constraining in magnetic moments

Message

eltonjose · #1 Post by **eltonjose** » Thu Jul 09, 2009 11:44 am

Hi Vasp forum,
my question concerns on the magnetic constraint calculation in vasp. I have followed the hand on tutorial of vasp workshop and just on section 4
http://cms.mpi.univie.ac.at/vasp-worksh ... tation.htm
I found problems with the convergence of the penalty functional. I have followed exactly the same steps putted in the file
http://cms.mpi.univie.ac.at/vasp-worksh ... dsonIV.pdf
but I can not converge the penalty energy for Fe dimer with constraining the direction of magnetic moments.

I_CONSTRAINED_M = 1
RWIGS = 1.0
LAMBDA = 10
M_CONSTR = 0 0 1 0 1 1

The result that obtained in file OSZICAR is

E_p = 0.00000E+00 lambda = 0.500E+02
ion MW_int M_int
1 0.000 0.000 0.000 0.000 0.000 0.000
2 0.000 0.000 0.000 0.000 0.000 0.000
DAV: 30 -0.953515183919E+01 -0.59393E-04 -0.46316E-04 60 0.688E-02
1 F= -.95351518E+01 E0= -.95425086E+01 d E =0.220703E-01 mag= 0.0003 2.4811 5.4780

E_p = 0.00000E+00 lambda = 0.500E+02
ion lambda*MW_perp
1 0.00000E+00 0.00000E+00 0.00000E+00
2 0.00000E+00 0.00000E+00 0.00000E+00

And my inputs are the same used by the tutorial (INCAR, KPOINTS, POSCAR, POTCAR).

So, do you have any glue for this issue?

Thanks in advance,

Elton.

forsdan · #2 Post by **forsdan** » Thu Jul 09, 2009 3:07 pm

It appears that you get no penalty energy whatsoever. Also you don't seem to get anything under MW_int and M_int. Just to check:

1. Are you using a binary compiled without the flags -DNGXhalf and -DNGZhalf, so that you can run non-collinear calculations?

2. Have you set LNONCOLLINEAR = .TRUE. ?

3. Have you set the LORBIT tag ?

Cheers,
/Dan

<span class='smallblacktext'>[ Edited Thu Jul 09 2009, 05:09PM ]</span>

eltonjose · #3 Post by **eltonjose** » Fri Jul 10, 2009 8:04 am

Exactly:
1. I have compiled without -DNGXhalf and -DNGZhalf flags. I runned almost all non-collinear examples in Vasp workshop ( http://cms.mpi.univie.ac.at/vasp-worksh ... dsonIV.pdf)

2. My INCAR file is:

cat > INCAR<<!
SYSTEM = Fe dimer

!Electronic minimisation
! ENCUT = 270.0 eV ! energy cut-off for the calculation (optional)
! ENAUG = 240.87 eV ! energy cut-off for the augmentation charges
! PREC = Accurate
!DOS related values
! ISMEAR = -5
LNONCOLLINEAR = .TRUE.
MAGMOM = 0 0 3 0 0 3
ISTART = 0
ISYM = 0
LORBIT = 11
VOSKOWN = 1
I_CONSTRAINED_M = 1
RWINGS = 1.2
LAMBDA = 50
M_CONSTR = 0 0 1 1 1 1
GGA_COMPAT = .FALSE.

My KPOINTS and POSCAR is:
K-points
0
Monkhorst Pack
1 1 1
0 0 0
!
cat > POSCAR <<!
Fe dimer
1.00 !universal scaling parameters or if negative total volume of the cell
8.000 0.0000 0.0000 ! lattice vector a(1)
0.000 8.0000 0.0000 ! lattice vector a(2)
0.000 0.0000 8.0000 ! lattice vector a(3)
2
Cartesian
3.00 0.00 0.00
5.00 0.00 0.00
!

Any suggestion?

Thanks in advance.

Cheers,

Elton.

forsdan · #4 Post by **forsdan** » Fri Jul 10, 2009 9:35 am

You have misspelled "RWIGS" in your INCAR file. If I remove the N your files work perfectly. If I leave the misspelling there it doesn't work. So the typo is probably the cause to why it doesn't work for you, since the RWIGS tag will not be explicitly defined in the INCAR file due to this.

Best regards,
/Dan

<span class='smallblacktext'>[ Edited Fri Jul 10 2009, 11:38AM ]</span>

eltonjose · #5 Post by **eltonjose** » Fri Jul 10, 2009 10:13 am

Thanks Dan.

It works properly now.

All the best,

Cheers,

Elton.