LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] Pressure with an applied E-field
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] Pressure with an applied E-field

From: Steve Plimpton <sjplimp@...24...>
Date: Thu, 15 Jun 2017 08:06:26 -0600

Not sure I undestand.  But if you apply an electric field
to a diatomic with + and - charges, then the larger the
field the more the 2 particles will want to violate the
SHAKE constraint, i.e. the farther they would move apart
or together on a timestep.  So the SHAKE constraint
has to apply larger forces to keep them at the proper
distance.  Hence the internal virial due to SHAKE increases.
Hence the pressure goes up.


On Wed, Jun 14, 2017 at 8:18 PM, Matthias Kahk <matthiaskahk@...24...> wrote:

I have been running some MD simulations using fix efield, and noticed that "something strange" was going on with the pressures that I was getting out... so I did some tests.

Consider a single, rigid, diatomic molecule, aligned with the z-axis, in a non-periodic simulation cell, at zero temperature. (I.e. all of the atoms are at rest.)

The molecule carries no net charge, but the individual atoms have charges of +1e and -1e respectively. There are no forces defined between the atoms, and the geometry of the molecule is kept constant using fix shake.

What do I do?

I run the simulation, from this starting point, using the NVE time integrator, and monitor the pressure.

What do I expect?

That the pressure should be zero, because the atoms are at rest, and there are no forces between them. Even under an applied (constant, uniform) field, a rigid dipole could experience a torque, but no net force.

What do I observe?

That the pressure returned by the "thermo" keyword scales linearly with the applied field.

Why might this be?