[Wien] Convergence Problem

[Laurence Marks]

In addition to what Peter said:

* Please read carefully
 http://www.numis.northwestern.edu/Presentations/DFT_Mixing_For_Dummies.pdf

* Why are you trying to converge ee to 0.00001? This may be too small to be
useful.

* The most common reason for convergence problems is a bad problem, for
instance RMTs too small, confused au for Angstroms.

* What version are you using? With a recent one did you uncomment "STIFF"?

* Check-mixing is much more useful than what you provided.

--
Professor Laurence Marks (Laurie)
Department of Materials Science and Engineering, Northwestern University
www.numis.northwestern.edu
"Research is to see what everybody else has seen, and to think what nobody
else has thought" Albert Szent-Györgyi

On Tue, Apr 25, 2023, 04:23 Peter Blaha <peter.blaha at tuwien.ac.at> wrote:

> Please check the relevant quantity. For a mbj calculation this is usually
> the gap.
>
> grep :GAP case.scf
>
> Is it converged and stable ?
>
> Also monitor   :DIS how it changes (smooth or with jumps or with long-term
> slow oszillations?), which may give further hints.
>
> mBJ calculations sometimes converge quite difficult, because the kinetic
> energy density (tausum) is mixed separately using PRATT and the density
> mixing (using MSR1) may be confused because it does not know about the
> tausum. In addition there is a "3rd mixing" in lapw0 (potential mixing,
> which depends on the "greed" of the last cycle - which cannot be changed by
> an input)
>
> Sometimes it helps to:
>
> i)  change the mixing factor in case.inm_tau  ( even larger when
> convergence is slow but smooth, smaller when oszillating !)
>
> ii) change for a couple of iterations (maybe 20) to PRATT in case.inm,
> then continue again with MSR1
>
>
> Am 25.04.2023 um 11:01 schrieb Hülya Gürçay:
>
>  Dear Prof. Dr. Blaha and WIEN2k users,
>
> I've been using Wien2k since 4 months. I am a physics master's student,
> calculating the magnetic and thermoelectric properties of quaternary
> Heusler compounds. I have a question,
>
> Energy and charge converged with PBEsol but not with PBEsol-mbj potential .
>
> Xc potential : PBEsol - mbj
> cc: 0.0001
> ec: 0.00001
> kmesh:20,20,20
>
> Below I am sharing the results got after 100, 150 and 200 iterations.
>
>     cycle 100 <>(Sat Apr 22 21:59:29 +03 2023) >(1/99 to go)
> :ENERGY convergence:  0 0.00001 .0001018050000000
> :CHARGE convergence:  0 0.0001 .0079100
> >   energy in SCF NOT CONVERGED
> >   stop
>
>     cycle 150 <>(Sat Apr 22 23:24:49 +03 2023) >(1/49 to go)
> :ENERGY convergence:  1 0.00001 .0000068950000000
> :CHARGE convergence:  0 0.0001 .0032008
> >   charge in SCF NOT CONVERGED
> >   stop
>
>     cyle 200 <>(Sun Apr 23 00:47:57 +03 2023) >(1/98 to go)
> :ENERGY convergence:  1 0.00001 .0000027900000000
> :CHARGE convergence:  0 0.0001 .0021041
> >   charge in SCF NOT CONVERGED
> >   stop
>
> What should I do to get convergence? Please give your valuable suggestions
>
> BestRegards,
> Hülya
>
> _______________________________________________
> Wien mailing listWien at zeus.theochem.tuwien.ac.athttp://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/20230425/a32f5364/attachment.htm>