This thread is locked.Only browsing is available.
Top Page > Browsing
k-point mesh
Date: 2007/05/01 02:06
Name: Peter


Dr. Ozaki,

How to define number of k-points which are actually used in calculation in spin-polarized case? I think, number of k-points in regular mesh (as implemented in openmx) much more than Monkhorst-Pack (MP) k-point mesh is used. Why don't use MP k-point mesh in OpenMX?

Regards,
Peter I.
メンテ
Page: [1]

Re: k-point mesh ( No.1 )
Date: 2007/05/02 18:35
Name: T.Ozaki

Hi,

Since OpenMX was originally developed for large-scale calculations,
we did not take care of an efficient choice of the k-points.
At certain stage, we may consider the development.

Regards,

TO
メンテ
Re: k-point mesh ( No.2 )
Date: 2007/05/02 22:33
Name: JessK


Dr. Ozaki,

I have a question too.

1) Do you consider to implement a variable cell relaxation in OpenMX?

2) I looked into a source code and see that there is a possibility to write output files in xcrysden format, however, the calling of this function (out_xsf) is disabled.
Does it actually work?

Regards,
jk
メンテ
Re: k-point mesh ( No.3 )
Date: 2007/05/18 00:25
Name: Vasilii Artyukhov

Dear Dr. Ozaki,

It seems that a minor modification of the code could be very helpful for those of us who need to use k-points. For instance, when a 4x4x4 grid is used, OpenMX permorms the calculation at k-points with coordinates {-0.75, -0.25, 0.25, 0.75}, while it would be a nice idea to use {-0.5, 0.0, 0.5, 1.0}, so that the computational resources are used more efficiently and the high-symmetry points are taken into account. Also, a simplest account of spatial symmetry to factorize equivalent points like (-0.5 0.0 0.0) and (0.5 0.0 0.0) would be very appreciated!
メンテ

Page: [1]