Hi, segmentation fault. I do not know why. I want to know whether you ran the simulation with all the files provided by me. And i want to know where can i find the version 30 June lammps. Looking forward to your reply. Thank you so much.
ffield.reax.CHNi (5.85 KB)
lammps.in (655 Bytes)
multilayer.dat (40.3 KB)
slurm-2217511.out (969 Bytes)
Version 30 June 2020 is available as “Development” version on the Sandia Download page.
All (older) tar files are available at: http://lammps.sandia.gov/tars
You can also download archives stable and patch releases from github at: https://github.com/lammps/lammps/releases
or just clone the git repository and check out the stable or unstable branch or the individual patch release you desire.
More details about the various options to download and install LAMMPS are in the manual.
Yes, I ran your input unmodified. Please note that your system is large for a system with ReaxFF potential, so the estimator for the size of the neighbor list storage can easily overflow with 32-bit integers. Unlike the core of the LAMMPS code (which uses 64-bit integers consistently for integers that may overflow), the USER-REAXC code has not been designed and tested for such situations. I have just added an additional check to the USER-REAXC code to error out when an overflow happens.
This should not happen when using more than 4 MPI ranks with the unmodified input.
Axel.