[Wien] [Wien2k Users] QTL Warning for Fe_cubic_Space Group 225

[Ghosh SUDDHASATTWA]

Hi Kurt, 
Thank you very much for the help. I have one last (hope so!!!!) querie on
this aspect. 
When actually do we have to incorporate LO's in case.in1? 
1. When the *******WARNING****** message comes? You mention that "The reason
for changing the values is that I wanted to tune the linearization energies
as much  
as possible to the 'real' energies, in order to be sure to lose the
warning". 
2. To make the value of ENE more accurate? 
3. Does such incorporation of LO's really matter when we consider
lanthanides? 
4. As suggested by Stefaan (Guide) we have to run using in1new switch, then
we automatically get the linearization energies. Is the same thing true for
lanthanides
5. In cases of lanthanides, there is one LO for L=3 and one for L=2 in the
default case.in1. Can we, in similar way add one more LO for L=2 in such a
case. 
6. You also mention that QTL warning can be actually ignored, but I have
found that it does give an error in the DOS calculations unless our case.scf
file is free of any such warnings. 
The issue is when to make such changes? 
Any suggestions would be of great help. 

Thank you once again Kurt, 
Suddhasattwa Ghosh 

-----Original Message-----
From: wien-bounces at zeus.theochem.tuwien.ac.at
[mailto:wien-bounces at zeus.theochem.tuwien.ac.at] On Behalf Of Kurt
Lejaeghere
Sent: Tuesday, April 06, 2010 8:01 PM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] [Wien2k Users] QTL Warning for Fe_cubic_Space Group 225

> I modify my case.in1 as per your suggestions to
> WFFIL        (WFPRI, SUPWF)
>   8.50       10    4 (R-MT*K-MAX; MAX L IN WF, V-NMT
>  0.20    5  0      (GLOBAL E-PARAMETER WITH n OTHER CHOICES, global
> APW/LAPW)
>  0    0.20      0.000 CONT 1
>  0   -6.50      0.001 STOP 1
>  1    0.20      0.000 CONT 1
>  1   -3.97      0.001 STOP 1
>  2    0.20      0.005 CONT 1
>  2    1.00299  0.000 CONT 1
> K-VECTORS FROM UNIT: 4  -11.0       2.5    13
>
> Now I have a few queries
> 1. You have given linearization energies for L=1 and L=2 as 0.40 and 0.35?
> The default case.in1 shows the same energy? Could you please tell me why
you
> have changed it? That too to two different values

The linearization energies were chosen as the average between the up  
and down spin cases in :EPH001 and :EPL001. They are not the default  
values from case.in1. It is possible that in your case they have to be  
changed a bit, since you work with another space group, although I  
suspect that the values shouldn't differ much. The reason for changing  
the values is that I wanted to tune the linearization energies as much  
as possible to the 'real' energies, in order to be sure to lose the  
warning.

> 2. The example of Fe in Stefaan's guide does mention something similar
like
> this? How do we translate this to f-elements? Can we consider two L=3 just
> like two L=2 orbitals in our case?

This is indeed the same as in Stefaan's guide. I am not sure whether  
the addition of f-elements will constitute anything new for a geometry  
optimization, since Fe is only filled up to d-orbitals. The results of  
the case up to L=2 should be more than good enough.

Make sure that you change your number of exceptions from 5 to 6 if you  
add an LO and that all white spaces are OK.

Kind regards

Kurt

_______________________________________________
Wien mailing list
Wien at zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien