[Wien] problem with output file of lstart (fwd)

[Peter Blaha]

> I was trying to send mail with both structure file and outputst file from
> lstart, but the letter was too large, so I left only structure file.
>
> I want to ask if the discussion between you and
> Valerio Bellini  can be related to my problem.

No, I don't think so. Then you should always have these problems.


> Dear Wien Users,
> I still can not understand the origin of the problem with file.outputst of
> lstart which  gives for Mn ion
>
> TOTAL CORE-CHARGE:                  12.0006130153816
>  TOTAL CORE-CHARGE INSIDE SPHERE:    31.9476572588406
>
> I have  checked the structure and I think it is correct.
> I am attaching to this mail both my structure file and outputst file and I
> will be very greatfull if somebody takes a look at it.

I tried your struct file:
First I noticed:

  WARNING !!!! For good atomic total energies you should probably change the rad
 ial mesh (reduce NRAD or increase R0), or increase PARAMETERS NPT and NPT00
         961  RADIAL MESH POINTS REACH UP TO   7.48321593033320       A.U.

Such WARNINGS can usually be ignored, if the number given is large (say gt.
12. bohr), but yours is only 7.48...

Thus I checked the struct file:
...
ATOM  -1: X=0.00000000 Y=0.00000000 Z=0.00000000
          MULT= 4          ISPLIT= 8
      -1: X=0.50000000 Y=0.00000000 Z=0.50000000
      -1: X=0.50000000 Y=0.50000000 Z=0.50000000
      -1: X=0.00000000 Y=0.50000000 Z=0.00000000
Mn         NPT=  781  R0=0.00500000 RMT=    1.9000   Z: 25.0
                         ^^^^^^^
You have a much to large R0 value!!
structgen of w2web would produce a much smaller value (more like 0.00005 )
and while it is not too sensitive, one cannot choose arbitrary values!
All calculations with such large R0 values are at least questionable!!!!!

Nevertheless, this does not solve your problem. As I would have expected, I
cannot reproduce your error, but get:

 TOTAL CORE-CHARGE:                  12.0006130153816
 TOTAL CORE-CHARGE INSIDE SPHERE:    11.9836622004387
 WARNING:   1.69508149428523D-002  CORE electrons leak out of MT-sphere !!!!

telling me that I should reduce E-seper from -6. to -7. Ry.

What is your case.inst ? Was it generated by structgen or instgen_lapw; or
by "hand" ? Maybe this causes the problem?


                                      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/
--------------------------------------------------------------------------