turbulent model
 

hello everyone

Any opinion on suitable turbulent model for highly swirling turbulent flow? Which of the model can capture the velocity structure best?

Thanks

Re: turbulent model
 

k-omega based models; RSM then is pretty good; LES ustimately.

What are you trying to model?

Re: turbulent model
 

Hello there,

Basically I'm modelling corotating disks in non-axisymmetrical enclosure. I can get a solution.. Actually irregradless of the model used, the velocity always converged to a particular solution. However this solution do not tally with the experimental result.. I can only fine-tune the model to suit the experimental result as I can actually get other results for comparison.. (The experiment was carried out by someone else). My question is what is the best model to capture the flow structure of the problem? I'm not really interested in LES as I'm modelling steady flow..

Re: turbulent model
 

Dear, I have used k-e models, but not k-w, the realizable and rng forms, givw better results in the whole by my experience. you can check out to see the best, and compare it as a good search...

Re: turbulent model
 

I would try k-w based models (such as SST for example) and then a RSM model (Sarkar Speziale and Gatski for example). The latter should show improved results; what you can do is do a re-start from one of your solution with this upgraded model and compare the evolution of the solution.

Re: turbulent model
 

Dear, I have used k-e models, but not k-w, the realizable and rng forms, givw better results in the whole by my experience. you can check out to see the best, and compare it as a good search...

Re: turbulent model
 

try this: segregated solver + RNG K-epsilon +PISO with neighbour and skewness correction Put a monitor of velocity in swirling zone. If RNG k-e does not converge switch to enhanced wall function but be sure that wall yplus is below 4.0

if still problems are there use Realizablek-e K-w models are for low reynolds no applications and SST for external flow simulations specifically. Hope this helps.