Issue with TB2J calculation not finishing + question about window selection

[Iram Bel]

Dear TB2J community,

I am currently working on VASP + TB2J calculations to extract the exchange interactions Jij​ for magnetic compounds. My workflow is as follows: I first perform a standard SCF calculation with VASP, and then run TB2J.

The problem I am facing is that TB2J successfully generates the Jij​ values and all the expected output files, but the calculation never actually finishes. It keeps running indefinitely without producing any new results or modifying the output. I have to manually stop the job myself; otherwise, it just continues without end. This behavior is quite concerning, and I was wondering if you could help me understand why this happens.

In addition, I would really appreciate a more detailed explanation or guidelines for choosing the proper energy windows. I have read the documentation, but I still struggle to make the correct choice for my systems.

Thank you very much for your support.

Best regards,
Iram

[Xu He]

(Dear Iram, 
Could you share the related files: The wannier90 files, POSCAR, and the TB2J command, and  the TB2J version?
For choosing the energy window, I recommend reading this webpage: https://www.wanniertools.org/tutorials/high-quality-wfs/  

(Sometimes it is a bit difficult to get access to this page. You can try a few times, or use a different web browser). 

Best regards, 

HeXu

--
You received this message because you are subscribed to the Google Groups "TB2J" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tb2j+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/tb2j/573b21e9-0d62-41d7-b8d9-4de4336b8212n%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Iram Bel]

Dear HeXu,

Thank you very much for your quick reply.

I have attached the folder containing all the relevant files. 

The command I used is the following:

python3 "${SCRIPT_DIR}/wann2J.py" \ --posfile "${INPUT_DIR}/POSCAR" \ --efermi 4.327 \ --kmesh 4 4 4 \ --elements Mn \ --prefix_up "${INPUT_DIR}/wannier90.1" \ --prefix_down "${INPUT_DIR}/wannier90.2" \ --emin -10 \ --emax 0 \ --rcut 9

The TB2J version used is TB2J 0.9.0.1.

Thank you in advance for your help.

Best regards,
Iram

 SrMnO3_save.tar.gz

To view this discussion visit https://groups.google.com/d/msgid/tb2j/CAHwPuaBPD%3Dks4nyRrOKuP8AO7SnbUAS6yTwT6S3sfGjBQFqL%3DQ%40mail.gmail.com.

[Xu He]

Hi, 

Could you try with the most recent version of TB2J. 

You can upgrade it with

pip install -U TB2J 

I run it with the files you sent and it works fine (it finishes in less than 10 seconds). 

Best regards, 

HeXu

[Iram Bel]

Hi, thank you for your help.

I tried with the latest version of TB2J and I encountered the same issue. I think I may have forgotten to mention that the first SCF calculation converges without any problem. The issue arises with the second DFT calculation, where I activate the Wannier90 parameters in the INCAR file.

In this case, the job keeps running indefinitely and never finishes. It generates all the expected Wannier90 output files correctly, but the calculation itself continues endlessly without producing any further output or modifications. I once let it run for more than a week, but it still did not progress.

Thank you in advance for your support.

Best regards,
Iram

[Xu He]

Hello, 

I thought the problem was with the TB2J calculation but apparently I misunderstood. 

You can go with the TB2J calculations once  the Wannier90 files are generated as these are the only things needed. So I guess it is safe to stop the VASP calculation. I have little expertise in VASP calculations. Perhpas you could try with a different version of VASP or compiled on a different machine. 

Best regards, 

HeXu 

To view this discussion visit https://groups.google.com/d/msgid/tb2j/27ae1a21-2a6a-400b-b6d9-ea71dcd03b46n%40googlegroups.com.