The ABINIT Users Mailing List ( CLOSED )

[889180013@s89.tku.edu.tw]

Dear Abinit users,
    I have a question about the occupation number in
spin polarized calculation. I think that total occupation
numbers should be the number of kpoints times number of 
electrons per unit cell in any kinds of systems. When I 
test ./Test/v3/t15.in it sums to 8 electrons for Fe atom 
even I change the kpt to 0.5 0.0 0.0 from 0.0 0.0 0.0.
But if I try to increase the numner of kpoints (3x3x3,that
is 6 special k-points), the sum of occupation number is 
not 48 electrons. I get 49.07 electrons. Is it OK?
I need some comment.

Regards,
Jerome

This is my input file:
# Fe atom in a big box
# Perform the computation of the Gamma dynamical matrix.

 ndtset 2

#Dataset 1 : SCF
  tolvrs  1.0d-13
   nstep  30

#Dataset 2 : RF at q=0 0 0   (this is always needed for IFCs)

#Common data

 ecut 22
 acell 3*6.50
 rprim 0 .5 .5  .5 0 .5  .5 .5 0

 ixc    1
 natom  1
 nband 10
 nsppol 2
 spinat 0.0 0.0 4.0
 ntypat  1
 occopt 4
 nkpt1  1
 kpt1 0.5 0.0 0.0
kptopt2 1
ngkpt2 3 3 3
# test in periodic crystal case

 wtk1 1.0

 typat  1
 xred  0.0 0.0 0.0
 znucl 26.0