[Wien] linearization energies

[Stefaan Cottenier]

> Thanks for your reply. The linearization energies were set acccording to 
> the steps described in the notes writen by you using the DOS for the 
> system. But if i use the -in1new option during run_lapw i get different 
> linearization energies. The difference is that there are additional LOs 
> and the "E_trial" is also different. The total energy is lower by 
> 0.00013 in the case where in1new option was used.
> 
> My question is, which is more reliable result or does it depend on some 
> other criteria also?

in1new is probably a bit more accurate, but for many 'normal' materials 
there will be no qualitative differences between the defaults and the 
energies by in1new. At first sight, I don't see anything problematic in 
your case.in1.

Conclusion: your original problem (forces do not converge) is not 
related to in1new. Try the other possibilities (kmesh, basis set) or 
provide more details about how exactly your force problem shows up.

Stefaan