ABINIT Discussion Forums

Dear all,

I am struggling with calculating the ground state energy of Co in HCP structure.
I want to get the energy of formation of some Al-C-Co-Ni compounds (E21 structure) by calculating the differences of ground state energies. The results are really good for Al-FCC, C-DIAMANT, Ni-FCC and their compounds (regarding lattice parameters of elements and comparison of structures with references).
But the problem arises when I want to caclulate something including Co. The lattice parameters of Co-HCP are off by about 0.7 % (almost three times the error of Al-FCC lattice parameter) and the energies of formation get systematically worse the more Co atoms there are in the structure (e.g. error(Co4C)>error(Co3AlC)>error(CoAl3C)).
The deviations from the reference values aren't that high. But what's concerning me is that there seems to be a systematic error.

I attached input and output files of Co-HCP calculations with converged input values for ecut and ngkpt.
I am using JTH - v1.1 PAW atomic datasets.

I would be very thankful if you could take a look at the input variables and tell me if I am missing something or could do any better.

Thank you in advance!