[Adyk]

Hi there,

i'm not sure if it's even possible: I'd like to run a case for a long time (in future probably up to 4 weeks, so that the endtime is around 2.400.000 seconds). The details in between are not that important to me, so the timesteps can be large.

In controlDict I tried to adjust the Courant Number maxCo (set it to 1000.000), timePrecision, writeInterval and deltaT, but every time i got errors.

So my questions are: is it possible at all to run a case for such long time interval (with large timesteps/writeInterval) and if yes, how do i do this?

Hope you can help me out!

[linnemann]

Hi

The best solution to large timesteps is to use the pimple algorithm the "right" way.

https://openfoamwiki.net/index.php/O...hm_in_OpenFOAM

https://holzmann-cfd.com/en/openfoam...rithm-analysis

https://holzmann-cfd.com/en/publicat...nd-openfoam-v7 (Worth the €9.99)

Most transient in-compressible tutorials utilize pimple, but in PISO mode.

Using a Co = 1000 might be a stretch.
You can maybe safely utilize a Co of 50 -> 200 depending on your flow type.

Beware this is only for in-compressible transient simulations. If you are doing combustion, interface flow (VoF) etc. a Co below or equal to 1 would still be needed.

[Adyk]

Hi Linnemann,
thanks for your quick reply!

Well, I'm using rheoInterfoam (from rheoTools), it's a solver for incompressible, two-phase flows (VOF). I'm not sure, but I think it's using pimple:

in "fvSolution":

 
PIMPLE
{
    nInIter         1;
    SIMPLEC         true;
    
    nCorrectors     1;
    nNonOrthogonalCorrectors 0;
}

Hm, oh, did not know, that the Courant number in VoF should be at max. 1.. What is the reason for that? Because with the higher couant number (e.g. 250) it's working fine

[linnemann]

Well it depends on the level of accuracy you expect from your solution.

https://www.researchgate.net/post/is...gh_in_openFoam

So even though you "can" have a high Co, the trade-off is between accuracy/stability and simulation time.