[Pw_forum] Re: Pw_forum digest, Vol 1 #1029 - 5 msgs

Nicola Marzari marzari at MIT.EDU
Fri Jun 16 06:23:00 CEST 2006

aarti srirangarajan wrote:
> Thanks for the reply.
> The degauss value I am working with is 0.003. What value should typically
> be used for Ni?
> regards
> aarti

Such small value will require a large amount of kpoints to
converge forces and energies. My rule of thumb is between 0.1 eV
and 0.5 eV (REMEMBER TO divide by 13.5068 to obtain the correct degauss
for pwscf).

0.1eV is really small, to be used for difficult/magnetic cases. 0.5eV
is large, for simple metals like Al.

The larger the smearing, the fewer kpoints needed for an accurate
integration (functions are smoother). But, the larger the smearing,
the further away you move from the zero-temperature target.

So it's a tradeoff, and you need to find the sweet spot.

You should use either methfessel-paxton or marzari-vanderbilt, as a
smearing scheme. I prefer the latter.


Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu

More information about the users mailing list