Issues with running TB2J calculation

[Kapil Sharma]

Dear Dr. He,

Thank you for your work on TB2J. I am new to TB2J, and I am running calculations with QE and Wannier90 before TB2J to find the exchange parameters for Samarium. 

I have an issue running the TB2J. Up to Wannier90, it is fine, yielding good converging results, with num_iter=0, and the Wannier centers are very close to the atoms, with small spreads. However, when I run TB2j, I receive an error message, which is attached as error_message.png. 

If I convert the scf. in lattice parameter in POSCAR like format, and change the posfile from scf.in to  POSCAR, the calculations run, but the Magmom is zero and the exchange value is nan, as attached as exchange_out*. I am puzzled by the result. Is it a unit (Armstrong to Bohr ) handling issues, or is it a different one?  I am seeking assistance with this matter. Thank you so much for your time and consideration.

Sincerely,

Kapil Sharma

[Xu He]

Hi, 

In the command you have both spin up and spin down Wannier functions with the prefix wannier90_up.  With this, the input spin up and down are exactly the same and will be non-spin-polarized. 

Best regards, 

HeXu