Magnetic moment

[saswata]

I want to do a calculation for a simple Mn dimer kept ina cubic cell of size 20X20X20 with magnetic moment 10( both 5).So my INCAR is like this---
ISTART = 0
ISMEAR = 0
SIGMA = 0.1
IBRION=2
POTIM = 1.0
NSW=20
PREC=Med
VOSKOWN=1
ISYM=0
ISPIN=2
NUPDOWN=10
LORBIT = 12
So as I fix the magnetic moment 10 with NUPDOWN=10, this moment should not change at all.I see in the OSZICAR the magnetization remains 10 all over.But due to the LORBIT=12 I can see the final magnetization of the system and which is like this----

magnetization (x)

# of ion s p d tot
----------------------------------------
1 0.147 0.009 4.100 4.256
2 0.147 0.009 4.100 4.256
------------------------------------------------
tot 0.294 0.017 8.201 8.512

here we see the final moment is 8.512.So my question is that even if I fix it 10 why does it change to 8.512?I feel I have done something wrong.So can you please tell me where is the fault?or what are the necessary flags that I need to use to calculate the optimization of two Mn dimer placed in a cubic cell each having magnetic moment 5?I also tried with MagMom tag by speciffying each moment as 5 instead of NUPDOWN=10.But I got the same result.In the final part of the OUTCAR file I got the magnetization something ~8.5 instead of 10.Please reply.

[forsdan]

There is nothing wrong. The magnetization for each atom as written to the OUTCAR file only takes into account the charge inside the Wigner-Seitz radius and hence the corresponding difference between spin-up and spin-down densities may appear different from your total magnetization in your cell. The total magnetization is still 10. If you want the magnetization as written to the OUTCAR to be closer to the real value, you can increase the RWIGS-value to account for more charge (see the RWIGS and LORBIT sections in the manual), but keep in mind that you should avoid overlap between the spheres.

Best regards,
/Dan

<span class='smallblacktext'>[ Edited Sun Feb 10 2008, 08:33PM ]</span>