CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   CFL in dual time stepping with implicit scheme (

dokeun May 8, 2013 23:01

CFL in dual time stepping with implicit scheme
Dear CFDers

I think I don't understand clearly about physical global time step and pseudo time step.

The problem is that when I uselarge CFL number (10^4~10^6) using implicit scheme for pseudo steady solution, then physical time marching is too fast and the solution arrives at final time I set.

I think I may have misconception about computing physical global time step.

What I understand about dual time stepping is as below

During pseudo time stepping, local time step is computed at every iteration and update pseudo steady solution.
If the unsteady residual is converged enough(order of 10^2~10^4), then get out of inner loop and compute physical global time step extract minimum value among the local time steps.

Even though I have minimum value of local time step(pseudo time step at converged moment) contains large CFL number and it makes physical time step large.

Please correct my misunderstanding.

thank you in advance


Meanwhile I found a clue (

In the last sentence, physical time is defined only by desired accuracy. Does it mean that global time step is just user defined value?

All times are GMT -4. The time now is 20:46.