Hello Aline,
Usually I don't relax the lattice vectors, I do relax the atomic positions. If
the pressure is too high as you say, perhaps it is a good idea to scan
different lattice constants i.e. loop your calculatiIation so that the lattice
vector{s} are expended or contracted by some small amount, < .1 Ang usually.
Which lattice vector{s} you relax may depend on whether the stress is uniaxial,
anisotropic or isotropic. My experience is that volume relaxations, that is
allowing all lattice vector to relax at once, is not only costly but also
sometimes results in weird structures.
I hope it helps.
Diana M. Otálvaro
Doctoral Candidate
Computational Material Science
MESA+ Institute of Nanotechnology
University of Twente.
Enschede, Nederland
________________________________________
From: [email protected] [[email protected]] on behalf of Aline M.
Pascon de Marque [[email protected]]
Sent: Wednesday, July 24, 2013 8:36 PM
To: [email protected]
Subject: [SIESTA-L] [Siesta-L] delta-TaN (111) surface lattice vectors
I'm building a TaN (111) surface and I'd like to know if after to build this
surface I need to relax the new lattice vectors of the structure, because the
static pressure is very high.
The lattice vectors of delta-TaN, that I'm using to build the (111) surface
were relaxed previously.
Tks
--
Aline.
