[Wien] Convergence Problem

[Laurence Marks]

Some points:

1) Except in very special cases (isolated transition metal atoms),
there is no reason to use the PRATT scheme, it is obsolete.

2) Depending upon which version you are using, in fact the "BROYD"
scheme is really the multisecant approach. (Do grep -e "MIXING SCHEME"
*scf). If you are in fact using BROYD, don't -- it has been obsoleted
by MSEC1.

3) My personal opinion is that the default mixing of 0.4 is too
greedy; I use 0.1. For a few problems this may require an extra
iteration or two, in general not.

4) Don't play around with the mixing parameter in MSEC1. Unlike the
PRATT or BROYD methods you don't need to. See PRB 78, 075114 
2008
for more details.

5) Convergence of the scf cycle depends upon how well you have posed
the physical problem. While with RMT's of 1.25 an RKMAX of 5.5 is OK
for a general calculation, I suspect that 1 k-point is too few. I
suspect that to get a convergence to an -fc of 0.1 you will also have
to push RKMAX higher and use more k-points.

On Mon, Jan 19, 2009 at 5:43 AM, ritabrata bhowmick <rito at rcais.res.in> wrote:
> Dear Wien users,
>                             I am performing phonon calculations of B13C2  with 2x2x2 supercell. After init_phonon_lapw  with RMT 1.25 (maximum possible), RMTkmax 5.50 , k points 1 and mixing factor 0.05(BROYD scheme). It converged with -fc 1.0 but  I am facing convergence problem with -fc 0.1.  Default mixing factor is 0.4. I tested the same with mixing factor 0.1 with BROYD and PRATT scheme separately but the result was same. I am waiting for your suggestions.
>
>                                                                                                  Ritabrata
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>



-- 
Laurence Marks
Department of Materials Science and Engineering
MSE Rm 2036 Cook Hall
2220 N Campus Drive
Northwestern University
Evanston, IL 60208, USA
Tel: (847) 491-3996 Fax: (847) 491-7820
email: L-marks at northwestern dot edu
Web: www.numis.northwestern.edu
Chair, Commission on Electron Crystallography of IUCR
www.numis.northwestern.edu/
Electron crystallography is the branch of science that uses electron
scattering to study the structure of matter.