VASP Forum

Hi,

I seem to have a consistent discrepancy when calculating energy using the VASP-implemented (IVDW=202) and LIBMBD MBD dispersion correction.

I am using VASP version 6.6.1 compiled with LIBMBD, and in this example compared two single point calculations, one using PBE and IVDW=202, and one using PBE and IVDW=14. I tried to manually supply the KPOINTS mash using the LIBMBD flags in the INCAR when using the LIBMBD version of the correction, as was suggested in another post in this forum, like so:

IVDW = 14
LIBMBD_XC = pbe
LIBMBD_K_GRID = 2 1 4
LIBMBD_K_GRID_SHIFT = 0 0 0

The resulting total energies are:

VASP-implemented PBE+MBD = -916.7127424 eV
LIBMBD PBE+MBD = -916.7171248 eV

This is about 0.005 eV difference.

I have attached the input and output files associated with both calculations.
Please let me know if I am doing anything wrong or missing anything, or weather this is a real issue. As these are both PBE+MBD@rsscs I expect the final total energies to be the same.

Thank you very much for your help.

Dear Roy,

two reasons for the discrepancies are given in this forum thread.
Does Fabien's post answer your questions?