The ABINIT Users Mailing List ( CLOSED )

[Xavier Gonze <gonze@pcpm.ucl.ac.be>]

Dear Xinjie Wang,

Xinjie Wang wrote:
> Dear, all
>
> I am doing non-collinear magnetism calculation of BCC Fe with spin-orbit 
> coupling (nspden=4, spinor=2).
> Abinit requires turnning off all symmetries. The computational time 
> increases by 48 times!  I am wondering
> if Abinit can identify the magnetic space group and if there is a way to 
> retain the symmetries in non-collinear
> magnetism calculation. Many thanks in advance!

Non-collinear magnetism (nspden=4, nspinor=2) works, but this does not
mean spin-orbit is activated. In order to use spin-orbit, you must
use non-default values for the input variable so_typat (so_typat(:)= 2, for 
HGH
pseudopotentials). But at present, you cannot use non-collinear magnetism 
(nspden=4)
with spin-orbit in ABINIT. Sorry for this. By contrast, you can do
spin-orbit calculation (nspinor=2) with time-reversal symmetry, and no 
magnetism (nspden=1).

Moreover, I just checked that no test prevent the user to set nspden=4 with 
so_typat(:)/=1 .
This will be corrected in a future version of ABINIT ... or non-collinear 
magnetism
will be implemented ...

Xavier