[Wien] (no subject)

[Stefaan Cottenier]

> When I calculate a tetragonal distorted cell, which is deviated from a
> cubic
> structure, I have to setup a structure with a slight distortion. Then, I
> initialize with a lower symmetry (I choose only a B lattice).
> But after init_lapw the distortion isn't slightly, because it changed for
> example from 5.72A for a (5.73A for c) to 7.6A and 10A after the program
> determined the tetragonal space group.
> My questions:
> Does the program find the right cell or have I to know which tetragonal
> space group is correct?
> Can I work with the atomic positions of the cubic cell during the
> initialization to find the tetragonal cell? Then the initialization
> program finds the new space group with new atomic positions and changes
> lattice parameters.
> The second way would be: First, I'm choosing the subgroup and start from
> the beginning with the deviated tetragonal cell, atomic positions and cell
> parameters?
> But there is still the problem that Wien2K is changing my preferred
> lattice parameters?

The case.struct generated by nn and sgroup should always give a lattice that
is *identical* to your original case.struct. Perhaps the lattice constants
look very different, but this is only because it has chosen a different
space group. The absolute positions of your atoms will be exactly the same
(check this by displaying both lattices with xcrysden). The case.struct from
sgroup has higher symmetry than your original choice, and is the preferred
one to work with.

Stefaan