I don't understand Non-Collinear results

[lahaye]

Hi,

I have a number of different structures that possibly have magnetic
properties, because in the spin-polarized calculations there are
1 or more unpaired electrons (the 'mag' value in OUTCAR is
typically between 1 and 2).

In order to learn more about the corresponding magnetic moment,
I was adviced by 'admin' to use Spin-Orbit coupling. See:
http://cms.mpi.univie.ac.at/vasp-forum/ ... php?4.2064

So I did a non-collinear calculation, according the vasp guide,
using the recommended approach at the end of:
http://cms.mpi.univie.ac.at/vasp/vasp/node144.html
which prescribes to use in INCAR:

ICHARG = 11
LSORBIT = .TRUE.
SAXIS = x y z ! direction of the magnetic field
GGA_COMPAT = .FALSE. ! improves precission of GGA
ISYM = 0


My results for one particular structure are as follows:

1) The 'mag' in OSZICAR is always 0 0 1, no matter what I
choose for the (x,y,z) direction or size of the magnetic field in INCAR.

2) In OUTCAR, there are only non-zero values for mag(z);
for mag(x) and mag(y) everything is zero.

How should I interprete these numbers?
Is the direction of the magnetic moment for this structure
always along the z-axis?

Thank you for your help.

Regards,
Rob.

[admin]

1) you got my answer wrong: non-collinearity has NOTHING to do with SOC.
2) your results indicate that the FM (or AFM, depending on the signs of the moments) configuration configuration is the most stable one, with the moments aligned along z