CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   diverge with low Re numbers (

hui April 9, 2007 05:43

diverge with low Re numbers

i am using RSM to calculate turbulent Couette Flow. My program works well with high Re numbers but diverged with low Re numbers.

Anyone has such experences? how can the Re number influent the performance of the program?

Phil April 10, 2007 04:36

Re: diverge with low Re numbers
don't have experience of exactly this but sounds like possibly the actual physics become unsteady at low Re and so you need to use the explicit solver. Run an unsteady simulation and compare the velocity contours at different time steps. You might see a wake that is changing position with time!

Otherwise, if you want a steady solution try bringing in relaxation factors towards the end of the solution AFTER the solution starts to become unstable but only bring down the relaxation factors a little for THE UNSTABLE EQUATIONS because it will take much longer to converge.

Please note- IF YOUR SOLUTION IS STILL CONVERGING AND ISN'T TOO UNSTABLE THEN IT MAY BE WORTH JUST LETTING IT CONVERGE ANYWAY. If this worries you then try it with and without the RF's. You may get the same result but in half the iterations by just leaving it! but depends whether the solution is still converging or not!

hope this helps.

hui April 10, 2007 07:16

Re: diverge with low Re numbers
Thank you very much!!

I will try again :)

All times are GMT -4. The time now is 09:54.