[lammps-users] pressure fluctuations in NPT

LAMMPS LAMMPS Mailing List Mirror
1

Hi,
I am working with NPT ensemble on a polymer chain (well not really, 104 atoms: carbon and hydrogen). My timestep is 0.01 fs. When I run it for 200,000 timesteps, pressure seems to settle around 1atm which is the defined pressure. However when I extend the run to 500,000 timesteps, pressure makes big fluctuations between 124 and -160 atm. which is either too big or small, which sound unrealistic.

I am not sure what is happening after 200,000 timesteps. The amplitude of the oscilations around 1 atm gets bigger towards 500,000 timestep on pressure vs timestep plot.

My fix command in the input file is:

fix 1 all npt 293 293 10 xyz 1.0 1.0 10 drag 2.0

Thanks for any feedback

Burcu

2

Burcu,

Why is your timestep so tiny? ~1.0 fs seems more
typical.

Pressure fluctuations are normally huge in liquid
phase molecular simulations. And more so for smaller
systems.

Also, your time constants on your npt command seem
small. Instead of:
fix 1 all npt 293 293 10 xyz 1.0 1.0 10 drag 2.0
Perhaps you could try something like:
fix 1 all npt 293 293 100 xyz 1.0 1.0 1000
Or even:
fix 1 all npt 293 293 1000 xyz 1.0 1.0 10000
(maybe a bit on the large side for the time constants,
but may be necessary with your tiny system)

I hope this helps!

Paul