Forum EPW
http://epwforum.uk/

a fatal error with EPW directory
http://epwforum.uk/viewtopic.php?f=6&t=696
Page 1 of 1

Author:  anindya@iiita [ Wed Aug 29, 2018 5:46 am ]
Post subject:  a fatal error with EPW directory

Dear Developers,
I am suffering in a big issue,I have calculated my wannier band structure of Graphene which is fully converged(zero band gap at k point) and I can see this in graphene_band_dat file which was generated by wannier90 calculation.But when I am using the same wannier90 values(disentanglement and wannier parameter values copied to epw.in) in epw.in script then I am getting wrong result.The graphene_band_dat file generated by epw.in during calculation is giving different values(non converging) than wannier90.Exact same wannier90 parameter values are used in epw.in script using wdata and dis_win,dis_froz still the file is giving wrong result and naturally my a2f output is giving unphysical values.Is this any kind of issue in the epw 4.3.0 version that I am using?Please let me know.Now it is becoming a serious issue.

Thanks and regards,
Anindya Bose

Author:  sponce [ Wed Aug 29, 2018 2:26 pm ]
Post subject:  Re: a fatal error with EPW directory

Hello,

It depends which version of Wannier you used.

EPW uses wannier2.1.0 in library mode.

As a first step you should indeed try the newest QE 6.3 that contains EPW 5.0

Differences between EPW and Wannier bandstructure might stem from the way the Wigner-Seitz cell is constructed.
In EPW 5.0, the phonon WS cell is constructed differently (better) but the electron WS is the same (Gamma centred).
This should reproduce the wannier90 results if you do not use the wannier guiding center input variable. In that case you can have different results.

Best wishes,
Samuel

Author:  anindya@iiita [ Wed Aug 29, 2018 4:56 pm ]
Post subject:  Re: a fatal error with EPW directory

Dear Dr. Samuel Poncé ,
I am unable to resolve this, please suggest me a way to rectify this.I have taken guided centre= true in wannier90 and epw.in.Using epw 5.0 will be the solution of this problem?

Thanks and regards,
Anindya Bose

Author:  anindya@iiita [ Fri Aug 31, 2018 6:55 am ]
Post subject:  Re: a fatal error with EPW directory

Dear Experts,
I am not able to understand the issue between wannier90 and epw wannier,if they share same wannier library then my wannier90 output(.wout file) and epw wannier output (.wout) should match exactly.But in terms of spreading they are entirely different for the same wannier input(calc.win file).I have uploaded all the wannier parameters(with the parameters I achieved correct and converged wannier band structure) exactly in the epw.in file but still the .wout file and calc_band_dat is entirely different from wannier90 result(non converging and comparatively large spread).Any suggestion in this issue is highly appreciated.Even I tried with epw 5.0 version but the results are same.

Thanks and regards,
Anindya Bose

Author:  sponce [ Sat Sep 01, 2018 3:16 pm ]
Post subject:  Re: a fatal error with EPW directory

Hello,

Some parameters cannot be passed to wdata. For example atoms positions etc are read from the .save folder and should not be input in epw.in

Carefully look at the two .wout files and try to see if there are difference that you do not expect. For example initial Wannier center positions etc should be the same.

Also I would avoid using guiding_center if you want to compare between EPW and Wannier90.

Best wishes,
Samuel

Author:  anindya@iiita [ Sat Sep 01, 2018 9:50 pm ]
Post subject:  Re: a fatal error with EPW directory

Dear Experts,
I have resolved this error by slightly changing the BZ coordinate decimal places.I think I should update about this.

Thanks and regards,
Anindya Bose

Page 1 of 1 All times are UTC
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/