The ABINIT Users Mailing List ( CLOSED )

[Gilles Zerah <gilles.zerah@cea.fr>]

According to what I understand, you wish to find the gs  wave function with a 
given value of the net
total
magnetisation. ($\vec M$)
Naturally, this would have a non trivial effect only by including spin-orbit 
coupling, which is
perfectly possible in the present version of the code.
To do this by brute force would be to impose a constraint of given magnetic 
moment, which is not
too dissimilar to a (generalized) orthonormality constraint (the spin density 
matrix should be given
a value.)
So, this might be feasible, but first let me know if this is what you want.
Gilles

(I had to resend this mail, because it was beginning by PHrom.....)
> Hello everyone,
>
> I'm trying to calculate Berry curvature and Chern number changing net 
> magnetization.
> In this forum, there is explanation about Berry phase approach to 
> polarization
> including derivatives of wavefunctions with respect to k or other variables.
>
> However, what I want to change is the direction of the net magnetic moment 
> so
> that by calculating Berry curvature along the arbitary direction of the 
> magnetic
> moment,which in general need not be the lowest energy direction, and 
> averaging it,
> I want to calculate the Chern number.
>
> Any comment or help would be greatly appreciated.
>
> Hongki Min