Energy, Pressure and Temperature reaching high values in few timestep

LAMMPS LAMMPS Mailing List Mirror
1

Hi lammps users,

I was trying to simulate cnt polyethelene nanocomposite in lammps. I created the initial structure in material studio and converted to lammps data file using msi2lmp tool. I used COMPASS potential in doing so. After creating the input data file, slight editing was done to the file to fit for lammps format. When I run the simulation, let it be any ensemble the system energy, pressure and temperature shoots up to a very high value and reaches inf in first four to five steps. I read in lammps forum that it might be due to bad initial configuration. I double checked the data file and it seems to be fine. Later I created a single propane molecule in material studio and used msi2lmp with class 2 compass potential. The same error occurred this time too. Since there are only 3 carbon atoms I could manually check the data file and I could not spot any error. I also checked the dump file using OVITO visualisation and it shows atoms moving out of box at the fourth or fifth timestep.

Please anyone help me out of this. I have attached the data file (propane molecule), input file and log file.

Lammps version : 3 March 2020

With regards
Abhiram

in.just_polymer (1.1 KB)

pol0.data (5.41 KB)

log.lammps (5.07 KB)

2

there are lots of things wrong with your input deck:

  • msi2lmp generates output and parameters for “real” units, not metal

  • your pair style should be lj/class2/coul/cut or lj/class2/coul/long (plus a suitable kspace style)

  • your pair wise cutoff is extremely too short (something like 12.0 angstrom would be more adequate)

  • your simulation box is far too small (just look at the extremely high pressure)

  • most of your communication and neighbor list settings are unsuitable. leave the comm settings on their defaults and use:
    neighbor 1.0 bin
    neigh_modify every 1 delay 0 check yes

  • since you have hydrogen atoms and are not using fix shake, you should use a timestep of 0.5fs or less.

also, it is generally best to not start with fix npt, if the initial structure has a high potential energy. better use fix nvt

axel.