CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM (
-   -   runTimeModifiable (

Djub November 12, 2012 06:10

Hi everybody,
runTimeModifiable seems to not work.
I am running pimpleFoam, and my calculation was running since a week with a maxCo of 10.
I changed it slightly (to 11). Then it started to do bad things. For example, my probes file are bad: at the midification time, the line is incorrect (the line is interrupted by another, and a nother, and another... and then it is OK again). And my fields start to be ugly, and illogical: before the run time modification, my maxCo was 10, with a deltaT=0,08 s. and my computer was needing 40 minutes to calculate one second signal. And after the modification (allowing maxCo to be 15), deltaT was 0,09s and time speed about 2H by second!

runTimeModifiable doesnot work in my case. Or I did bad things?

akidess November 12, 2012 08:38

Does the case behave any better if you start with maxCo 15? I'd guess not, and that maxCo 10 is the best you can do for your case.

Djub November 12, 2012 09:45

hi akidess,
I just started the same calculation, from the point when I tried to runTime modify my controldict.
And strating again, it works fine, with maxCo=15. thus, it is not a problem of maxCo, but a problem of modifying on run time the dictionnaries.

Djub December 13, 2012 11:27

OK I've got it!
In fact, I was doing really bad things: I was running something like
mpirun -n 6 pimpleFoam
instead of
mpirun -n 6 pimpleFoam -parallel

The first line is completely idiot: it launches 6 times the same calculations! BUT working on the same files ! Thus, when I was changing the controlDict, the different processes didnot read it at the same time so it make horrible things. But I was not because of runtimeModifiable ; it was because of the strange behavior of MPIrun lauched on the non-parallelized pimpleFoam application...

Thanks for help!

All times are GMT -4. The time now is 03:42.