[Pw_forum] Slab optimization with -24kbar stress!!!

[Nicola Marzari]

On 05/12/2014 07:01, David Foster wrote:
> Dear Paolo and Nicola
>
> Thank you for replies. However:
>
> 1- Nicola, As you know, for a slab with vacuum, we don't use vc-relax to reduce the stress. It is not a good idea.
> Cell parameters are only optimized for bulk systems.


Hi David,


1) you need to relax the slab in the plane parallel to the surface -
how you do it (by hand, constrained vc-relax, anything else) is a different
matter

2) you need to understand that freezing some atoms has a consequence on
the stress - if you force two planes of atoms to be very close, you'll
get a stress both in plane and out of plane

3) experiment with simple systems, trying to understand better the
relations between cell relaxations, atoms relaxations, stress, and
the presence or not of vacuum.

Or, at a minimum, relax all your atoms in your unit cell, and change
the in-plane dimensions to get zero stress. Then, see what is the effect
of forcing different planes of atoms move in the direction perpendicular
to the plane.


				nicola



----------------------------------------------------------------------
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL