[Wien] Finding the best Rmt*Kmax and kmesh

[Stefaan Cottenier]

> I have done a energy convergence test to find the best RmtKmax and
> k-mesh for a wide gap semiconductor by following DFT and the family of
> LAPW methods: a step by step introduction by S.Cottenier whose graph I
> am attaching with.
> Are 500 kpoints (whole of BZ) and Rmt*Kmax=8 are good enough?

Hard to say, for these two reasons:

1) The meaning of 'good enough' depends on what you need. So first 
determine how accurate you want your energy to be, and then inspect 
whether with 500 k-points you meet that requirements.

2) The effect of the basis set size on the accuracy of the total energy 
cannot be tested on absolute total energies. You need total energy 
differences (see fig. 7.3 [bottom] in the 2nd edition of the text you 
quote).

Stefaan