Sorry Azar, can't help with this one: I've never performed transiesta
calculations.
To me however, a 1x1x1 M&P grid looks odd, unless you have an
isolated molecule or a very large, bulky, cell. I guess yours is
a wire geometry, and the wire must be well relaxed before starting
the transport calculations, right?. Thus such a grid is "unwise"
to say the least.
Cheers,
Roberto
On 11/24/2015 02:57 PM, azar ostovan wrote:
Dear Roberto
Thank you so much for your kind reply.
Do the number of kgrid_Monkhorst_Pack block ( or periodicity in each
direction) could differ in structural relaxation, transiesta and tbtrans
calculations? Or it must be the same?
I mean can I for time saving used the below block in the optimization and
transiesta
%block kgrid_Monkhorst_Pack
1 0 0 0.0
0 1 0 0.0
0 0 1 0.0
%endblock kgrid_Monkhorst_Pack
but for tbtrans extend the structure in z-direction?
Warm regards
azar
On Tue, Nov 24, 2015 at 4:22 PM, RCP <[email protected]> wrote:
Hi,
On 11/24/2015 09:20 AM, azar ostovan wrote:
Dear, siesta users
The kgrid-cutoff parameter no need in the calculation with
kgrid_Monkhorst_Pack block?
Correct, kgrid-cutoff sets the M&P grid for you. Roughly you may
think of it as the density of k-points: the larger the value the
tighter the grid.
I couldn’t understand the sentences in the user manual
/“This parameter is not used if //kgrid Monkhorst Pack //is specifiedâ€
/
/Â /
Means they are mutually exclusive, but if you specify bnth, then M&P
takes precedence.
And if I define a value for this parameter during the calculation it may
result in wrong results?
Of course not ;)
Regards,
Roberto
