[Wien] R0 to big

[Peter Blaha]

Dear "beginners" of WIEN2k,

Not even in my worst dreams I imagined that somebody would specify in the structgenerator
not only the "name of atom", but also ZZ (I assumed, nobody would look this up for a
heavier atom, where one usually does not memorize the Z of a certain atom.)
It is NOT NECESSARY !!!

Thus I never understood how people could "create" these R0 errors (showing up in lstart)
unless they copied an old struct file and changed it by hand.

Anyway, once the problem can be reproduced, the solution is fairly simple and was more or less
pointed out in some previous emails (Thanks !!!).

So there is a clear message:

a) Never specify ZZ, only the correct !!! name of the atom.
    (PS: You can still modify ZZ later on for "virtual crystal approx.", but
     do it only AFTER a first initialization  (instgen !!)

The attached SRC_w2web/libs/struct.pl file will fix this problem and set R0 properly in any case.
However, this implies another restriction:

b) The w2web structeditor will NOT tolerate nonstandard R0 (if you change R0 by hand,
    you MUST NOT open the structeditor later on. But this is not recommended anyway !!!)

I hope, this settles the "R0 to big" problem once and forever.


Am 02.08.2011 10:53, schrieb Martin Kroeker:
> Dear Maurits
>
> You could try moving the four lines that set "ro" (ro=0.00005 and the
> three "if" statements for various z ranges) from the end of the
> Element() function in struct.pl to immediately after line 309 and leave/readd
> the (if $zz[$i]==0) of the original version.
> That way, Z should be set automatically only if it is empty, and R0
> should always be set according to Z, no matter where the Z value came from.
>
> Another problem would be how to handle user-defined R0 - I am beginning
> to suspect that the original intention of the code was to have the user
> either leave Z and R0 blank to get the default values, or enter correct
> values for both, and it is only the combination of user-defined Z and
> empty R0 that is not handled properly. On the other hand I have not even
> tried whether a user-defined R0 would survive structgen with the
> original code (with the change proposed above, it would obviously be
> replaced).
>
> Martin

-- 

                                       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/
--------------------------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: struct.pl
Type: application/x-perl
Size: 71701 bytes
Desc: not available
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20110803/82e74bec/attachment.pl>