[gmx-users] COM motion removal problem

[Erik Lindahl]

Christoph Freudenberger wrote:

> Erik Lindahl wrote:
>
>> Any special reason why you don't remove it every step? I would at 
>> least try that and seem if that's the problem.
>
>
> I'm also not quite sure how to handle this. The manual is not really
> unmistakable about this:
>
>
> | nstcomm: (1) [steps]
> | frequency for center of mass motion removal>/dd> motion removal
> | (should only be used for vacuum simulations)
>
> How is that meant?

It means that we remove the center-of-mass motion every "nstcomm" steps. 
So, if you set it to 10000 you could get quite a buildup of kinetic 
energy during the first 9999 steps, and then it is reset back.

In most cases I'd recommend to remove it every step, i.e. setting 
nstcomm to 1.

Cheers,

Erik