[Pw_forum] Regarding controlled dynamics

[Matic Poberznik]

Dear Vipul,

> Hi all,
> I am planing to work on interfaces. In doing so, I want to keep few layers
> of a structure fixed while few to relax. In short this is what I am
> planning :


If I understood correctly and you're doing a relaxation the simplest way
to fix atomic positions is to add 0 0 0 at the end of the specifed
coordinate in the ATOMIC_POSITIONS card, (this will fix each component
in x y y directions respectively) if you need to free up one direction
replace the 0 by 1 i.e.

ATOMIC_POSITIONS
A x y z 0 0 0

or to allow movement in one direction (x)

A x y z 1 0 0

> 10A vacuum
> 2layers fixed
> 4layers to relax
> 2layers fixed
>

It's difficult to say without further information but make sure that the
largest cell parameter (typically where you put the vacuum, is oriented
in the z direction (due to parallelization)). I don't think esm and
assume_isolated is needed in this case as long as your vacuum is large
enough so that there is no interaction between the periodic slabs.

Hope this helps and best regards,

Matic
-- 
Matic Poberznik
J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia