[Sax-project-users] kgrid

[Giovanni Bussi]

Dear Ali,

in a GW calculation there a many parameters that needs to be optimized.

- kpoint grid: as a starting point, the same used for DFT calculation
should be fine. but be aware that you may need to increase it
(especially if you go to BSE calculations)

- cutoff_* parameters, the crucial ones are:
cutoff_polarizability: default 6 Ry. this is very system dependent,
try to increase it (if you can afford it!).
cutoff_fock: default 6 Ry. use something of the order of the
wavefunctions cutoff

Concerning the spectra, after a GW calculation you can only use the
calculation_kind=RPA option, which neglects excitonic effects. If you
need to include them, you also have to perform a bse calculation,
followed by a spectra calculation with calculation_kind=RPAEXC

Giovanni



On Wed, Jul 22, 2009 at 1:27 PM, <kazempoor at ph.iut.ac.ir> wrote:
>
> Dear developer
> For doing fine GW calculation, in addition to using optimized values for
> ecut in initial step, which parameters would be also optimized? How many
> kpoints is enough?
> another question:
> by runing spectra.x it is possible to obtain dielectric constant. If after
> GW.x, we do spectra.x , how much are differences with when doing spectra.x
> after bse.x?
>  Thanks a lor
> Ali Kazempour, Fritz-Haber Institute,Berlin ,Germany
>
> _______________________________________________
> Sax-project-users mailing list
> Sax-project-users at qe-forge.org
> http://qe-forge.org/mailman/listinfo/sax-project-users
>