[Wien] supercell problem

[Stefaan Cottenier]

> I have a case with tetragonal unit cell in rutile structure (a=b, c)
> and then I built up a supercell 1x 1x 2.
> After I checked the new positions, the number of inequivalents atoms and
> also broken the symmetry
> replacing one atom by impurity.
> The problem is when I start the calculation the nn program complains
> about multiplicity and creates
> a new struct file but if I use the new struct file the sgroup program
> change the shape of the structure
> to triclinic cell. Moreover,  if I keep the old struct file the sgroup
> change the lattice constants and creates
> an orthorhombic base centered cell.
> Does anyone know how I can fix it?

Most likely you don't have to fix anything. Go through nn and sgroup
repeatedly and accept all their changes until they do not protest any more.
Then look at the resulting cell by xcrysden (convential cell). Probably you
will see the cell that you wanted. Only init_lapw has reduced it to be as
small as possible, and has transformed it into a format that is consistent
with the WIEN conventions.

Stefaan