Dear all,
I’m running the LUNAR EPON862 example from the official GitHub repository (EXAMPLES/EPON862) using:
- LAMMPS version: 29 Aug 2024
- Operating system: macOS 15.4.1
- PCFF
After some simulation steps, I get the following error:
ERROR on proc 4: Bond atoms 11544 11567 missing on proc 4 at step 41630 (…/ntopo_bond_all.cpp:59)
Last command: run 200
I checked the dump file just before the crash, and it seems that the atoms involved (and nearby atoms) suddenly start to move erratically — getting very close or far apart very fast, which I suspect causes the instability.
To try fixing it, I already attempted several things:
- Reducing the timestep
- Lowering force field stiffness (K values in the data file)
- Using run every to progressively update bond topology
However, the issue persists.
Also, I noticed something strange: when I launch several simulations (same input), some of them work, or crash at different steps. The error seems to have some randomness to it, which makes it harder to debug.
Has anyone encountered a similar issue with LUNAR outputs or this specific example? Any suggestions for stabilization strategies or known issues with this setup?
Thank you very much in advance for your time and support.
data.lmp (324.2 KB)
in_reacter.lmp (3.7 KB)
slurm.out (100.3 KB)