[Wien] [test_results] kgen and kgen -fbz

[Dr. K. C. Bhamu]

Dear Wien2k Users,

I have tested a case for it and below is my conclusion:

with "x kgen"  and 40000 k-points I can calculate thermoelectric properties
without any problem while with "x kgen -fbz" and 40000 k-points I get the
error:

forrtl: severe (41): insufficient virtual memory

The case.energy file created from later is large and may be cause of error.

Any comments will be helpful.


Sincerely
Bhamu


On Mon, Jun 12, 2017 at 2:00 PM, Dr. K. C. Bhamu <kcbhamu85 at gmail.com>
wrote:

> Dear Wien2k Users
>
>
> I have a doubt for x kgen  and need some clarification.
> For calculating DOSS and optical properties we need a fine mesh and many
> numbers of k points for thermoelectric properties.
>
> What we observed that: If we use only x kgen with 2000 k points, we get 72
> kpoints with 12 12 12 mesh siz (in our case otherwise it depends on the
> case).
>
> If I use x kgen -  fbz it gives 1728  total number of k-points with same
> mesh size (12 12 12).
>
> For thermoelectric properties, if I use "x kgen" for 40000 k points, it
> gives me 1059k-point with a 34 34 34 mesh size while with kgen -fbz I am
> getting 39304 k-point with same mesh size 34 34 34.
>
>
>
> As far as I know  "x kgen" represents the IBZ which gives image mapping of
> full BZ with 3D repetition while kgen with -fbz stands for whole BZ.
> .
> My query is:
>
> Which type of BZ I should use for thermoelectric, optical and DOSS. If I
> use 40000 kmesh with -fbz it takes much more time.
>
> Would there be any difference for simply "x kgen" or "x kgen -fbz"
> calculations?
>
>
>
> Bhamu
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20170613/8a6f3f2c/attachment.html>