[gmx-users] zero cut offs

[Mark Abraham]

On 14/02/2012 6:26 AM, Juliette N. wrote:
> Hi all,
>
> I have two simple question on in vacu calculations.
>
> 1- Say we start from rc=1, and go all the way to rc=0.1 then we say 
> cut off has been reduced from 1 nm to 0.1 nm meaning that atoms fall 
> within 0.1 nm interact with each other. Now when we set rc=0, why dont 
> we consider this as zero cutoff and that No interaction is cacluated 
> for in vacu simulation. When one sets r_vdw= 0, r_coulomb=0, why 
> people call this infinite cutoff and not zero cutoff (=no nonboded 
> interaction calculated).

Because it is defined that way. See manual section 7.3.9. In theory, you 
could have another .mdp variable "cut-offs = yes/no" to serve this 
purpose, but since the value zero serves no useful purpose, it is used 
in the above way.

>
> 2-For calculation of heat of vaporisation why do we need single 
> molecule in vacu? why cant we leave cut off as in liquid phase?
>
> 3- Also why do we turn off PBC for in vacu runs?

Usually efficiency. Finding the nearest neighbours takes time, and if 
there are few atoms at all (as is often the case with vacuum 
simulations) it can be cheaper just to compute all the interactions.

Mark