AMBER Archive (2005)

Subject: Re: AMBER: cut-offs in non-periodic simulations

From: Andreas Svrcek-Seiler (svrci_at_tbi.univie.ac.at)
Date: Mon Oct 03 2005 - 14:15:19 CDT


Hi,
> Does it mean that it is not possible to truncate Coulombic
> interactions with for exemple a cubic spline
> when running non-periodic simulations ? What would be the reason for not
> doing so ?

As for the reason for not doing so: Just plot E_coul(r) for a pair of
charges (e.g. monovalent ions if you like units) and smoothly truncate
that (graphically, no calculus is needed). Then, again graphically, sketch the
(absolute of the) corresponding force - the negative derivative of
E_coul. Now you see a strong force in the splined (=smoothed,
interpolated) region. This is the reason - it is unphysical and worse than
the ugly 'hard' cutoff.

good luck
Andreas

             )))))
             (((((
            ( O O )
-------oOOO--(_)--OOOo-----------------------------------------------------
               o Wolfgang Andreas Svrcek-Seiler
               o (godzilla)
                        svrci_at_tbi.univie.ac.at
       .oooO Tel.:01-4277-52733
       ( ) Oooo.
-------\ (----( )--------------------------------------------------------
         \_) ) /
               (_/

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu