CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   why I can't get Second order convergence (http://www.cfd-online.com/Forums/fluent/28541-why-i-cant-get-second-order-convergence.html)

mike huang May 18, 2001 21:41

why I can't get Second order convergence
 
I use fluent to compute a ejecting flow in pipe, the main flow is about 4 mach.I can only make it converge to the first oder upwind scheme,but can't make it converge to the default second order upwind scheme. please give me some advice. thanks.

John C. Chien May 18, 2001 22:11

Re: why I can't get Second order convergence
 
(1). Are you saying that you are computing the Mach 4 flow in a pipe and you can get converged solution with the first-order upwind scheme only? (2). What is your Reynolds number?, mesh size and distribution?, turbulence model? boundary conditions? (3). How about the shock structures in the pipe? I think, these will affect the stability of the numerical solution. The first-order method will give you higher numerical viscosity to stabilize the flow.

Volker Pawlik June 8, 2001 06:04

Re: why I can't get Second order convergence
 
If you mean residuals lower than 1E-3 with "converged", then you should try to solve the problem with the unsteady solver. This mostly helps due to the instationary behaviour of jets if you have more than one. They reason is that jets are interacting which leads to a periodic phenomenum.


All times are GMT -4. The time now is 19:18.