Interactions between atom groups under different fixes

Dear LAMMPS community,
Suppose I have two groups of atoms that are not separated in space, i.e., should interact with each other.
I specify fix nvt for one group and fix nve for the other.
Will there be interaction between these two groups? Or since the two groups are under different fixes, those fixes perform time integration only for their own groups, not computing forces between atoms in their own group and other group?

Time integration fixes know nothing about how forces are computed. They just take the computed force and change atom velocities and positions accordingly.

Fixes do not generally calculate forces (unless they’re designed to). That’s the job of the “force field” styles, namely the pair, bond, and so on.

When you use a Nose-Hoover thermostat, there is a difference between thermostatting different parts of your simulation separately, due to the velocity rescaling by the thermostat multiplier. Specifically, for a LAMMPS simulation set up as

group set1 <some definition>
group set2 subtract all set1

and thermostatting settings

fix nvt all nvt ...

the center-of-mass motions of set 1 and set 2 will be correlated (as they must, to conserve momentum), whereas with

fix nvt1 set1 nvt ...
fix nvt2 set2 nvt ...

their center-of-mass motions will be relatively less correlated. This can be seen and deduced from the equations of motion.

1 Like