[Wien] Inconsistency in kgen
Peter Blaha
peter.blaha at tuwien.ac.at
Tue Mar 19 22:48:58 CET 2024
Hi,
Yes, my fix was not correct. The bct and bco lattices are special cases
and are also discussed in literature.
While the direct lattice vectors in bct have all the same length, the
reciprocal lattice is face-centered tetragonal and thus they have
different length. Nevertheless as far as I understand it is usually
recommended to use the same divisions of them, when constructing a
k-mesh. This was also enforced in WIEN2k, only the recent addition of
specifying a delta-k was breaking this, but led to wrong multiplicities.
The algorithm for finding the multiplicity does not work for bct, when
the divisions are not the same.
The present fix is to go back to the original bravai.f and use the new
basdiv.f, which enforces equal k-divisions in the bct/bco case.
Thanks for checking.
Peter Blaha
Am 19.03.2024 um 08:09 schrieb balabi via Wien:
> Dear Prof. Peter Blaha,
>
> I think I found another issue with kgen -1 mode especially concerned
> with I4/mmm symmetry.
>
> As previously mentioned, the '-1' mode of kgen generates uneven k
> divisions, as you rightly pointed out. However, I've observed potential
> issues with the weight of the uneven k mesh, particularly in the context
> of I4/mmm symmetry (and possibly other symmetries as well). This
> discrepancy appears to lead to inaccuracies in determining the Fermi energy.
>
> Here is how I benchmarked:
>
> 1. run_lapw -p till converge for CaFe2As2 and 10 10 10 shift kmesh.
> save_lapw -d scf
> 2. x kgen generate 9 9 9 non shifted mesh | x lapw1 -p; x lapw2 -p
> -fermi | get Ef=0.5076760931
> 3. x kgen generate 9 9 13 non shifted mesh | x lapw1 -p; x lapw2 -p
> -fermi | get Ef=0.5028504280
> 2. x kgen -fbz generate full 9 9 9 non shifted mesh | x lapw1 -p; x
> lapw2 -p -fermi | get Ef=0.5076760047
> 3. x kgen -fbz generate full 9 9 13 non shifted mesh | x lapw1 -p; x
> lapw2 -p -fermi | get Ef=0.5077417577
>
> You can observe the significant difference in the Ef obtained from the 9
> 9 13 kmesh in the irreducible Brillouin zone (IBZ). However, the Ef
> calculated from the full Brillouin zone (FBZ) using the same mesh
> appears normal. Consequently, I suspect there may be an issue with the
> weight of k-points in the 9 9 13 mesh generated by your modified
> kgen. On the other hand, equal division kmesh 9 9 9 gives consistent Ef
> for both ibz and fbz.
>
> What is more, for 9 9 13 IBZ kmesh, when viewing fermi surface. xcrysden
> 1.6.2 gives error. xcrysden 1.5.60 gives completely wrong fermi surface.
>
> Please help, thank you very much!
>
> best regards
>
> ------------------------------------------------------------------------
>
>
>
> ------------------ Original ------------------
> *From:* "A Mailing list for WIEN2k users" <peter.blaha at tuwien.ac.at>;
> *Date:* Mon, Mar 11, 2024 08:33 PM
> *To:* "wien"<wien at zeus.theochem.tuwien.ac.at>;
> *Subject:* Re: [Wien] Inconsistency in kgen
>
> Ups.
> Here it comes.
>
> Am 11.03.2024 um 13:26 schrieb balabi via Wien:
> > Dear Prof. Peter Blaha
> >
> > Thank you so much for your reply! But I can not find your attachment of
> > bravai.f.gz
> >
> > best regards
> >
> > ------------------------------------------------------------------------
> >
> >
> >
> > ------------------ Original ------------------
> > *From:* "A Mailing list for WIEN2k users" <peter.blaha at tuwien.ac.at>;
> > *Date:* Mon, Mar 11, 2024 04:03 PM
> > *To:* "wien"<wien at zeus.theochem.tuwien.ac.at>;
> > *Subject:* Re: [Wien] Inconsistency in kgen
> >
> > Hi,
> >
> > Thank you very much for your report. I can confirm the problem.
> >
> > Both, for bct and bco lattices (body-centered tetragonal or
> > orthorhombic) kgen enforced in default modes equal divisions of the
> > reciprocal lattice vectors (as it should be for bcc). This was not a
> > good choice and the selection made by option "-1" (mesh density in
> > bohr^-1) is correct.
> >
> > I attach a modified bravai.f.gz file, which should be copied and
> > unziped in SRC_kgen, then recompiled (make; cp kgen ..).
> >
> > PS: The prevous setting was not a problem if your k-mesh is converged
> > (besides the larger computational effort), but may lead to extra
> > inaccuracy for non-converged meshes.
> >
> > Peter Blaha
> >
> > Am 10.03.2024 um 14:48 schrieb 巴拉比 via Wien:
> >> Dear wien2k developers and users,
> >>
> >> I am using wien2k 23.2 and working with CaFe2As2 structure which has
> >> I4/mmm symmetry. I am trying to generate klist using kgen. The kgen
> >> has several mode:
> >>
> >> the 1st mode is to specify k-mesh density
> >>
> >> /NUMBER OF K-POINTS IN WHOLE CELL: (0 for 3 divisions of K, -1 for
> >> delta-K)/
> >> /-1/
> >> / length of reciprocal lattice vectors (bohr^-1): 0.898 0.898
> 1.203/
> >> / Specify density of k-mesh in bohr^-1:/
> >> /0.2/
> >> / Shift of k-mesh allowed. Do you want to shift: (0=no, 1=shift)/
> >> /0/
> >> / 17 k-points generated, ndiv= 5 5 7/
> >> / delta-K (bohr^-1): 0.1796 0.1796 0.1719/
> >> /KGEN ENDS/
> >> /0.004u 0.016s 1:03.31 0.0% 0+0k 0+88io 0pf+0w/
> >>
> >> As you can see, the ndiv=5 5 7
> >>
> >> The 2nd mode is to specify number of k points
> >>
> >> /NUMBER OF K-POINTS IN WHOLE CELL: (0 for 3 divisions of K, -1 for
> >> delta-K)/
> >> /1000/
> >> / length of reciprocal lattice vectors (bohr^-1): 0.898 0.898
> 1.203/
> >> / Shift of k-mesh allowed. Do you want to shift: (0=no, 1=shift)/
> >> /0/
> >> / 102 k-points generated, ndiv= 10 10 10/
> >> / delta-K (bohr^-1): 0.0898 0.0898 0.1203/
> >> /KGEN ENDS/
> >> /0.026u 0.003s 0:09.41 0.2% 0+0k 0+344io 0pf+0w/
> >>
> >> as you can see unlike -1 mode, the ndiv=10 10 10 which is even.
> >>
> >> The 3rd mode is to specify ndiv explicitly, and here comes the problem
> >>
> >> /NUMBER OF K-POINTS IN WHOLE CELL: (0 for 3 divisions of K, -1 for
> >> delta-K)/
> >> /0/
> >> / length of reciprocal lattice vectors (bohr^-1): 0.898 0.898
> 1.203/
> >> / Specify 3 mesh-divisions (n1,n2,n3):/
> >> /5,5,7/
> >> / Lattice symmetry requires equal mesh in x and z direction/
> >> / Specify 3 mesh-divisions (n1,n2,n3):/
> >>
> >> If I set ndiv as 5 5 7 as -1 mode. It just does not work. kgen insists
> >> equal mesh on x and z. So I must input even ndiv like 5 5 5.
> >>
> >> So the current issue is why the ndiv obtained with the -1 mode in kgen
> >> is not uniform for the CaFe2As2 system, whereas ndiv obtained with the
> >> k-point mode is uniform, and using the 0 mode, it's impossible to set
> >> non-uniform ndiv at all？
> >>
> >> finally, I am attaching the struct file of CaFe2As2 as below
> >>
> >> /blebleble /
> >> /B LATTICE,NONEQUIV.ATOMS: 3 139 I4/mmm /
> >> / RELA /
> >> / 7.383336 7.383336 21.922849 90.000000 90.000000 90.000000/
> >> /ATOM -1: X=0.00000000 Y=0.00000000 Z=0.00000000/
> >> / MULT= 1 ISPLIT=-2/
> >> /Ca1 NPT= 781 R0=0.00005000 RMT= 2.50000 Z: 20.0 /
> >> /LOCAL ROT MATRIX: 1.0000000 0.0000000 0.0000000/
> >> / 0.0000000 1.0000000 0.0000000/
> >> / 0.0000000 0.0000000 1.0000000/
> >> /ATOM -2: X=0.50000000 Y=0.00000000 Z=0.25000000/
> >> / MULT= 2 ISPLIT=-2/
> >> / -2: X=0.00000000 Y=0.50000000 Z=0.25000000/
> >> /Fe1 NPT= 781 R0=0.00005000 RMT= 2.28 Z: 26.0 /
> >> /LOCAL ROT MATRIX: 0.7071068-0.7071068 0.0000000/
> >> / 0.7071068 0.7071068 0.0000000/
> >> / 0.0000000 0.0000000 1.0000000/
> >> /ATOM -3: X=0.00000000 Y=0.00000000 Z=0.63490625/
> >> / MULT= 2 ISPLIT=-2/
> >> / -3: X=0.00000000 Y=0.00000000 Z=0.36509375/
> >> /As1 NPT= 781 R0=0.00005000 RMT= 2.17 Z: 33.0 /
> >> /LOCAL ROT MATRIX: 1.0000000 0.0000000 0.0000000/
> >> / 0.0000000 1.0000000 0.0000000/
> >> / 0.0000000 0.0000000 1.0000000/
> >> / 16 NUMBER OF SYMMETRY OPERATIONS/
> >> / 1 0 0 0.00000000/
> >> / 0 1 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 1/
> >> /-1 0 0 0.00000000/
> >> / 0-1 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 2/
> >> / 0-1 0 0.00000000/
> >> / 1 0 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 3/
> >> / 0 1 0 0.00000000/
> >> /-1 0 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 4/
> >> /-1 0 0 0.00000000/
> >> / 0 1 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 5/
> >> / 1 0 0 0.00000000/
> >> / 0-1 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 6/
> >> / 0 1 0 0.00000000/
> >> / 1 0 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 7/
> >> / 0-1 0 0.00000000/
> >> /-1 0 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 8/
> >> /-1 0 0 0.00000000/
> >> / 0-1 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 9/
> >> / 1 0 0 0.00000000/
> >> / 0 1 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 10/
> >> / 0 1 0 0.00000000/
> >> /-1 0 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 11/
> >> / 0-1 0 0.00000000/
> >> / 1 0 0 0.00000000/
> >> / 0 0-1 0.00000000/
> >> / 12/
> >> / 1 0 0 0.00000000/
> >> / 0-1 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 13/
> >> /-1 0 0 0.00000000/
> >> / 0 1 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 14/
> >> / 0-1 0 0.00000000/
> >> /-1 0 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 15/
> >> / 0 1 0 0.00000000/
> >> / 1 0 0 0.00000000/
> >> / 0 0 1 0.00000000/
> >> / 16/
> >>
> >>
> >> best regards
> >> ------------------------------------------------------------------------
> >>
> >> _______________________________________________
> >> Wien mailing list
> >> Wien at zeus.theochem.tuwien.ac.at
> >> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >> SEARCH the MAILING-LIST
> at:http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
> >
> > --
> > -----------------------------------------------------------------------
> > Peter Blaha, Inst. f. Materials Chemistry, TU Vienna, A-1060 Vienna
> > Phone: +43-158801165300
> > Email:peter.blaha at tuwien.ac.at
> > WWW:http://www.imc.tuwien.ac.at WIEN2k:http://www.wien2k.at
> > -------------------------------------------------------------------------
> >
> >
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> > SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>
> --
> --------------------------------------------------------------------------
> Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-1-58801-165300
> Email: peter.blaha at tuwien.ac.at WIEN2k: http://www.wien2k.at
> WWW: http://www.imc.tuwien.ac.at
> -------------------------------------------------------------------------
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
--
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300
Email: peter.blaha at tuwien.ac.at WIEN2k: http://www.wien2k.at
WWW: http://www.imc.tuwien.ac.at
-------------------------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bravai.f.gz
Type: application/x-gzip
Size: 2826 bytes
Desc: not available
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20240319/80b7d63b/attachment.gz>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: basdiv.f.gz
Type: application/x-gzip
Size: 1494 bytes
Desc: not available
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20240319/80b7d63b/attachment-0001.gz>
More information about the Wien
mailing list