[Q-e-developers] Proposed code additions for 2D systems

[Richard Andrew]

Dear Layla Martin-Samos,

Just a note: 

vc-relax with cell_dofree='2Dshape' should be run with press=0D0 in the input file.  Either the user must know this else the following can be placed in input.f90 to force the requirement.

Insert after 
CALL init_dofree ( cell_dofree ) 
on line 1424 the following code:

if(cell_dofree.eq.'2Dshape') press_ = 0d0

Regards

Richard Charles Andrew

Postgraduate Student
Physics Department
University of Pretoria
South Africa
>>> Layla Martin-Samos  08/31/12 2:41 PM >>>
DEar Richard, thank you very much for your contribution. We will insert it in the trunk as soon as possible.

best regards

Layla

2012/8/31 Richard Andrew <Richard.Andrew at up.ac.za>
Dear QE team,

You requested we send any code additions we may have.   I work in 2D systems and I noticed most DFT codes have no way for you to relax the 2D cell shape while keeping the area constant.  The cell_dofree='shape' option in QE relaxes all degrees of freedom for the cell shape at constant volume which for 2D systems means also changing the vacuum layer.  I have modified the cell_base.f90 module to include a cell_dofree='2Dshape' option which will relax the in-plane (xy plane) cell shape while keeping the area (therefore the vacuum layer) constant.  This also required a change in the deviatoric.f90 code in the flib.  Please find attached these modified files for your consideration.  All modifications are marked between '! RICHARD' comment lines.

I also included in cell_base.f90 a cell_dofree='2Dxy' option where the xy  plane is truly free as you had suggested in your code comment for cell_dofree='xy'.

Regards,

Richard Charles Andrew

Postgraduate Student
Physics Department
University of Pretoria
South Africa



_______________________________________________
Q-e-developers mailing list
Q-e-developers at qe-forge.org
http://qe-forge.org/mailman/listinfo/q-e-developers




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20120904/043df62f/attachment.html>