[Wien] setrmt and init_lapw

[Mikhail Nestoklon]

Dear Prof. Blaha,
Thank you for the clarification.

My naive impression was that the problem was in the small RMT (and maybe
HDLO), but it was not.
After some more experiments with this case I realized that the main problem
was my mistake which led to incomplete initialization of the structure
symmetry. Proper initialization of the structure which fully takes into
account the structure symmetry before running init_lapw removes the problem.
The tricky part in finding this mistake was that init_lapw did put some
symmetry (which is lower than the correct one) to the structure
automatically.
It was unexpected to have QTL-B due to slightly lower symmetry as well, but
now I am aware that this can happen.

Sincerely yours,
Mikhail Nestoklon


On Fri, 21 Mar 2025 at 20:57, Peter Blaha <peter.blaha at tuwien.ac.at> wrote:

> The larger the RMTs are, the larger is the "linearization error".
>
> Thus we have set certain limits for RMTs, depending on the required
> accuracy.
> In setrmt this is 2.5 bohr.
> In init_lapw with prec 2 and 3 this is further reduced to minimize any
> linearization error  (it does not make sense to use eg. a fully
> converged RKmax, but use large spheres (or a lousy k-mesh or a low
> L-VNS). The -prec XX  selection tries to balance these settings.
>
> PS: The next   init_lapw will have a -noreduc option, since the HDLOs
> are quite efficient to reduce linearization errors also for larger spheres.
> At the moment you would have to run:
> init_lapw -prec 2
> edit the struct file and increase rmt again.
> x lstart
> x dstart
> run_lapw ....
>
> PPS: For which atom happens the ghostband problem ??
>
> Am 21.03.2025 um 08:45 schrieb Mikhail Nestoklon:
> > Dear wien2k community,
> > I am a bit confused about RMT settings when using init_lapw with "-prec"
> > switch.
> > It seems, for "-prec 2n" or "-prec 3n" the script sets the RMT to its
> > minimal value. As a result, in one of my cases, simple "-prec 2n/3n"
> > results in ghostbands (problems caused by HDLO on one of atoms which is
> > automatically added due to small RMT).
> > The setrmt script, as well as "-prec 1n", on the contrary, sets the
> > spheres close to maximum possible values.
> >
> > What is the current strategy (and reasoning behind it) of the "-prec"
> > option on setting RMTs?
> >
> > Thank you in advance.
> >
> > Sincerely yours,
> > Mikhail Nestoklon
> >
> >
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> > SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>
> --
> -----------------------------------------------------------------------
> Peter Blaha,  Inst. f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-158801165300
> Email: peter.blaha at tuwien.ac.at
> WWW:   http://www.imc.tuwien.ac.at      WIEN2k: http://www.wien2k.at
> -------------------------------------------------------------------------
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20250323/5320caab/attachment.htm>