[Wien] Lapw1 error

[Peter Blaha]

> 1. How can one know "a priori" when it is possible to drop an automatically
> generated LO without adversly affecting the accuracy of the calculation?

In normal cases this does not occur, in difficult cases (RMT small) you 
can't. But you can always analyse a posteriory the help files and 
find out what causes the problem.

> 2. It seems the 2s/3s problem here was analogous to the 3p/4p example
> discussed in the UG(12.1). In the FAQ it is mentioned that "you need to remove
> a LO or increase the energy seperation between LO and L/APW energies".
> Removing the LO works well; however if one were to increase the energy
> separation between the LO and L/APW energies instead, I assume it is the
> energy of the 2s LO that needs to be changed (increased relative to the
> 'Global E parameter' for that atom?)
> 
> Does a general "prescription" exist on how to choose the expansion energy of
> LO's?

The "general" prescription is built into the code. Only some exceptions 
must be handled manually.

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------