You should add a smearing, even if you think that your system is not metallic (if you think it is, you MUST add a smearing). Something like occupations='smearing', smearing='gaussian', degauss=0.01 to 0.05 or so. You might also try a spin-polarized calculation.
Paolo On Wed, Dec 27, 2017 at 3:51 AM, lt15 <[email protected]> wrote: > > Hi > > I was doing a 'scf' calculation for a defected h-BN structure as shown in > the attached figures. I would like to obtain the charge density of the > layer especially around the defect. But I found the calculation couldn't > converge as the 'estimated scf accuracy' in the output file keeps > fluctuating. In the input file, I tested different mixing_mode like 'plain' > and 'local-TF'. or the mixing_beta value like 0.01. 0.1, 0.3, 0.7. But > still, it didn't converge even after 8 hours running. In structure in-a, > the defected boron atom was far away from its original lattice position, as > shown in in-a.png. In structure in-b, the defected boron atom plus a > nitrogen atom around the edge of layer were deleted, as shown in in-b.png. > In structure in-b, the layer still maintained neutral electricity and kept > the void defect in the center. However, neither structure can converge. > Could you give some suggestions about the convergence issue for this > defected structure? Thank you. > > best, > Lei > > _______________________________________________ > Pw_forum mailing list > [email protected] > http://pwscf.org/mailman/listinfo/pw_forum > -- Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche, Univ. Udine, via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222
_______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
