CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   NUMECA (http://www.cfd-online.com/Forums/numeca/)
-   -   A puzzle about turbulence of B.C. in inlet (http://www.cfd-online.com/Forums/numeca/51037-puzzle-about-turbulence-b-c-inlet.html)

Yijin Li April 26, 2007 22:05

A puzzle about turbulence of B.C. in inlet
 
Hi All,

In the Help of NUMECA/Fine7.1, when defining turbulent Boundary Conditions, the help suggests that Vt/v=1 to 5 in internal flows on Spalart-Allmaras Model, but in K-epsilon Model Muit/Mu =50.

I followed this advice. In NUMECA/Cfview, the value of Turbulent Viscosity(Mut/Mu)in inlet boundary in K-epsilon model is 10 times of that in Spalart-Allmaras model.

what's wrong? Is the value of Vt/v between 1 to 50??


Yijin Li April 27, 2007 20:05

Re: A puzzle about turbulence of B.C. in inlet
 
How to make a consistent result with Spalart-Allmaras Model and k-epsilon Model? Can someone give me some advices? Thank you!

Andy Hill April 29, 2007 16:31

Re: A puzzle about turbulence of B.C. in inlet
 
Hello,

If you are using the S-A boundary condition, you don't define nu_t/nu, you define nu_t itself. In my calculations I have to define a nu_t variable in CFV using mu_t/mu and density, temperature, etc. I have also learned that using nu_t=0 at walls will cause a divergence before even the first iteration of a solution. Ironically, the S-A model specifies nu_t=0 at a solid wall, but nu_t~1e-9 will even work.

I'm not sure if this will help, but I struggled with matching these inlet conditions when I made my first attempt at this. Best of luck.

Yijin Li May 2, 2007 21:31

Re: A puzzle about turbulence of B.C. in inlet
 
Thank you!


All times are GMT -4. The time now is 12:27.