CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS (http://www.cfd-online.com/Forums/ansys/)
-   -   Convergance problems with FLOTRAN (ANSYS multiphysics) (http://www.cfd-online.com/Forums/ansys/69578-convergance-problems-flotran-ansys-multiphysics.html)

Cirion0000 October 27, 2009 21:43

Convergance problems with FLOTRAN (ANSYS multiphysics)
 
I know this may seem like a vague question at first but bear with me. I am having serious woes trying to get any convergant solution for just about any geometry involving FLOTRAN, especially turbulence models. Specifically, I'm trying to model the lift/drag forces on various airfoils. No matter what solution algorithm I use, no matter how many nodes I have (if anything, more nodes makes it harder...), and no matter how I play with the stabilization, it can never converge. My models are not even that complicated. I can't even get a solution for this simple geometry:

Simple box wind tunnel
Circular "airfoil"
Inlet parameter of 1m/s of air, water, or whatever (have tried both)
Top/bottom BC's of 0 velocity
BC's around the circular "airfoil" of 0m/s
BC of exit line of 0 Pa
Mesh: Quad (Fluid 141 2-D)

Like I said, I've tried different solution algorithms (the tutorials that come with ANSYS multiphysics say that SUPG can be hard to converge... well the others do not converge either), and tried just about every combination of stability parameters. If it is this hard just to get a single analysis of a single shape (simple one at that to work) then I am going to kill myself before I can get the analysis done that I really need to.

Do I have a bad mesh? If so what's wrong with it? Perhaps I should send a copy of my test file so someone can analyze it.

I know my boundary conditions should work because I have gotten convergant solutions with laminar (but I have had lots of woes with laminar, even.)

Is the solver in ANSYS really that bad or am I just bad at setting up the solution? I can't imagine having to analyze dozens of airfoils if it takes countless hours of trying different stabilizations to get a single one to converge. It takes long enough, if you have hundred of thousands of nodes... only to diverge...

Thanks.

mechanicaldesign December 11, 2009 02:39

Convergance problems with FLOTRAN (ANSYS multiphysics)
 
Hi Cirion0000,

If you are agree to send me an power point with some explicit slides with your problem, I will try to help you. Your slides must include all input data (please not send me your error). If you are agree, send me only the geometry of the model which must be study. I waiting your answer here, and if you are agree, you can send me all what I specify here, on my personal email.

Best regards.:)

P.S. I using ANSYS Classic 11.


All times are GMT -4. The time now is 19:31.