[gmx-users] Shift function: rcoulomb-switch

[Mark Abraham]

On 19/02/2012 2:01 AM, Juliette N. wrote:
> Hi,
>
> I am trying to figure out the best setting for cut offs and r_switch 
> for my system based on shift functions.
>
> coulombtype         =  Shift
> vdw-type                =  Shift
> rcoulomb-switch     =  1
> rvdw-switch         =  1
>
> ;               Cut-offs
> rlist               =  1.35
> rcoulomb            =  1.1
> rvdw                =  1.1
>
> The above setting works well but just wanted to make sure setting
>
> rcoulomb-switch     =  1
> rvdw-switch         =  1
>
> is not technically wrong. The default rcoulomb   is 0 and setting it 
> to 1 does not affect my result. Is there specific reason for default 
> value of rcoulomb-switch     =  0       or I can leave 0.1 nm distance 
> between rcoulomb-switch and cutoff of 1.1 as above?

Your force field should be dictating your settings. If you do something 
else, then the burden of proof that it makes a reasonable model is on 
you. :-) The position of the switch changes the shape of the function. 
What that does to the model is anyone's guess.

Mark