double-grid (dg) analysis for BSE

Moderators: amolina, andrea marini

Post Reply
sitangshu
Posts: 117
Joined: Thu Jan 05, 2017 8:08 am

double-grid (dg) analysis for BSE

Post by sitangshu » Sun Nov 05, 2017 8:56 am

Dear Sir,

I have a small query about using double-grid (dg) analysis for BSE :

I tested an Si bulk BSE calculation on double-grid (4 4 4) by generating the Doublegrid netcdf file (ypp -m) and it worked fine.
However, when I tested the same by including the electron-phonon correction in BSE, the code complains while solving linear algebra part after kernel (used inversion solver with full perturbative). I realized that my elph was done on a lower scf and nscf grid (2 2 2) while BSE is on dg (4 4 4), so there could be a grid-mismatch. Am I correct?
Is there any trick such that elph is done on a lower scf and nscf grid say 2 2 2 but compatible with BSE done on dg, say 4 4 4?
Your answer would be very helpful for me...

Thanks.
Sitangshu
Sitangshu Bhattacharya
Indian Institute of Information Technology-Allahabad
India
Web-page: http://profile.iiita.ac.in/sitangshu/
Institute: http://www.iiita.ac.in/

User avatar
amolina
Posts: 128
Joined: Fri Jul 15, 2011 11:23 am
Location: Valencia, Spain
Contact:

Re: double-grid (dg) analysis for BSE

Post by amolina » Mon Nov 06, 2017 10:17 am

Dear Sitangshu,

I think you can use the el-ph ndb.QP in a BSE calculations even the grids are different. Have you made the test?

Another option is to use an interpolation to go from 2x2x2 grid to a 4x4x4 grid but that is not yet implemented.

Cheers,

Alejandro.
Alejandro Molina-Sánchez
Institute of Materials Science (ICMUV)
University of Valencia, Spain

Post Reply