[gmx-users] removal of COM

[Mark Abraham]

Sun Joo Lee wrote:
> Hello Users
> 
> I am trying to measure the drift of molecules in the system.
> When I ran the simulation,  the removal of center-of-mass of the whole 
> system took place at every step.
> 
>  From the mailing list, I've found that removal of COM works by making 
> the velocity of the COM to be zero by subtracting the velocity from 
> every particle velocity.
> 
> It would be very grateful if you can help me to make two points clear;
> 
> 1) Would the magnitude to be subtracted  from every particle velocity  
> be the same (or different depending on the mass of the particle)?

The same. The COM is the mass-weighted center of geometry, so COM 
velocity is already the rate of change of the displacement of the 
mass-weighted center of geometry.

> 2) Could removal of COM make one type of molecule move in an opposite 
> direction of the other molecules that move in one direction to keep 
> maintain the COM of the system at its original position?

Sure. Consider a system with three particles moving to the left with 
speed 3 and one particle moving to the left with speed 1.

Mark