[petr.f.]

Hello Foamers,

I'm little bit confused with turbulence modelling options in multiphaseEulerFoam. It seems that by default there is only LES included in the source code (#include "LESModel.H"). I'm not sure what to do when I want to use RASModel - if I have to modify the code by including RASModel.H or if there is any option for LESModel to simulate RAS instead of LES. Does anyone of you have any experience with that?

Thanks, Petr.

[kwardle]

You are correct that mpEF only does LES by design. This is because the intended application was where the sharp interface and interfaceCompression is used and thus time steps are already small and LES is little to no added cost. You would have to modify the code for RAS. One problem is deciding whether to calculate turbulence only for the continuous phase or have RAS model per phase with inter-phase turbulence transport--this is more general and more complex. Again, depending on your application you might choose one or the other. Based on questions you have asked elsewhere it seems like you will be using interfaceCompression in which case I wonder why you are interested in RAS models.

[petr.f.]

Again - thank You very much for the reply Kent. Originally, I intended to use the RAS models in order to stabilize the computation (I don't need to resolve behaviour of flow in turbulent boundary layer). But as I have run into problems with computational time, which vastly exceeds the time available, I have decided not to use any turbulence modelling. At least until I find a way how to implement some sort of fully implicit coupled block-matrix solver into multiphaseEulerFoam (or interFoam...).

Considering the interfaceCompression and RAS models - are there any problems in using them simultaneously?

[kwardle]

You can use RAS in interFoam, so functionally there is no issue.
Not sure what type of problem you are looking at, but I have implemented the LTS solution method (as in LTSInterFoam) into an older version of mpEF and it works to accelerate the solution. This is only valid/stable when you have a problem with a stable, steady-state liquid surface. If you want to go that route I could probably give some pointers.
-Kent

[tikulju]

Hi all!
I'm trying to incorporate actually a general turbulence model into the solver, so that both RAS and LES could be used. The idea is to use the mixture field U to model to turbulence. For now all I have done, is to replace the

LESModel

with

turbulenceModel

in the code and in UEgn.H-file the nuEff accordingly. Is there more, which I should take care of, or is this enough?

- Timo