[gmx-users] Box Size in MD

Justin A. Lemkul jalemkul at vt.edu
Thu Mar 22 17:28:23 CET 2012

Lara Bunte wrote:
> Hello
> What size should a box have in that you do your MD? I always read that for
> short Lennard Jones interactions one should do a cut off with the half of the
> box size but what how to know a good box size?

This is not correct.  The cutoffs are dictated by the force field you've chosen 
to use.  There are predefined values that should (for the most part) be adhered 
to.  What you need to make sure of when setting up the box is that the longest 
cutoff cannot exceed half the shortest box vector.  If it does, you get 
duplicate force evaluations from a violation of the minimum image convention. 
Use of an NPT ensemble will cause the box dimensions to fluctuate (due to 
pressure) and thus it is unwise in such cases to ever have a box that is only 
twice the size of the longest cutoff, as violations are almost certain to occur.



Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list