The ABINIT Users Mailing List ( CLOSED )

["Anglade Pierre-Matthieu" <anglade@gmail.com>]

Dear Konstantin

Have you tryed to optimize diemac and dielng ? On bulk materials
Kerker's preconditioner should behave almost perfectly and almost for
free...
Otherwise do you use an electronic temperature as zhangting does ? No
electronic temperature can also prevent convergency on systems with
for instance dangling bonds like a defect in an otherwise bulk Si
cell.

regards

PMA

On 10/23/06, Konstantin Rushchanskii <konstantin.rushchanskii@cea.fr> wrote:
> Dear Matthieu,
>
> I have had the same problem with bulk materials.
>
> Regards,
> Konstantin
> >
> > GGA can be more sensitive to vaccuum (certain terms can diverge when n
> > ->0 though this should be treated properly in abinit kudos for
> > Xavier), need higher ecut or other convergence parameters. Try other
> > iscf values, nline variables, or just be happy with 10-10 which is not
> > bad for tolvrs.
> >
> > Matthieu
> >
> >


--
Pierre-Matthieu Anglade