Why openfoam can not write time folder when the processor number increases?

yzf1215 · September 26, 2013, 21:51

When I was using 32 processors for parallel computing, openfoam can write time folder as I defined in controlDict. However, when I increased the processor number, say, to 64, it stoped writting and the log is not written after printing the coefficients.

musahossein · September 26, 2013, 22:26

That should not be the case. Openfoam can deal with over a 1000 processors. What is your mesh size - if your mesh size is not sufficiently large then this will impede the processing. You should start with smaller number of processors and then add a few processors incrementally so that you can see how the mesh is being decomposed and at which threshold the solver stops writing to the log.

September 26, 2013, 21:51	Why openfoam can not write time folder when the processor number increases?	#1
yzf1215 Member Zifei Yin Join Date: Sep 2012 Location: Shanghai & Ames Posts: 33 Rep Power: 13	When I was using 32 processors for parallel computing, openfoam can write time folder as I defined in controlDict. However, when I increased the processor number, say, to 64, it stoped writting and the log is not written after printing the coefficients.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Multiple floating objects	CKH	OpenFOAM Running, Solving & CFD	14	February 20, 2019 09:08
pisoFoam with k-epsilon turb blows up - Some questions	Heroic	OpenFOAM Running, Solving & CFD	26	December 17, 2012 03:34
New OpenFOAM Forum Structure	jola	OpenFOAM	2	October 19, 2011 06:55
OpenFoam doesn't write the solution time folders	jose21	OpenFOAM	2	December 8, 2010 09:44
Modeling in micron scale using icoFoam	m9819348	OpenFOAM Running, Solving & CFD	7	October 27, 2007 00:36

September 26, 2013, 22:26		#2
musahossein Senior Member musaddeque hossein Join Date: Mar 2009 Posts: 309 Rep Power: 18	That should not be the case. Openfoam can deal with over a 1000 processors. What is your mesh size - if your mesh size is not sufficiently large then this will impede the processing. You should start with smaller number of processors and then add a few processors incrementally so that you can see how the mesh is being decomposed and at which threshold the solver stops writing to the log.