[QE-users] k point symmetry and open_grid.x error

[Zeeshan Ahmad]

Hi,

The open_grid.x code with automatic grid of k points in the scf input file produces k points between -0.5 and 0.5 (crystal coords). However, the EPW code requires wave functions for full set of k points from 0 to 1 in the first BZ and gives an error when using k points between -0.5 and 0.5: 

Error in routine epw_setup (1):
coarse k-mesh needs to be strictly positive in 1st BZ 

Unfortunately this is a hybrid functional calculation, so I cannot perform nscf calculation or use all k points in scf due to high memory requirement.

Therefore, I tried using the kpoints.x code in the PW/tools to generate IBZ k points between 0 and 1 and specified the weights in scf file. On running open_grid.x on this calculation, I get the error: 

Error in routine exx_grid_init (1):
     wrong EXX q grid

Is it not possible to use open_grid.x other than for automatically generated k-points (runs fine in that case)? And is there a way to generate the files for full set of k-points between 0 and 1 instead of -0.5 and 0.5?

The files for reproducing the error are at: https://gitlab.com/ahzeeshan/qe-issues/-/tree/master/open_grid <https://gitlab.com/ahzeeshan/qe-issues/-/tree/master/open_grid>

I used a 4 by 4 by 2 grid of k-points to generate the IBZ using kpoints.x code.

Thanks for your help,
Zeeshan




--
Zeeshan Ahmad
Postdoctoral Scholar
Pritzker School of Molecular Engineering
The University of Chicago

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20220205/13c7c42b/attachment.html>