[Pw_forum] Strain calculation through quantum espresso (Giovanni Cantele)

[Eduardo Ariel Menendez Proupin]

Hi DS

Strain is a geometrical property. If you want a fast calculation, just use
the Bilbao Crystallographic Server.
http://www.cryst.ehu.es/cryst/strain.html
there you may input two sets of lattice parameters and get the strain
relating both lattices.

Take a look at Fast et al, PRB 51, 17431 (1995). In equations (1) and (2)
are explained the relation between lattice vectors and strain. In  Kittel's
textbook it is explained too, but I am not sure if in all the editions. In
the article of Fast it explains what you can do with Q-E and similar codes.

In Q-E you input the lattice parameters or the lattice vectors, and all the
other stuff, and get the stress tensor as explained in previous posts.

Maybe what you need is to find the strain of a crystal under an specified
stress tensor. As far as I know, with Q-E you cand do it only in the case of
hydrostatic stress (when the stress tensor is the pressure times the
identity matrix), using calculation='vc-relax'  and setting the variable
press. E.g.
 &control
    calculation = 'vc-relax'
..........
&CELL
        cell_dynamics='damp-pr',
        press = 35.0, (the value of pressure in kbars)
        cell_dofree='all',


For a non-hydrostatic stress, you may need to find the elastic constants (as
in Fasts's paper) first and then solve the equations of elasticity. If the
stress is big, you may need an iterative process.

If you need the strain under a fully specified stress tensor, i.e., the six
independent components of the tensor, you may device a minimization
algorithm implemented in a shell or python script, that run pw.x  to get the
energies and stress.
An alternative is  to find the elastic constants (as in Fasts's paper) first
and then solve the equations of elasticity. If the stress is big, you may
need an iterative process to find the elastic constants under a stress that
is close to your derired stress.

In simple cases, such as uniaxial stress and orthorombic lattices, you do
cell relaxations with constraints (cell_dofree='xxxxx' ) keeping one lattice
vector fixed, and by trial and error you can obtain the lattice that produce
the desired component of the stress.




-- 
Eduardo Menendez
Departamento de Fisica
Facultad de Ciencias
Universidad de Chile
Phone: (56)(2)9787439
URL: http://fisica.ciencias.uchile.cl/~emenendez
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090708/4abbc09b/attachment.htm