My case runs very slow and it can not run with mpi(mpirun noticed that process rank 0 with PID 0)

MFiX
#1

I simulate a conial spouted bed with 15000 particles but it is estimated that 27 days is required with one core to simulate 15 s.

When I use mpi run, the case always broke showing the following messge.
“mpirun noticed that process rank 0 with PID 0 on node gadi-cpu-clx-2303 exited on signal 9 (Killed)”.
I tried differnt core division, such as 222, 181, 242, but the error stays.
However, it can runs well if i use only one cores.
I do not even have one clue about this problem.
MIT.zip (5.9 MB)

#2

Start with a time step of 1E-3s and set DT_Max=0.01s (you are starting with DT=1E-5s and DT_MAX=1E-4s). Also set the fluid normalization (norm_g) to 0 in the numerics pane.

With these settings I can run your simulation (mpi run) with no error on 40 cores (2x10x2 partition) in 41 minutes.