Page 1 of 1
New Coulomb cutoff in version 5.1
Posted: Wed May 25, 2022 6:36 am
by sdwang
Dear developers,
I noticed a new cutoff strategy has been implemented in the new version 5.1. As for the "slab z" cutoff of 2D system, how to set the cutoff size along the z direction? Could we just set it as the size of "box xy"? And another question is if we want to take the benefit of the rimW integration, as described in the paper "arXiv:2205.11946v1", to do GW calculation, which parameter we should activate? Does this rimW technique accelerate BSE calculations w.r.t the k-point also?
Thanks!
Best
Shudong
Re: New Coulomb cutoff in version 5.1
Posted: Wed May 25, 2022 8:46 am
by Daniele Varsano
Dear Shudong,
how to set the cutoff size along the z direction?
There is no need to set any zcut, it is automatically set by the code.
About the rimW a tutorial will be published soon. Please find attached an input example.
The parameter to set to activate it are:
Code: Select all
RIM_W. # as runlevel
RandQpts=3000000 # [RIM] Number of random q-points in the BZ
RandGvec= 97 RL # [RIM] Coulomb interaction RS components
RandGvecW = 15 RL
CUTGeo= "slab z"
The meaning of each variable should be clear from the paper you mention.
Code: Select all
Does this rimW technique accelerate BSE calculations w.r.t the k-point also?
You can try, but it is discouraged, still some systematic testing is needed.
Best,
Daniele
Re: New Coulomb cutoff in version 5.1
Posted: Wed May 25, 2022 9:12 am
by sdwang
Dear Daniele,
Thank you for the quick reply. I will try it.
Best
Shudong
Re: New Coulomb cutoff in version 5.1
Posted: Mon May 30, 2022 3:17 pm
by Sumit
Dear Developer,
Is no box sides are needed for this slab z method as found in your shared input_RIMW.in file?
CUTGeo= "slab z"
% CUTBox
0.000000 | 0.000000 | 0.000000 | # [CUT] [au] Box sides
%
Thank You
Regards
Sumit Kukreti
Ph.D. Scholar
IIT Jodhpur INDIA
Re: New Coulomb cutoff in version 5.1
Posted: Tue May 31, 2022 7:47 am
by Daniele Varsano
Dear Sumit,
no, the slab cutoff does not require any box size, as the cut is done on the Wigner-Seitz cell as explained here:
https://arxiv.org/pdf/2205.11946.pdf
Best,
Daniele
Re: New Coulomb cutoff in version 5.1
Posted: Wed Jun 29, 2022 2:40 am
by sdwang
Dear all,
I am trying the newly published Coulomb cutoff and it stopped with anything complaint in the log and the report(It seems the problem is with RIM integrals). My system is a 2D hBN.
Attached are the input, log, and report files.
Thanks!
Best
Shudong
PS: When I finished the chi calculation with RIM_W and then remove it, the sigma part goes smoothly and I got the final GW results. But this would miss something about rim integral.
Re: New Coulomb cutoff in version 5.1
Posted: Mon Jul 04, 2022 3:48 pm
by Alberto Guandalini
Dear Shudong,
please see my answer in the discussion:
viewtopic.php?p=12029#p12029
It applied to your problem also.
Best,
Alberto