[gmx-users] Question about center-of-mass removal

[Mark Abraham]

> Dear all;
>
> I have very simple question about center-of-mass removal,
> which would be very easy for the expert like you.
> When we dealing with the system whose center-of-mass is moving
> (e.g. initially the fluid is at one side, and as time goes on,
> it spreads to other direction), do we still have to remove the
> center-of-mass motion (e.g. nstcomm = 1) ?

Unless you're building in some other asymmetry, your diffusion will be in
both directions, so your COM motion should still be zero. You can see this
clearly if you consider the separate COM of your two fluids - they aren't
changing on average. Thus I guess that enforcing COM is probably OK.

Mark