Project

General

Profile

Bug #592

Long-range LJ correction for test particle insertion

Added by Kevin Daly about 9 years ago. Updated almost 9 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Erik Lindahl
Category:
mdrun
Target version:
Affected version - extra info:
Affected version:
Difficulty:
uncategorized
Close

Description

When inserting a molecule with multiple atoms, Gromacs does not appear to compute <C6> properly for the long-range correction. Instead, it computes <C6> for only one of the atoms in the inserted molecule. For instance, if I insert a TIP4P/2005 water molecule, which has one site with nonzero LJ parameters and three others with zero-valued LJ parameters, the <C6> value is nonetheless calculated to be zero. In the tpi.xvg output file, U_{dispcorr} is also reported to be zero.

input_files.tar.gz (277 KB) input_files.tar.gz Input files Kevin Daly, 10/10/2010 07:59 PM
forcerec.c (56.8 KB) forcerec.c fix Berk Hess, 10/12/2010 09:39 AM

History

#1 Updated by Kevin Daly about 9 years ago

Created an attachment (id=552)
Input files

#2 Updated by Berk Hess about 9 years ago

I think I fixed it with commit c99629bef0e89518ddef7b740e0081bd027fd29a
Could you test the attached fix?

Thanks,

Berk

#3 Updated by Berk Hess about 9 years ago

Created an attachment (id=553)
fix

#4 Updated by Berk Hess almost 9 years ago

Could you please confirm that I fixed this issue?
Then we can close the bug.

Berk

#5 Updated by Kevin Daly almost 9 years ago

(In reply to comment #4)

Could you please confirm that I fixed this issue?
Then we can close the bug.

Berk

Thank you for fixing the bug. I have confirmed that it works by comparing it to a "brute-force" calculation involving the difference in the LJ correction between a configuration with N+1 water molecules and one with N water molecules.

-Kevin

Also available in: Atom PDF