trouble using rigid bodies in simulation (lost atoms)

Dear LAMMPS users,

  I am doing a MD simulation for CO2 in porous carbon models. I model
the CO2 molecule as rigid. However, I am getting "Lost atoms" as
error. I am attaching the log file. I will appreciate if somebody can
provide some help.

Best Regards,
Surendra

co2-cs1000a-24cooh.log.lammps (107 KB)

Dear LAMMPS users,

  I am doing a MD simulation for CO2 in porous carbon models. I model
the CO2 molecule as rigid. However, I am getting "Lost atoms" as
error. I am attaching the log file. I will appreciate if somebody can
provide some help.

you have posted the same question before and have received suggestions
on how to resolve this.
before asking the same question again, please also describe what you
have done on your own to address the problem.
as far as i can see, your input is still cluttered with irrelevant,
commented out commands, your timestep seems large and your langevin
thermostat time constant small. the latter can hide other problems,
e.g. bad force field parameters.

axel.

Dear Axel,

  many thanks for your reply. There was some problem with my data
file. Some atoms were outside the simulation box. I added the
appropriate image flags. The simulation was running fine for NVE
ensemble.

However, when I added langevin thermostat it gave problem. I am using
a timestep of 0.5fs and my langevin time constant to 1.0 to dissipate
the energy quickly. Do you suggest to use a larger langevin time
constant.

Many thanks,
Surendra