[Wien] fixed-spin-moment and normal spin-polarized calculations

[carl]

Dear Dr. Blaha

Many thanks for your kind help.
I checked following your suggestion, for the FSM calculation, the 
:CHA=117.229/112.000.
and the :NEC01 =0.9965 (in the normal spin-polarization case, this value is 
1.0002)

By now I have also finished another test using very high k-point 
mesh-containing 28 point in the IBZ for the 32-atom-cell.
the :CHA=117.282/112.000 and the :NEC01=0.99636.
Again, the converged mangetic moment is 5.2 mu_B, not the expected 4 mu_B! 
So it not likely due to the BZ integration ??
In both calculations, I used tetrahedral mothod. I dont expect other 
mothods, such as Gauss and Temp help, right?

As I said earlier, this problem is also not likely due to the structure 
(using the same structure, normal spin-polarization works perfectly), nor 
likely due to the installation ect, as the FSM for bcc-Fe also works fine.

Now, any other possible clue to fix this problem? Has anybody else 
experience similar problem? (I checked the mailing list but got no answer)

thanks again and with my best wishes
Carl Cui

School of Physics
The University of Sydney, Australia


>>Hard to say without further details (scf file).
>>I'd expect the problem could be in the BZ integration (fermi energy,..)
>>
>>Check the scf file for :NOE   and ':CHA  :' and :NEC01
>>For NOE=232/224 you should get CHA=116/112 respectively. If you don't, most
>>likely also the total charge (NEC01) is wrong.
>>
>>                                       P.Blaha