[Thomas Baumann]

Hi all,

I have implemented an algebraic stress model for 2D turbulent flows near equilibrium in OF 1.6. It's implemented with an explicit formulation of the Reynolds-stresses and the equations are orginally from the paper: "Towards the developement of second-order closure models for non-equilibrium turbulent flows" from Charles G. Speziale and Xiang-Hua Xu. It's an explicit version of the SSG model.

Here's the link to the paper:

http://www.sciencedirect.com/science...8be3a1fd1158ee

But it's a basic model for 2D flows near equilibrium solved with the equations 10, 13 and 14 of the paper. I implemented it with respect to the standard k-epsilon model, so you will need wallfunctions for k, nut, and epsilon, but not for the Reynolds-stresses.

You can download it here:
http://www-isl.mach.uni-karlsruhe.de/~hi204/OpenFoam_Forum/kEpsilonASM.C.tar.gz

If you have any hinds, tips or ideas for improvement let me know.

Regards Thomas

[cfdmarkus]

Hi Thomas,

I am just wondering whether you have looked into alternatives to the EASM you have used.
Other EASMs solve a non-linear equation for the production to dissipation ratio and hence avoid the singular behaviour in non-equilibrium flows. However, I am not sure how well they perform for strong non-equilibrium flows. Any Experience on that?

MArkus

[Thomas Baumann]

Hi Markus,

I have no big experience with EASM until now, but I hope this will change in future. In the next steps I want to implement other ASM, too, especially some for 3D ,non-equilibrium, low Reynolds flows and later with turbulent heat fluxes (Reynolds flux models).

Thomas

[Thomas Baumann]

Hi all,

I have tried to implement another ASM model from: "On explicit algebraic stress models for complex turbulent flows" from Gatski and Speziale, Journal of Fluid Mechanics. It's the regularized algebraic stress model on page 70.

http://www-isl.mach.uni-karlsruhe.de/~hi204/OpenFoam_Forum/SSGRegularized.tar.gz

Solving with this model a square duct you get the secondary flows at the edges, but they rotate with the wrong direction. The normal velocities nearby the walls are wrong. Maybe I need a boundary condition for them?

wrong direction:
http://www-isl.mach.uni-karlsruhe.de/~hi204/OpenFoam_Forum/square_duct.png

Changing -2(S_ik*S_kj) to +2(S_ik*S_kj) changes the secondary flows to rotate in right direction.

right direction:
http://www-isl.mach.uni-karlsruhe.de...duct_right.png


Has anybody an idea what's wrong?

Thanks very much, Thomas

[Eliasjal]

Hi Thomas, I'm actually not very familiar with openfoam but I would like to use Ansys fluent for simulating a very similar case and if possible I would like you to convert the data file or even the mesh file ( that you linked here in the zip file) from Openfoam to Ansysfluent ..... thank you so much for your help
Elias,