I am running a simulation with periodic box in all directions. I set up the temperature control with following parameters. The program (submitted to a cluster using 200 cores) terminates with the following message:
ERROR: Lost atoms: original 5410715 current 5401770
May you please help me to fix it? Thanks!
Here is the information about the input
Hi Wei
I am also new to lammps but I think you should try making the following changes:
fix 1 all nvt 300.0 300.0 0.2 drag 0.0
in this fix 0.2 seems too low, change it to 1 or more... I am using a
value of around 5-10...
timestep 0.002
From temp it looks like u r using real units, so this time step is too
low (it is in fs not ps) .. try
timestep 0.5
or more...
Navdeep
Yujie,
Generally, lost atoms means some of atoms are moving out of the simulation cell. It could be sth unphysically happened. Maybe some atoms are too closed to each other and the next time step they move too far away and out of the cell boundary. It also depends ono what kind of potential you are using. The input script you provided is not enough for a guess.
Best,
Ajing
If you don't think they should become lost, then you
need to figure out why they are moving (possibly far)
outside the box. If lost is OK, then you need to swap
the order of your thermo_modify and thermo_style commands.
From the thermo_modify doc page:
IMPORTANT NOTE: These options apply to the currently defined thermo
style. When you specify a thermo_style command, all thermodynamic
settings are restored to their default values, including those
previously reset by a thermo_modify command. Thus if your input script
specifies a thermo_style command, you should use the thermo_modify
command after it.
Steve