Epw becomes silent after a2f calculation in the Eliashberg case

[timonmosko]

Dear epw experts,
I am trying to calculate anisotropic superconducting gap of NbSe2 monolayer.
In the eliashberg calculation I bump into the problem with calculation of spectral function. I do not use special flag for a2f, I follow the way used in hands-on tutorial from this year (https://docs.epw-code.org/_downloads/22 ... /T.H.1.pdf). In my case the code realizes that a2f was not calculated so it calls particular subrutine to calculate it. It really creates prefix.a2f , but it becomes still. The job does not crash, it just remains silent.

Code: Select all

     ===================================================================


     Finish writing dos file aiida.dos


     Finish writing phdos files aiida.phdos and aiida.phdos_proj


     ===================================================================
     Solve anisotropic Eliashberg equations
     ===================================================================


     Finish reading freq file

                  Fermi level (eV) =     0.0000000000E+00
     DOS(states/spin/eV/Unit Cell) =     1.7045262504E+00
            Electron smearing (eV) =     1.0000000000E-01
                 Fermi window (eV) =     4.0000000000E-01
     Nr irreducible k-points within the Fermi shell =        92 out of       109

           2 bands within the Fermi window


     Finish reading egnv file


     Max nr of q-points =       129


     Finish reading ikmap files


     Start reading .ephmat files


     Finish reading .ephmat files

     a2f file is not found to estimate initial gap: calculating a2f files

I attach the epw input file and hope for some ideas, how to resolve this.
Thanks
Timon

[hmori]

Dear Timon,

I would like to know more about the files output from EPW in this job.
Especially I want to know whether prefix.a2f was an empty file, and whether any newer files than prefix.a2f were output.
Since you use iverbosity == 2, some optional files such as prefix.lambda_pairs and prefix.lambda_XX.cube are supposed to be output following prefix.a2f.

Incidentally, these optional files are not required to solve the Eliashberg so that you can run without iverbosity == 2. I don't think this is the solution to your problem, but just FYI.

Best regards,
Hitoshi

[timonmosko]

Dear Hitoshi,

in a meanwhile I have managed to solve the problem. Actually, code was not stopping, the problem was in the density of the grid I have used for scf and phonon calculations together with quite hight ecutoff. So, epw was calculating all of the files, but it took hours, which I did not expect.
However, thanks for your hint, my prefix.a2f was calculated very quickly, so the iverbosity will solve it for me even with for more fine parameters.
Thanks a lot.
Timon