Minimization turn -INF

Hello everyone
    I want to simulate the seawater.
    To avoid the atoms overlay, I create the seawater orderly. I think they will become random after minimization. It looks well at the beginning, but suddenly the potential energy turn to minus. After a few step, it is -inf and the program stops without error messages. The attachment is the script and the log file.

h2o.txt (442 Bytes)

in.lammps (4.31 KB)

log.lammps (15.7 KB)

dump.atom (602 KB)

Visualize your trajectory to examine what is happening. There is likely bad dynamics due to bad initial structure or bad force field parameters.