[lammps-users] Simulating One Atom

LAMMPS LAMMPS Mailing List Mirror
1

Dear LAMMPS Tech-Support,

I am trying to understand how LAMMPS works by going through some simple simulations. One of these is simulating a single particle in a box undergoing a random walk. My files are copied below. When I try simulating this, I get a thermodynamic output with no potential or kinetic energy (even though I specified a velocity for the atom AND I have placed it in a langevin fix! Can you tell me whether simulating one atom has some special feature that I need to concern myself with? When I add a second particle, it seems to work fine (in terms of energies at least anyway).

Thanks!

Alvin

2

There is no potential energy for a single atom. The temperature
is computed for N atoms by subtracting off 3 degrees-of-freedom
(for the center-of-mass motion), so when N=1, that isn't what
you want. The compute_modify extra 0 command should fix that.

Steve

3

By default, the "velocity create" command sets the center-of-mass momentum
to zero. If you have only one particle in the box, its velocity will be
zero.

  Lutz