cavity flow
 

Hi, I am checking the turbulence model in CFX for solving the cavity flow. If I select k-epsilon model or any model with a scarable wall function, I don't get a good solution. I get the good solution for any turbulence modle with the automatic wall function. calcuration conditons are the followings; Re=1000. rho=1. mu=0.001 H=1m V=1m/s

what's the problem? how can I change the scarable wall function to standard wall function in CFX?

sincerely yours

Sikida

Re: cavity flow
 

For such low Reynolds number the only thing you should do is to generate a two zonal layer mesh with y+ near 1 and implementing a low-Reynolds turbulence model, which could support the two-layer approach. A good choice is the k-w or even better SST (if you increase ther Re number in other runs). It is not woth using wall function meshes with these Re. Finally, do a further research whether your flow is turbulent, which seems not to be from your data! Good luck

Re: cavity flow
 

Dear Steve

Thank you for your advice. I misunderstand the turbulece model in CFX. I want to know the difference of applying the turbulence model(k-epsilon, k-omega and so on) to cavity flow. If the turbulence model doesn't work because of a laminar flow condition, I think that it's result is similar to the solution of a laminar flow condition. In the time of applying the turbulence model(SST, or k-omega) with the option like "automative", I got the solution similar to it in a laminar condition. but, in the time of applying to k-epsilon model, I can't reasonably a solution. I am wondering about it. what's the problem? What do I misunderstand?

sincerely yours

Re: cavity flow
 

k-e is not able to produce a proper solution for your problem. It is only valid for highly turbulent flows Re>5e4. Look into a general book about flow modelling to find out why.

Gert-Jan

www.bunova.nl

Re: cavity flow
 

Dear Gert-Jan Thank you for your advice.

sincerely yours