[Pw_forum] Problem With WC Pseudopotentials in "Relax" Process

[Nicola Marzari]

Dear Mojtaba,


1) for ferroelectric perovskites, use PBEsol, preferably, or LDA as an 
xc functional

2) for the particular case of the Ba, Ti and O pseudos, use either
those three elements from the PBESOL pslibrary 1.0 (use a cutoff of 45 Ry,
dual of 8), from
http://www.qe-forge.org/gf/project/pslibrary/frs/?action=FrsReleaseView&release_id=147
or the ultrasoft PBESOL GBRV library (use a cutoff of 40 Ry, dual of 8),
http://www.physics.rutgers.edu/gbrv/

3) make sure you use atomic+random as initialization for the wavefunctions

Not sure if your problems are related, but fix 1+2+3 above before anything.

				nicola



On 14/03/2015 17:35, Mojtaba Mirseraji wrote:
> Dear All
>
> I am using WC pseudopotentials (THEOS, PSLibrary) for relaxing BaTiO3.
> These potentials are so good for cubic structure but when I use it for
> tetragonal structure of barium titanate with some errors !!!!!
> such as "not converged eigenvalues" , "S matrix not positive definite"
> , "fail to eigenvectors" and .....!!!!!!!
>
> I only changed force_conv_thr and etot_cov_thr=1.D-2 are without error !!!!
>
> I do not know how to do !!!
> PLEASE, help me
>
> Thanks in advance
>
> Mojtaba Mirseraji
> Ph.D. Candidate
> Theoretical Condensed Matter Physics
> Arak Univ.
> I.R.IRAN
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>

-- 

----------------------------------------------------------------------
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL