[gmx-users] Coul-LR in PME and in Reaction Field

[Berk Hess]

>From: pascal.baillod at epfl.ch
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: gmx-users at gromacs.org
>Subject: [gmx-users] Coul-LR in PME and in Reaction Field
>Date: Wed, 26 Apr 2006 12:13:38 +0200 (MEST)
>
>Hi everybody!
>
>Some more electrostatic questions, if I may..
>
>1) I think I understood the meaning of the coul-LR term now, but would like 
>to
>be sure.. The Coul-LR term is only printed in PME runs if rcoul > rlist. In 
>PME,
>coul-LR would be the (short range) real sum term computed between rlist and
>rcoul, whereas interactions after rcoul would be computed in reciprocal 
>space,
>in a term that is only computed for the entire system (coul. recipr.). 
>Therefore
>it is not possible to get coul. recipr. for part of the system..
>

Yes, but with PME you should preferably use rcoulomb=rlist.
I have added a check for this in the grompp of version 3.3.1.

>2) Would coul-LR be the long range "interaction to dielectric x 
>environment"
>contribution, in the reaction field method?

No.
This term contains the interactions between rlist and rcoulomb.

Berk.