[Pw_forum] pbe with norm conserving pp convergence problem

[Jia Chen]

O norm-conserving usually need 70Ry, since you have d electrons in
your system, they may require higher cutoff. Some people calculate
lattice constants to justify their choice of cutoff, but It finally
depends on what do you study. Not much more I can say about it...

On Mon, Sep 10, 2012 at 9:48 PM, Peng Chen <pchen at ion.chem.utk.edu> wrote:
> Dear Jia,
>
> I checked those pp and there are not recommended cutoffs. When I use
> occupation=fixed, the results are similar as those from occupation=smearing
> (reach minimum at 80 Ry) . And the scf doesn't converge when ecutwfc=60 Ry.
>
> Ecutwfc   energy     total force
> 50 -1129.7209 0.7937
> 60 no  convergence
> 70 -1136.6081 0.7066
> 80 -1137.3776 0.7059
> 90 -1137.3342 0.6997
> 100 -1136.9523 0.7056
>
> On Mon, Sep 10, 2012 at 5:17 PM, Jia Chen <jiachenchem at gmail.com> wrote:
>>
>> Dear Peng,
>>
>> For AFM case, you can set tot_magentization = 0, for FM case, you need
>> to know tot_magnetization from experiments or previous calculations.
>> Or you can use the int(absolute magnetization) from you AFM
>> calculations.
>>
>> Trying to converge total energy with respect to wave function cutoff
>> is usually not a good idea. If you just want to choose a good cutoff
>> energy for calculations, first thing you can do is to look into those
>> pseudopotentials you are using. If you can find recommended cutoffs,
>> you probably can just use the highest cutoffs in your system.
>>
>> Bests
>> Jia
>>
>> On Mon, Sep 10, 2012 at 5:02 PM, Peng Chen <pchen at ion.chem.utk.edu> wrote:
>> > Dear Dr.  Giannozzi,
>> >
>> > Thanks for your reply. I use smearing for computing the band structure
>> > in
>> > afm and fm states. It looks there is still a lot to learn. If I use
>> > occupation=fixed to calculate afm/fm states, shall I just set
>> > tot_magnetization=0 or 1?
>> >
>> > If I didn't set ecutrho, the default value of ecutrho is 4*ecutwfc. So
>> > the
>> > results are the same as the following, and the energy doesn't decrease
>> > monotonically. I tried to decrease degauss to 0.005 Ry, I got the
>> > similar
>> > results. What I can do to make the energy converge?
>> >
>> > ecutrho=4*ecutwfc
>> > ecutwfc     energy            total force
>> > 30 -1086.672528 1.335253
>> > 40 -1119.972723 0.690327
>> > 50 -1129.973333 0.690484
>> > 60 -1134.622157 0.640747
>> > 70 -1136.892534 0.617201
>> > 80 -1137.666471 0.616635
>> > 90 -1137.620919 0.610216
>> > 100 -1137.238099 0.615361
>> > 110 -1136.89911         0.62063
>> > 120 -1136.866006 0.622763
>> >
>> > On Mon, Sep 10, 2012 at 3:45 PM, Paolo Giannozzi
>> > <giannozz at democritos.it>
>> > wrote:
>> >>
>> >>
>> >> On Sep 10, 2012, at 16:35 , Peng Chen wrote:
>> >>
>> >> > I tried to calculate an insulator Ni3V2O3 using pbe norm conserving
>> >> > pp from pslibrary. But the energy related with ecutwfc is not
>> >> > converged. It reaches minimum at ecutwfc=80 Ry. But when I fix
>> >> > ecutrho=320 Ry and increase  ecutwfc, it starts to decrease
>> >> > monotonically. I am not sure if the system reaches the convergence
>> >> > at ecutwfc=80 Ry.
>> >> > (smearing=mv, Degauss=0.01 Ry and kpoints = 4 4 3)
>> >>
>> >> if it is an insulator, why are you using smearing? Apart from this:
>> >> with norm-conserving PP, ecutwfc is the only parameter defining
>> >> the basis set. You should not set ecutrho.
>> >>
>> >> P.
>> >> ---
>> >> Paolo Giannozzi, Dept of Chemistry&Physics&Environment,
>> >> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
>> >> Phone +39-0432-558216, fax +39-0432-558222
>> >>
>> >>
>> >>
>> >>
>> >> _______________________________________________
>> >> Pw_forum mailing list
>> >> Pw_forum at pwscf.org
>> >> http://www.democritos.it/mailman/listinfo/pw_forum
>> >
>> >
>> >
>> >
>> > --
>> >   Best Regards.
>> >         Peng
>> >
>> > _______________________________________________
>> > Pw_forum mailing list
>> > Pw_forum at pwscf.org
>> > http://www.democritos.it/mailman/listinfo/pw_forum
>> >
>>
>>
>>
>> --
>> Jia Chen
>> webpage: www.princeton.edu/~jiachen
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
>
> --
>   Best Regards.
>         Peng
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>



-- 
Jia Chen
webpage: www.princeton.edu/~jiachen