Reply To: Particle periodicity running with MPI.
Dear Rookie,
how many blocks do you have in your system if you run the simulation sequential? Try using as many blocks as you do, when using the parallel system. Does the same problem occur in this case?
It seems that you have a simplified case to work on this problem. Can you perhaps add an output of the particle position and cuboid at the end of applyBoundary? Does that align with your expectations? That could narrow down the possible causes.
In general, I still don’t see how the periodic boundary could introduce the behavior shown in your image, because for me it seems that the periodic boundary shouldn’t be located at the intersection of these blocks. If it is, that might already be the problem that should be solved.
Perhaps I misunderstand your setup. Could you provide some details or perhaps a sketch of the basic domain? Where are the periodic boundaries? How does the particle move in the images above, just from left to right?
Is the fluid also periodic?
Best regards,
Jan