[Pw_forum] "not orthogonal operation" during vc-relax when optimizing graphene

[Giovanni Pizzi]

Dear Peng Tao,

as Mohnish Pandey has reported, cell_dofree = 'xy' is not the right option for you.

In recent versions of QE (I don't know if already in 5.0.2, for sure in the SVN version) there is the
cell_dofree = '2Dxy'
option that allows the first two cell vectors to freely change on the plane (i.e. both the x and y coordinates of the first two vectors are free to move), that is what you want.

If you have an old version that does not accept this option, you may want to switch to a recent version of QE, or simply add the following lines to you Modules/cell_base.f90 file (and recompile)

            CASE ('2Dxy')
              iforceh      = 0
              iforceh(1,1) = 1
              iforceh(2,2) = 1
              iforceh(1,2) = 1
              iforceh(2,1) = 1

(see here to discover where to put these lines:
http://qe-forge.org/gf/project/q-e/scmsvn/?action=browse&path=%2Ftrunk%2Fespresso%2FModules%2Fcell_base.f90&view=markup
)

Best,
Giovanni Pizzi

--
Giovanni Pizzi
Post-doctoral Research Scientist
EPFL STI IMX THEOS
MXC 340 (Bâtiment MXC)
Station 12
CH-1015 Lausanne (Switzerland)
Phone: +41 21 69 31124




On 3 May 2013, at 02:26, mohnish pandey wrote:

Dear Peng,

The cell you might be using is 'hexagonal' and the 'vc-relax' that you are doing has two degrees of freedom in x and y direction which are orthogonal due the flag 'cell_dofree = xy' you set. Since the symmetry is breaking in the second step due to relaxation in x and y direction it is changing the number of k-points. That's why you are getting that error. A simple work-around for this is you use a bit bigger cell with orthorhombic symmetry (in which you have orthogonal 'a' and 'b' vectors). In orthorhombic cell symmetry won't break due to 'cell_dofree = xy'.


On Fri, May 3, 2013 at 4:13 AM, Peng Tao <ptao10b at imr.ac.cn<mailto:ptao10b at imr.ac.cn>> wrote:
Dear all,

Recently, I just want to putting 'vc-relax' to optimize the graphene structure,
but it repots an error:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from checkallsym : error #         2
     not orthogonal operation
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

And I set 'cell_dofree = xy' in the vc-relax run.

I know, it is no need to apply vc-relax to graphene, and only relax can aquire
my result. But if I want to relax bi-layer and mutilayer graphene and my ultimate
goal is to get their phonon dispersion curves(so the symmetry should not be broken),
how could I do? Is there any convienient method to relax them?

Thanks a lot.

Yours,
Peng Tao

--
-------------------------------------------------------------------
PH.D. candidate Peng Tao
Magnetism and Magnetic Materials Division
National Laboratory for Material Science
Institute of Metal Research, Chinese Academy of Sciences
Phone  +86-024-83978751<tel:%2B86-024-83978751>
-------------------------------------------------------------------



_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org<mailto:Pw_forum at pwscf.org>
http://pwscf.org/mailman/listinfo/pw_forum



--
Regards,
MOHNISH,
-----------------------------------------------------------------
Mohnish Pandey,
PhD Student,
Center for Atomic Scale Materials Design,
Department of Physics,
Technical University of Denmark
-----------------------------------------------------------------
_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org<mailto:Pw_forum at pwscf.org>
http://pwscf.org/mailman/listinfo/pw_forum

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20130504/ae00eb43/attachment.html>