Learning Forum

[Patrick Kirby]

Hello,

I've been looking into a mixing tank problem in which the tank is not baffled and swirlling type flow is likely. I've played around with different turbulent models in particular the RSM and LES models and have found them to work rather well. My issue is that as I look to move into multiphase flow LES seems to be uncompatable with several multiphase modules and RSM also has some drawbacks. Is there a better (more robust) option for these kind of mixing problems? Thanks

[Nikhil Narale]

Hello, 

 

Which multiphase model are you considering?
"RSM also has some drawbacks" : Would you mind elaborating a bit. 

You may also want to explore DES turbulence model, which is compatible with Mixture and VOF multiphase model. DES resolves the large-scale turbulent motions using LES and models the smaller, unresolved turbulence using RANS.

Thanks!

Nikhil

[Patrick Kirby]

Hello Nikhil,

Thanks for the response.

Relating to RSM the issues that I have found are :

1. The computational cost
2. The robustness of the model
3. Underestimation of the turbulent energy dissipation

I was considering DES and also investigating using k-omega GEKO as alternatives.

Thanks

Pat 

[Patrick Kirby]

Hi again Nikhil,

I just realized I did not address your full question. My multiphase model will include a liquid and a gas phase. There may be more phases down the road, but I need to get the two-phase models going first. I was thinking of applying population balance methods ( QMOM, inparticular) to model the gas phase. While buidling the model, I discovered that the population balance models are not compatable with LES (which makes sense).  Therefore, I was moving towards making use of DPM to incorporate the gas bubbles in the model if I use LES or DES as you mentioned.

 

Thanks

 

Pat

[Rob]

RSM should be fine. The issue tends to be that it's transient, but most multiphase models are too so it's not a problem. The main point about multiphase turbulence is the phase/mixture level effect: do the two phases have very different Reynolds Numbers, and does the second phase promote or damp turbulence in the first. There's probably still a PhD or few in this field so "best practice" may change between versions. 

[Patrick Kirby]

 

Hello Rob,

Many thanks for the response. Indeed its a very interesting/multitiered problem.

Thanks

Pat