Nose-Hoover Damping Time

LAMMPS LAMMPS Mailing List Mirror
1

Hello,

I’m curious about the suggestion for the Nose-Hoover damping time to be around 100 time steps. From what I’ve read, this damping time should be around the system’s smallest vibration or relaxation time, but not be smaller. Is the assumption behind your suggestion that people are setting their timesteps to about 1/20 of that smallest relaxation time for energy conservation purposes, and then you multiplied that by 5 since it’s better to err on the too-high rather than too-low side?

Just wanted to check my understanding since I’ve seen lots of literature about basing the damping time on the system’s largest characteristic frequency, but I haven’t come across literature about basing the damping time on the timestep.

2

The suggestion wasn’t specific to Nose/Hoover. Ditto for Langevin, etc.

I don’t think there is a “right” answer to the Q. You don’t want the

damp time too small or the thermostat is forcing the system unphysically

(and may go unstable). You don’t want it too big, or the temp may drift,

esp if there are other driving forces. I suggest making it big if the system

temp behaves well. 100 steps often works well in practice.

Steve