[Wien] question about the Wien2k_14

[Gavin Abo]

> I have two questions about the new version.
>
> 1.How to avoid "EFG-MATRIX is a NULLMATRIX"?
> For cubic cell it is said running "runsp -s -so lapw1" before normal 
> scf can avoid the problem. But for v14 it seems does not work. What is 
> the problem? Is the message "EFG-MATRIX is a NULLMATRIX" important for 
> the calculation or in the scf iteration this problem can be corrected 
> automatically? For other lattice, such as hcp, how to avoid the problem?

Did you search the mailing list archive and read the previous reports 
about that error? For example:

http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg04409.html
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg06048.html
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg09677.html
http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg03446.html

> 2.For the eloss calculation of fcc Ni the results are different from 
> those calculated by v13. Has some important change been made to the 
> optic package? The eloss calculated by v13 is more resonable than the 
> one by v14 (in both calculation no error reported).

There are no changes between SRC_optic of v13 and v14; as you can see, 
no changes have been made as SRC_optic is not listed under v14 on the 
updates page of the Wien2k website [ 
http://www.wien2k.at/reg_user/updates/ ].  So the difference must be 
coming from something else, a difference in input file(s) or output(s) 
of the steps (packages) leading up to the use of the optic package.  I 
currently don't know the underlying details of the optic package as 
intimately as others might, so it is hard for me to say where the 
differences might be coming from.  You would likely have to compare the 
input and output at each step in the calculation until you are able to 
identify where the differences are coming from.