CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   k-w Turbulence model implementation (

suneesh November 20, 2005 11:29

k-w Turbulence model implementation
Dear Friends,

I am working on the implementation of k-w turbulence model for 2d flows and I have a few problems, which I would like to clear. The turbulence model of Wilcox(1988) has been implemented by me in a coupled fasion, with implicit time stepping.

1. The code blows off after reaching a residue of 1e-4 on fine meshes.

2. k and w are becoming negative.

There are a few references that talk about limiting k and w, but the procedure is not well explained. I believe that the blow up of the code and the simultaneous degradation in the contours is because of improper limiting. It would be a big help if anyone could clarfiy these doubts and/or point to references that contain substantial discussions in this regard.

Thanks in advance, Suneesh

pkm November 21, 2005 03:15

Re: k-w Turbulence model implementation
normally the k starts attaining -ve value first.then obviously solution blows up. adopt linearisation of k -eqn. also ensure attainable convergence of the continuity eqn.

peter.zhao November 23, 2005 11:12

Re: k-w Turbulence model implementation
There are a lot of lower limits on K and omega to prevent them below unphysical value. For example, From, CFL3d imposes the below limit on omega and K: If either nondimensional omega or non-dimensional K drops equal to or below 1.0e-20, they will be reset to 1.0e-20. From, WIND-US software sets kmin and εmin to 10'10 and 10'12, respectively. You can also find a more complicate formula to limit omega and K from Which way is the best rests on the problem you solve

suneesh November 23, 2005 12:19

Re: k-w Turbulence model implementation
Dear Peter Zhao and pkm,

Thanks for the suggestions. I shall check them out. I would also like to know if some upper limits are provided for k and w, and also like to know if there are limits on the eddy viscosity.

Thanks in advance,

Regards, Suneesh

FELIPE November 23, 2005 18:35

experimetal or numerical data channel flow
Dear Friends

I am working on the simulation of turbulent mixed convection in vertical channel flow , I have validated my code for forced convection and the natural convection in a tall and square cavity, I want to validate my code for mixed o natural convection in the vertical channel flow, Please anyone can send to me experimental or numerical data for the validation code.

sincerely thank a lot

All times are GMT -4. The time now is 23:55.