fix gcmc with full-energy

Dear Lammps Users,

I have installed lammps-16Mar18 package.

I am trying to run some gcmc simulations. I have run a test of gcmc simulation for CO2 or H2O in the example folder.

However, I got the error ‘fix gcmc does currently not support full_energy option with molecules on more than 1 MPI process. (…/fix_gcmc.cpp:721)’

Please find attached the log files for more details.

Whereas, the gcmc simulation for LJ simple fluid in the same example folder runs well.

How can I solve this problem?

I wish to hear back from you soon.

Thank you.

Regards,

Anh

log.lammps (4.64 KB)

Dear Lammps Users,

I have installed lammps-16Mar18 package.

I am trying to run some gcmc simulations. I have run a test of gcmc simulation for CO2 or H2O in the example folder.

However, I got the error ‘fix gcmc does currently not support full_energy option with molecules on more than 1 MPI process. (…/fix_gcmc.cpp:721)’

Please find attached the log files for more details.

Whereas, the gcmc simulation for LJ simple fluid in the same example folder runs well.

How can I solve this problem?

you have to find a way to reliably do fix gcmc with molecules and the full_energy activated (which is implicit when using long-range coulomb) and then implement it in the fix gcmc source code. older versions of LAMMPS did not have this check, but would cause incorrect trajectories.

so either you must not use MPI parallelization or figure out a way to run fix gcmc correctly with molecules across multiple MPI ranks.
LJ particles are not affected, because they neither require the full_energy option nor are they molecules.

axel.

If you have access to GPU’s, or simply a powerful many-core machine, you could use the packages GPU, KOKKOS (omp or gpu), and USER-OMP to gain some extra performance.

Anders

Thank you Axel and Anders for your replies.

Regards,
Anh