly costly in terms of computational resources. (Assuming we have variety of particle's sizes)

Do people do these type of simulations? Or is it not feasible at all to do in terms of time consumed?

I will be thankful to you for any of your comments on it.

Glenn Horrocks February 14, 2008 17:48

Re: One clarification please

Yes, multiphase modelling will significantly increase the cost of an LES mode. The expense of a LES model is also strongly dependant on the Re of the flow. Low Re are not too bad as the turbulent eddies you have to resolve are large. For high Re flow you need much greater resolution. This is the primary factor influencing the size of an LES simulation.

Whether it is feasible depends on how big a simulation you need to run to get useful results, how big a computer you have access to and how long is your patience.

Glenn Horrocks

All times are GMT -4. The time now is 06:25.