The default direction for magnetic moment in a collinear calcultaiotns.

Message

Neutrino · #1 Post by **Neutrino** » Sat May 14, 2011 10:09 am

The manual says that the default is in z-direction (the 3rd vector). Also According to the literature the first meta-stable state in DFT+U collinear calculations for UO2, the fluorite structure should be distorted such that the c-axis becomes shorter than a,b and (a=b). It is claimed that the spin in c-direction causes this contraction.

I tried to reproduce this and I got everything correct except that the contraction occurred in a-direction (the first vector) rather than the c-direction. I did not change the defaults for the the magnetic moment direction. Any hint about why I get that.

I use VASP 4.6.31 and I start my calculations from an Anti-ferromagnetic guess as it should be for UO2.

beck · #2 Post by **beck** » Sun May 15, 2011 12:14 pm

Did you comile VASP with the NGXhalf instead of NGZhalf?

Neutrino · #3 Post by **Neutrino** » Mon May 16, 2011 4:18 am

I did not compile it myself. Is there a way to check which was used in compilation from the output files?

#4 Post by **admin** » Mon May 23, 2011 7:41 pm

sorry, but I fear I don't understand: the fluorite structure is cubic, and unless you impose some external field, strain,... in a certain direction, there is no principal axis which could be identified as c (with respect to what would you choose?)