CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Convergence second order (https://www.cfd-online.com/Forums/fluent/46322-convergence-second-order.html)

 KS October 15, 2007 04:13

Convergence second order

Hi, why when increase the proportional number of cells in my mesh the problem does not converge in second order? thanks

 Woo Meng Wai October 15, 2007 08:41

Re: Convergence second order

Hard to say why this happens. Have you tried interating in 1st order first and then switching to second order?

But what is important to take note is that having higher number of cells does not mean giving better results. From some of the simulations that we are doing, having too many cells might actually result in unrealistic simulations.

Might be good to do a mesh independence test to determine which range of 'cell number' is suitable for your case.

 KS October 15, 2007 08:51

Re: Convergence second order

hi, thanks so much! Yes, I started with first order and later I passed to second order but he did not converge how I make this mesh independence test?

 Woo Meng Wai October 15, 2007 20:02

Re: Convergence second order

To test for mesh independence:

One way is to just start off with an appropriate mesh size that you think is appropriate. Get a converged solution. Plot a certain parameters for comparison which is relevant to your case.

[Eg. In my case for a cylinder-on-cone spray dryer, I plot the axial velocity at a certain elevation for the whole cross section.]

Now reduce / increase the mesh size and repeat the procedure. Compare the plots of the parameters and see if there are any changes. If the changes is minimal, ~5%, then it is mesh independent.

P.S. Not sure on the convergence of your case. You can try adjusting the underrelaxation factors.

 All times are GMT -4. The time now is 18:47.