2D Surface Icutcoul post m_coulombian.F90 fix

GW, Bethe-Salpeter …

Moderators: maryam.azizi, bruneval

Locked
JEJohns
Posts: 55
Joined: Sun May 02, 2010 5:30 pm

2D Surface Icutcoul post m_coulombian.F90 fix

Post by JEJohns » Wed Oct 27, 2010 12:47 am

Hello,
I'm working on computing some quasiparticle calculations for a 2D system, and was running through the convergence tests with respect to ecutwfn and wanted to truncate the coulomb expansion perpendicular to my surface. So I used
icutcoul 2
vcutgeo 1 1 0
rcut 9.5

I'm currently using version 6.2.2, so I made the change in m_coulombian.F90, changing test to check in lines 373 to 379 as suggested, recompiled, and I'm still getting an error. The error I'm getting is

-6.759E-06 1.1707E-05 9.9208E-06
Found q-points with non-zero components along non-periodic direction
This is not allowed, see notes in cutoff_surface.F90
ACTION : Modify the q-point sampling

But my Q-mesh for the screening and kpt grid are 0 in the kz direction. I've attached the log file and input file. Thanks for any suggestions.
--James
Attachments
C4F_Icutcoul.in
(1.76 KiB) Downloaded 324 times
C4F_Icutcoul.log
(45.07 KiB) Downloaded 304 times

Robin
Posts: 61
Joined: Thu Jan 14, 2010 3:18 am

Re: 2D Surface Icutcoul post m_coulombian.F90 fix

Post by Robin » Wed Oct 27, 2010 11:20 am

Dear James,

You can set the following parameters:

gw_nqlwl 1
gw_qlwl 0.00001 0.00001 0.00000

Sincerely,
Guangfu Luo

JEJohns
Posts: 55
Joined: Sun May 02, 2010 5:30 pm

Re: 2D Surface Icutcoul post m_coulombian.F90 fix

Post by JEJohns » Tue Nov 09, 2010 8:30 pm

Sorry I didn't post anything earlier. But this definitely fixed my problem. Thank you
--James

Locked