Hello,
I have a problem related to the following pair-style
pair_style lj/gromacs/coul/gromacs 9.0 12.0 0.0 12.0
A few months ago I have produced a few trajectories with this pair-style
[LAMMPS (17 Nov 2010)]. I am now [LAMMPS (21 Mar 2012)] just
restarting them but I found discrepancies in potential non-bonding energies.
Exploring the mail list archive I understand that a bug was discovered, related
to a cut-off, but apparently the bug was fixed before I used the
pair_style it for
the first time. Any hints about possible additional problems?
Yours sincerely,
Stefano
Stefano Mossa
CEA Grenoble - INAC/SPrAM
17, Rue des Martyrs
38054 Grenoble Cedex 9
France
E-mail - [email protected]...
www - http://www.lptl.jussieu.fr/user/mossa/Welcome.html
Tel - +33 438 783577
FaX - +33 438 785691
Try diffing the older and newer pair style
files. If they are essentially the same, then
that's not the issue.
Steve
Try diffing the older and newer pair style
files. If they are essentially the same, then
that's not the issue.
there also was a change by aidan that affects the
energy calculation as "git blame" shows
3ad56870 (sjplimp 2008-05-15 20:49:38 +0000 160) if (rsq < cut_ljsq) {
b511a755 (sjplimp 2010-11-05 15:57:43 +0000 161) evdwl =
r6inv * (lj3[itype][jtype]*r6inv - lj4[itype][jtype]);
7d858516 (athomps 2011-08-26 00:17:54 +0000 162) evdwl +=
ljsw5[itype][jtype];
3ad56870 (sjplimp 2008-05-15 20:49:38 +0000 163) if (rsq >
cut_lj_innersq) {
b511a755 (sjplimp 2010-11-05 15:57:43 +0000 164) eswitch
= tlj*tlj*tlj *
7d858516 (athomps 2011-08-26 00:17:54 +0000 165)
(ljsw3[itype][jtype] + ljsw4[itype][jtype]*tlj);
3ad56870 (sjplimp 2008-05-15 20:49:38 +0000 166) evdwl += eswitch;
3ad56870 (sjplimp 2008-05-15 20:49:38 +0000 167) }
3ad56870 (sjplimp 2008-05-15 20:49:38 +0000 168) evdwl *= factor_lj;
3ad56870 (sjplimp 2008-05-15 20:49:38 +0000 169) } else evdwl = 0.0;
commit 7d8585160725d2badfb1ba40e81a88050ba9ff83
Author: athomps <[email protected]...>
Fixed error in lj/gromacs
git-svn-id: svn://svn.icms.temple.edu/lammps-ro/[email protected]...
f3b2605a-c512-4ea7-a41b-209d697bcdaa
axel.