Question on DFT-D3 vdw corrections

[Alex Brown]

Dear Cp2k experts,

I have a question around DFT-D3 VDW correction in cp2k. How do we set the parameter R_CUTOFF?  

I started with 7.5 angstroms (half of my cell size), but I observed energies change by 5 Kj/mol when I increase the cutoff to 10 angstroms. Can someone explain me please how to choose this parameter properly to get reliable results?

Please find attached my input and topology file.

Thanks,

Alex

[Jürg Hutter]

Hi

interactions are included if two atoms are within 2*R_CUTOFF. The default value for
R_CUTOFF is 10 Angstrom. You can use larger values (there is no restriction of a
minimum image convention), but you should probably not reduce it.

regards

Juerg Hutter

________________________________________
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of Alex Brown <alexbro...@gmail.com>
Sent: Wednesday, April 13, 2022 3:38 AM
To: cp2k
Subject: [CP2K:16846] Question on DFT-D3 vdw corrections

Dear Cp2k experts,

I have a question around DFT-D3 VDW correction in cp2k. How do we set the parameter R_CUTOFF<https://manual.cp2k.org/trunk/CP2K_INPUT/ATOM/METHOD/XC/VDW_POTENTIAL/PAIR_POTENTIAL.html#R_CUTOFF>?

I started with 7.5 angstroms (half of my cell size), but I observed energies change by 5 Kj/mol when I increase the cutoff to 10 angstroms. Can someone explain me please how to choose this parameter properly to get reliable results?

Please find attached my input and topology file.

Thanks,
Alex

--
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 on the web visit https://groups.google.com/d/msgid/cp2k/58a36afe-6cd5-4aad-b35e-0fc828b93407n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/58a36afe-6cd5-4aad-b35e-0fc828b93407n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Alex Brown]

Dear Prof. Hutter,

Thanks a lot for your response.

Regards,

Alex