CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   understanding dual-time stepping (

srinath August 8, 2008 09:48

understanding dual-time stepping

I have just started reading about dual time stepping. I want to implement the Weiss and smith preconditioner, with dual time-stepping to maintain temporal accuracy.

I have 1 fundamental doubt When the multistage algorithm for the dual time stepping is implemented, how do we know that d/d(tau)(..)=0? I understand how backward difference is used for the physical time term and how the residual is calculated etc.

How many stages should there be to ensure convergence of the dual-time stepping?

Thanks Srinath

Vinayender August 11, 2008 11:05

Re: understanding dual-time stepping

you can decide wether d/d(tau)(..)=0 is achieved or not by either 1) monitering residual for each inner loop (that is each intermediate steady state problem) and this residual computation is same as you do for a normal implicit procedure to solve a steady state problem. 2) Or you can say d/d(tau)(..)=0 is achieved just by some experince that it may achieve after N no of iterations (this method helps you to get rid of additional residual computation).

Thanks, vinayender.

All times are GMT -4. The time now is 11:05.