[Wien] Questions about running WIEN2k with amorphous materials

[Lyudmila Dobysheva]

On 07.04.2015 10:47, YOOSOO YI wrote:
> The first question ... is it possible to calculate the electronic
> structures of amorphous silicate oxides using WIEN2k?
> ...if I make a sufficiently large super cell which is
> containing enough number of atoms that can ignore the boundary effect
> due to the periodic boundary condition. Is it right?

Yes, you can also make an averaging over few calculations of few 
different unit cells.

> The second question is about using SGROUP program for the amorphous
> materials. ... In this case, can I use this
> case.struct_sgroup generated from SGROUP? If not, what is the reason?

Usually, you can, but I'd prefer to not use in order that all unit cells 
for the above mentioned averaging be the same.

> The third question... the
> unexpected ERROR is continuously occurred on the LAPW1 in the first step
> of SCF calculation with a simple error message in the lapw1.error file:
> “Error in LAPW1”.

You are to find a diagnostic. If it was a parallel calculation - try a 
sequential, the diagnostic in lapw1.error will be more clear. If it was 
a sequential - maybe the calculation was not ended due to a hardware 
problem. Usually (sequential case) this file contains more information. 
Try to obtain it.

> ...I
> had thought that this unexpected ERROR on the LAPW1 might be caused from
> using the case.struct_sgroup file generated by SGROUP.

There can be many reasons. Try and find more diagnostic.

> This is the fourth question. After researching some references, I have
> tried to calculate a partial DOS for amorphous silicate oxides without
> using case.struct_sgroup file, and the problem was looked like solved.
> However, after the first step of SCF calculation is finished, another
> ERROR is occurred on LAPW2 with an error message related to the ghost
> band.

This can be a result of an earlier problem. Look very attentively 
through case.scf and warnings.

> To solve this problem, I had tried to modify the case.inm file to
> change the mixing scheme from MSR1 to MSEC1, but it did not help.

Because this way is wrong.

> Then I
> thought that I had to modify the energy parameters in case.in1c file,
> especially “EF=###” and “GLOBAL AE-PARAMETER with n OTHER CHOICES”.

I'd not change EF.
As a principle, working with E-PARAMETERs is a right way, but I have a 
feeling that something wrong happened earlier.

> Should I retain/use this “EF=0.1” value to restart the SCF
> calculation
 > Or, should I change all the energy parameters in case.in1c
 > file “0.30” to another value?

You can try, no recipe here without information.

> I want to understand the differences among these following cases.
> Set-A) Using -6.0 Ry as a core-valence separation energy
> Set-B) Using -11.0 Ry as a core-valence separation energy
> Set-C) Using -6.0 Ry as a core-valence separation energy & Modifying
> case.in1c file same as in case.in1c file in Set-B

Set-C looks for me wrong. case.inc, case.in1c and case.in2c change with 
-6 -11 in accordance. You should not combine them by hit and miss. Read 
in userguide what makes the choice of core-valence separation energy.

Best wishes
   Lyudmila Dobysheva
------------------------------------------------------------------
Phys.-Techn. Institute of Ural Br. of Russian Ac. of Sci.
426001 Izhevsk, ul.Kirova 132
RUSSIA
------------------------------------------------------------------
Tel.:7(3412) 432045(office), 722529(Fax)
E-mail: lyu at ftiudm.ru, lyuka17 at mail.ru (office)
         lyuka17 at gmail.com (home)
Skype:  lyuka17 (home), lyuka18 (office)
http://ftiudm.ru/content/view/25/103/lang,english/
------------------------------------------------------------------