Problems with long lines in INCAR

Message

olle · #1 Post by **olle** » Fri Oct 27, 2006 12:29 pm

Hi

I'm trying to run noncollinear supercell calculations, with a supercell of 96 atoms. I want to specify MAGMOM for each atom. This means I a line with 96*3 elements, and vasp doesn't seem to be able to read this, no matter if I specify it as 1 1 1 1 ... or 288*1, it only gives me
IERR=0, found 198 elements.

Am I doing something stupid or is it a known limitation?

By the way, I have no problems with cells with smaller number of atoms

Olle

lahaye · #2 Post by **lahaye** » Sat Oct 28, 2006 8:15 am

With 'elements' do you mean 'degrees of freedom'?

As far as I understand, you need to specify only one
number per atom in the MAGMOM line. So in your case:

MAGMOM = 96*1

I think it specifies the 'majority' vs. 'minority' per atom, and
not the vector moment per degree of freedom per atom.
R.

olle · #3 Post by **olle** » Mon Oct 30, 2006 12:50 pm

Yes, I do mean degrees of freedom. And no, in this case three numbers has to be specified, if not vasp won't run. For example, for a system with a single atom

MAGMOM = 1 0 0

Gives an initial moment in the x-direction. This works fine, the problem occurs when I have large supercells.

lahaye · #4 Post by **lahaye** » Sat Nov 04, 2006 4:41 pm

The VASP Guide seems to tell a different story.
Check the example given in:
http://cms.mpi.univie.ac.at/vasp/guide/ ... 0000000000
This system with two atoms, has

Code: Select all

MAGMOM = 1 -1

I myself have a system with 11 atoms, and I use successfully:

Code: Select all

MAGMOM = 10*2 1*1

Rob.

#5 Post by **admin** » Tue Nov 14, 2006 9:45 am

to read long input lines (> than 255 characters) from INCAR, please set
#define LONGCHAR
in subroutine RDATAB (drdatab.F) (in vasp.4.lib)
and re-compile the library and the code.
then the sting length is increased to 32767 characters