[Wien] AFM supercell

[ROBERTO LUIS IGLESIAS PASTRANA]

Dear all

I'm trying to run an AFM calculation in a 16 atom Cr supercell. I want to have the Cr magnetic moment values for reference purposes, in order to compare to other systems of the same size and later larger ones. The normal bcc unit cell AFM calculation runs fine, following the procedures outlined in the UG, section 4.5.4, and I get the same magnetic moments for both Cr atoms in the cell but with different signs.
 I set up a 16 atom supercell and then flip the spins of the Cr atoms in consecutive pairs in case.inst. I also tried to flip just the second 8 atoms, unphysical as it may be, and consequently there is a crash in dstart. When I perform either a runsp_lapw or a runafm_lapw calculation I always get zero, or close to zero magnetic moments in all sixteen Cr atoms, as if the system was in a nonmagnetic state. The question is: What else do I need to do in order to get final magnetic moments with the same absolute values but with alternate different signs, as for the bcc unit cell? Or am I writing nonsense?
I'm using WIEN2k_08.3 and running in 8 processors via k-point parallelization.

Thanks in advance

Roberto

---------------------------------------
Roberto Iglesias
Physics Department
University of Oviedo
Calvo Sotelo, s/n
33007 Oviedo
SPAIN
Tel: +34985102898
Fax: +34985103324
email: roberto at uniovi.es
---------------------------------------

--------------------------------------