[Wien] total energy question

[Stefaan Cottenier]

> Is the total energy, as listed in the .scf ouput file, the energy per
> number of atoms as specified in the .struc file?  For instance,
> in the example TiC we have 2 atoms in the .struc file at 4a and
> 4b positions (space group 225). Now, we have 3 additional Bravais
> translations since this is the FCC lattice (0,1/2/,1/2; 1/2,0,1/2;
> 1/,1/2,0).  Should I then understand that my energy in the .scf
> file is per 2 atoms or rather per 8 atoms (as in the unit cell)?
> What would happen if there were more than one (4a and 4b in 225
> have only one) equivalent positions for a given Wyckoff site?
> Would then the total energy include all the equivalent atoms as
> well?

The total energy in case.scf is the total energy for a *primitive* unit
cell, which contains always exactly as many atoms as there are X-Y-Z
coordinate lines in case.struct. The primitive unit cell of TiC contains 2
atoms (there are 2 coordinate lines in case.struct), and the corresponding
conventional unit cell contains 8 atoms. But the energy is for the 2 atoms
of the primitive cell. Positions with multiplicity larger than 1 mean that
you have more coordinate lines, hence, also the energy is for more atoms (2
positions each with multiplicity 2 is 4 atoms in the primitive cell).

Stefaan