Clarification of R_CUTOFF in VDW_POTENTIAL
emerson p l
Dear cp2k community, I hope you’re all doing well.
I’m writing to ask for clarification on the choice of the cutoff radius parameter when using periodic boundary conditions in VDW_POTENTIAL in AIMD and geometry/cell optimizations:
&VDW_POTENTIAL
POTENTIAL_TYPE PAIR_POTENTIAL
&PAIR_POTENTIAL
PARAMETER_FILE_NAME dftd3.dat
TYPE DFTD3
REFERENCE_FUNCTIONAL PBE
R_CUTOFF [angstrom] 10
&END
&END VDW_POTENTIAL
Suppose I have a cubic simulation box with side length angstrom, and I (mistakenly) set R_CUTOFF = 10 angstrom. Would setting RCUTOFF = 10 angstrom in a 5 angstrom cubic box cause the cutoff sphere to extend beyond the primary cell and overlap with periodic images, resulting in unphysical, multiply-counted interactions?
My understanding from classical MD is that we must obey the minimum-image convention, which requires R_CUTOFF ≤ 2 Lmin. I don't know if R_CUTOFF has this physical meaning of classical MD.
Thank you very much for your time and insights.
Best,
Emerson
Jürg Hutter
CP2K does not use MIC for vdW corrections. You can specify any cutoff radius.
Atoms within 2*R_CUTOFF will interact. You can check the convergence of the vdW
energy by increasing the cutoff .
regards
JH
________________________________________
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of emerson p l <emers...@gmail.com>
Sent: Thursday, June 26, 2025 9:41 PM
To: cp2k
Subject: [CP2K:21590] Clarification of R_CUTOFF in VDW_POTENTIAL
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/0afee8cd-252c-4c9e-b0c1-e2142551930bn%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/0afee8cd-252c-4c9e-b0c1-e2142551930bn%40googlegroups.com?utm_medium=email&utm_source=footer>.