Problem with scGW in VASP.5.3.2

Problems running VASP: crashes, internal errors, "wrong" results.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
mythids

Problem with scGW in VASP.5.3.2

#1 Post by mythids » Wed Dec 05, 2012 5:31 pm

Hi, I'm trying to do scGW calculation but it doesn't work.

In Vaspwiki it says if I want to do scGW calculation, I have to turn on the LPEAD tag so that the WAVEDER file can be recalculated with the upgraded wavefunction. But with the tag on, the process does not go any further right after 'GW_DIAG' in OUTCAR.

I tested systems with moderate or wide band gaps to make sure the LPEAD tag works fine, and all the lines in my INCAR file are exactly same as what vaspwiki recommands. I have tried to find where the problem occured, and it seems like the setfermi loop in dos.F makes an infinite loop for some reason, but I have no idea why it happens.

Could you tell me why it happens and how to fix it(or at least bypass the problem)?
Thank you very much in advance for you time.



<span class='smallblacktext'>[ Edited ]</span>
Last edited by mythids on Wed Dec 05, 2012 5:31 pm, edited 1 time in total.

gbal
Newbie
Newbie
Posts: 17
Joined: Mon Oct 12, 2009 2:25 pm
License Nr.: 5-82
Location: Stockholm, Sweden

Problem with scGW in VASP.5.3.2

#2 Post by gbal » Thu Dec 13, 2012 1:34 pm

It can be several things. Not going to guess what, but i notice that one can use the flag LCALCEPS instead of LPEAD in the GW part to get optical outputs.
Maybe it can help, maybe not.
For my tests, if you omit the optical flags, you also got the same results for EIGENVAL/DOSCAR (but i normally calculate GW0).
Good luck
Last edited by gbal on Thu Dec 13, 2012 1:34 pm, edited 1 time in total.

Post Reply