[lammps-users] NPT

I’m trying to simulate a 9000 atoms system (of polypropylene) using npt dynamics. I’ve searched messages that have been posted on this forum before about the pressure fluctuations in the Lammps-NPT algorithm, however I get around 1000 atm (I’m using real units) of fluctuation and it makes no sense to me. This is the part of the input file to lammps with the npt fix (I’m using a higher temperature for equilibration, I later reduce the temperature using alternate steps of unfix-fix commands).

Is there a “numerical-stability” related issue involved?

I would appreciate any pointers you have for me.


initial velocities

velocity all create 1000.0 4928459 mom yes dist gaussian


fix 1 all npt 1000.0 1000.0 100 xyz 1.0 1.0 500 drag 2.0

fix 2 all rdf 1000 rdf.dat 100 1 1 1 2 2 1 2 2


timestep 1.0

thermo 100

thermo_style custom step pe etotal press vol temp

So long as the average of your fluctuations is 1.0 (the desired pressure),
then you're fine. Pressure does fluctuate by huge amounts in a small
MD simulation of 9000 atoms. That's not a LAMMPS thing; it's due to
the definition of instantaneous pressure.