[Wien] tetrahedron BZ integration

[Fred Nastos]

On May 10, 2004 01:46 pm, Jorissen Kevin wrote:

> that's what I suspected.  I want to use the linearization scheme of P.
> Bloechl like it's used in some places in the wien code.  So I guess the
> weights I'm looking for are the ones given by kgen (== multiplicity  of the
> irreducible point) times some factor which is, I think, given in his paper.
> I was just hoping someone has done this before and knows exactly which
> factor I need.

I guess the exact factor depends on what exactly you want to
do.  You can use the standard linearization scheme, in which
case, there are many papers that present the weights.  I believe,
the actual code where it is computed is in SRC_lapw2/fermi.f.
Look for SUBROUTINE WEIGHT(VOL,E,EF,WGHT) at around
line 1000.  Bloechl's scheme includes a correction for the
quadratic errors of the linear scheme.  This is at around line
1114.  I do not exactly understand this correction, although I
can see the "spirit" of it.  Without this correction, I think
everything reduces to the standard LATM method.

Cheers,  Fred