The ABINIT Users Mailing List ( CLOSED )

[Ponniah Ravindran <ponniah.ravindran@kjemi.uio.no>]

Dear Storr,

        If you know the number of valence electrons involved
in the calculation, you can calculate number of electrons in
the occupied part of the band structure. We know that each
band is occupied by 2 electrons. Hence the rule of thump is
Nband = (valence el.)/2

For metals some of the bands cross the Fermi level, and hence

nband = Nband + 20% of Nband.

I hope you may now able to sentup nband for your case.

Best regards
Ravi

On Wed, 26 Mar 2003, Riad Shaltaf wrote:

> Hi..
> I dont know for which compund you are calculationg the energy bands,
> but most probably it is a metallic one. in metalic cases you should not
> rely on default values of nband.
> nband value  must be increased in such a way you include 20%-30% of the
> original number of bands as extra ones , with this the energy may
> converge.
>
> I hope this might help.
>
> On Wed, 26 Mar 2003, M.T.Storr wrote:
>
> > Hi,
> >
> > I am just starting out in the world of DFT and am therefore trying to set
> > up calculations on some compounds of interest. However when I try to run
> > an energy calcualation on a fixed geometry the run keeps crashing and
> > states increase the nband value.
> > My question is how doe sone determine the value of nbands to use? I am
> > sure that it will be someting straightforward, but I have not found out
> > how at present.
> >
> > Many thanks and aplogies if this is too simple a question
>

*****************************************************************
Dr.P.Ravindran, Associate Professor    |
Department of Chemistry                |Tel work: +47 22 85 56 06
University of Oslo, PO Box 1033        |     mob: +47 45 27 07 37
Blindern, N-0315 OSLO, NORWAY          |Fax: +47 22 85 54 41/55 65
Email: ponniah.ravindran@kjemi.uio.no  |
URL : http://folk.uio.no/~ravi/        |
*****************************************************************