CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   AdjustTimeStep question (

danvica January 11, 2008 06:03

Hi, I need just some explanat
I need just some explanation about adjustTimeStep.

As far as I understand you need to have a Co<1 in order to obtain trustable results.
In general Co->0 => perfect

The problem arise when you don't have a "good" mesh.

In this case it can be that in some area of the mesh the computation brings to higher Co and there's nothing you can do to lower the value, apart refining the mesh or changing the specification of the problems (BC,transport,etc.)
This takes OF to use 1e-12 time step,

I believe that OF uses some interpolation techs to increase the resolution of the mesh, from the computation point of view, but it fails if the mesh is not refined enough.

That's ok. But sometimes it forces to increase the number of cells just to solve a specific area problem.
I know that it's possible to locally refine the mesh but it's not always clear where to refine it at the very beginning.
Due to the fact that I'm using a single machine my processes last about 3-4 days and sometimes OF stucks (but it's not its fault) after 2 days !

In few words: Would it be possible to have a lower limit for the time step to overpass the limit of Co ? Usually I work with Co<0.5 but in some area, for some time i can accept higher values.
When I'll have the results I can see where to, eventually, refine the mesh and check if there are differences in the area I really care to simulate.

Sorry for the long thread, I'm trying to use OF since one month and actually i'm very new to CFD in general.
I would appreciate some simple confirm/refuse of my idea, thanks.

danvica January 11, 2008 06:32

I forget to mention that I'm u
I forget to mention that I'm using rasinterfoam with water/air and would like to go with lesinterfoam.

All times are GMT -4. The time now is 16:04.