[Pw_forum] constrained magnetization calculation to converge at the ferromagnetic state

[sreekar guddeti]

@Arles

> Dear Sreekar Guddeti,


Respected Mr. Rebaza,
  Thank you for the reply.

if you want to calculate a constrained magnetization
> with antiparallel direction between the atom 2 and atom 8, you must to put
> the "starting_magnetization" tag with opposite value, for example:
>
>  starting_magnetization(2)=  0.3091,
>  starting_magnetization(8)= -0.3091,
>
>
 I wanted to perform a constrained calculation for the parallel spin
configuration.
 --------------------------------------------------------
@Gabriele

> Dear Sreekar,
>
> Respected Mr. Sclauzero,
  Thank you for  the suggestions and the pointers.

I would first try to reduce mixing_beta  and also lambda (say, to 0.2 or
> 0.1). You can then restart the calculation with a larger value of lambda.
>

   Don't forget that for USPP you need to specify also ecutrho, and perhaps
you don't need such a high ecutwfc.

> If your USPPs are those on the QE website, I would use ecutwfc=35 and
> ecutrho=350, for instance.
>
>
I am using USPPs sourced with the package. It seems like I was using an
incorrect algorithm to converge scf calculation. I converged with ecut
keeping the default ecutrho at gamma point, then with ecutrho, then with
kpoints and finally with smearing width.
Questions:
1.  Would I be right to assume there is no need to converge the calculation
with the mixing_beta?
2.  How significant it is to start with different sets of
initial_magnetizations for an unconstrained calculation considering its
arbitrariness and the increased cost of performing them. Quantitatively, how
would one determine the density of the sampling from the possible
configurations in the light of the accuracy of the calculation involved.

Sincerely,
Sreekar Guddeti
Undergraduate (5th year)
Engineering Physics
IIT Bombay
India

> Respected Sir/Madam,
> > 1. I would like to perform a constrained magnetization calculation for a
> system, whose unconstrained spin polarized calculation gives an
> antiferromagnetic state, to converge at the ferromagnetic state.
> > 2. The system is a zigzag graphene nanoribbon(
> http://en.wikipedia.org/wiki/Graphene) and I would like to flip the atomic
> moment at one of the edge site and constrain both the edge site moments.
> > 3. I calculated the antiferromagnetic edge site moment  using PDOS
> calculation. I used 'lambda' parameter to constrain the moments.
> >
> > 4.Questions: Is 3) the right way to do it? If yes, what should be value
> of lambda? I used lambda =1 and the scf output (given below) is
> disturbing... I would like to ask what does "constrained moment :
> -20.000000"
>
> Indeed this looks suspicious... one should look into the code to understand
> what's going on. You can have a look into add_bfield.f90, it should not be
> very difficult to understand how it works.
>
>
Good luck
>
>
> GS
>
>
> > mean and why is the constrain energy so high.  If it is not the right
> way, I would be grateful if you could suggest an alternative.
> >
> > Thanking you,
> > Yours sincerely,
> > Sreekar Guddeti
> > Undergraduate (5th year)
> > Engineering Physics
> > IIT Bombay
> > India
> >
> > PS: The system namelist for scf calculation and the sample output are
> given below and the complete input file is attached.
> >
> >
> > --------------------------------------
> > ...
> >
>  ==============================================================================
> >      atom number   15 relative position :   -3.5000   6.0622   0.0000
> >      charge :     0.983302
> >      magnetization :          0.000256
> >      magnetization/charge:    0.000260
> >      constrained moment :     0.309100
> >
> >
>  ==============================================================================
> >
> >
>  ==============================================================================
> >      atom number   16 relative position :   -3.0000   5.7735   0.0000
> >      charge :     1.783674
> >      magnetization :         -0.385046
> >      magnetization/charge:   -0.215873
> >      constrained moment :   -20.000000
> >
> >
>  ==============================================================================
> > ...
> > ...
> >
> > constraint energy (Ryd) =   6095.83663339
> > ---------------------------------
> >
> >
> >
> > --------------------------------
> > &SYSTEM
> >                        ibrav = 12,
> >                    celldm(1) = 4.608737,
> >                    celldm(2) = 12.0,
> >                    celldm(3) = 10.0,
> >                    celldm(4) = -0.50,
> >                          nat = 18,
> >                         ntyp =10,
> >     starting_magnetization(2)= 0.3091,
> >     starting_magnetization(8)= 0.3091,
> >                   occupations='smearing',
> >                      smearing='gauss',
> >                       degauss=0.01,
> >                         nspin=2,
> >     constrained_magnetization='atomic'
> >                       lambda = 1
> >                      ecutwfc = 50.D0 ,
> >                       report = 2
> >
> >  /
> >
> >
> > --
> > Sreekar Guddeti
> >
> > <grapheneznrn10j.scf.in>_______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
>
>
> ? Gabriele Sclauzero, EPFL SB ITP CSEA
>   PH H2 462, Station 3, CH-1015 Lausanne
>
>
>
>
>
>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> http://www.democritos.it/pipermail/pw_forum/attachments/20110825/ba3beab2/attachment.htm
>
> ------------------------------
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
> End of Pw_forum Digest, Vol 50, Issue 59
> ****************************************
>



-- 
Sreekar Guddeti
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20110826/adea30f8/attachment.html>