Skip to content

Reply To: Particle periodicity running with MPI.


Dear Jan,

Thank you for answering my question. I am indeed using the subgrid3DLegacyFramework in my code. Although these codes are not easy to use, they do contain the functionality I need. Why aren’t you using this part of the code anymore? If this makes you uncomfortable, I won’t bother you with this issue again. I want to thank you for telling me not to add additional communication. From the results of inserting two particles, it seems that particles cannot move from one block to another. The problem may be that when particles pass through the boundary of the domain, they are not copied. Particle periodic boundaries should indeed not be called there. The question about whether the origin (0,0,0) for fluid and particle periodicity is set consistently. I set the origin of the fluid simulation domain to (0,0,0), but because of the fluid periodicity setting, an extra layer is added outside the geometric region, so it becomes (-Δx, -Δy, -Δz). I output the extend and minPhys of the particle periodicity, and these values ​​do not change in both serial and parallel modes. I only set periodicity in the x and y directions, so the extend values ​​for these two directions are correct. I want to emphasize that periodicity can run successfully in serial mode, so could this be the reason for termination in parallel mode?

[SuperGeometryStatistics3D] materialNumber=1; count=39520; minPhysR=(0,0,0.00266667); maxPhysR=(0.250667,0.0826667,0.0346667)
[SuperGeometryStatistics3D] materialNumber=2; count=6080; minPhysR=(0,0,0); maxPhysR=(0.250667,0.0826667,0.0373333)
[prepareGeometry] Dimension of the channel in meters: x = 0.250667 ; y = 0.0826667 ; z = 0.0373333
[prepareGeometry] Prepare Geometry … OK
[main] noOfCuboid=8
[prepareLattice] Prepare Lattice …
[setInitialConditions] Set initial conditions …
[setInitialConditions] Set initial conditions … OK
[prepareLattice] Prepare Lattice … OK
[main] Prepare Particles …

Best regards,