[Wien] problem in analysis function & problem with "isolated band" in super cell

[Dong Su]

Hi, Stefaan:
Thank you for you suggestions.
1 During running. There were no error files in the super cell case.I do see cycle in .dayfile. The last cycle of .dayfile shows:

>   lapw2 -c  -p 	(21:40:56) running LAPW2 in parallel mode
      saguaro-10-7 219.805u 27.321s 4:11.31 98.3% 0+0k 0+0io 0pf+0w
      saguaro-10-7 218.701u 26.917s 4:08.51 98.8% 0+0k 0+0io 0pf+0w
   Summary of lapw2para:
   saguaro-10-7	 user=438.506	 wallclock=499.82
450.124u 72.836s 4:45.96 182.8%	0+0k 0+0io 0pf+0w
>   lcore	(21:45:42) 3.200u 3.032s 0:06.93 89.8%	0+0k 0+0io 0pf+0w
>   mixer	(21:45:55) 21.825u 16.753s 0:51.61 74.7%	0+0k 0+0io 0pf+0w
:ENERGY convergence:  1 0.0001 .0000475000000000
:CHARGE convergence:  0 0.0000 .0056402
ec cc and fc_conv 1 1 1

>   stop

Does it look good right? I will try the super cell with 80 atoms according to your suggestion to see what happened.


2 
I have calculated this ABO3 case of single unit cell. The results look like good. I have .scf files in single unit cell case as follows:

> case.scf0;
> case.scf1;
> case.scf1-1; case.scf1-2;....case.scf1-36
> case.scf2; case.scf2-2;....case.scf2-36
> case.scf2p
> case.scfc
> case.scfm
I was using 32 nodes in calculation.
I still don't see the .scf file. Also I see the .dayfile is normal converged after 10 cycles.


Thanks


Dong





The following are from the last e-mail

_______________________________________________________________

> But when calculating the super cell:
> 1 I got message "> energy in SCF NOT CONVERGED "in .dayfile. 
>  
That only means that the maximum number of iterations you specified was
reached, without meeting your energy convergence criterium. If the
calculations was converging, just continue.
> 2 I reduced the K points from 20 to 10. the SCF converged but the DOS of atoms shows isolated bands instead of a continuous spectrum. It is just like the XRD pattern of a single crystal.
>  
Could you successfully obtain a DOS for a smaller test case, and did you
repeat exactly this procedure? If so, then perhaps you have not enough
k-points for a good DOS.

Stefaan
_________________________________________________________________
If you do not have a case.scf file, you did not complete even a single 
iteration. There must be *.error files present. Familiarize yourself 
with how to use this code by taking a related simpler case (bulk ABO3 
?) before going to a supercell with 160 atoms.

Stefaan

Quoting Dong Su <Dong.Su at asu.edu>:

> Dear Wien2k users:
>
> I got a problem in using analysis function in Wien2k.(V 7.3)
> I am calculating a ABO3 super cell in parallel mode with 32 nodes.  
> The unit cell has 160 atoms.
> I tried to use analysis to see the "ENE FER DIS NEC-new NEC-old  
> MMTOT" from case.scf file but have not got any result.
>
> In the SCF files,I only found:
> case.scf0;
> case.scf1;
> case.scf1-1;
> case.scf1-2;
> case.scf2
> case.scf2p
> case.scfc
> case.scfm
> In each scf file, I do not find any :ENE or other phrase.
>
> Does it look normal?
> Could you please tell me the possible problem there?
_____________________________________________________________________









Dong SU,Ph.D
Dept. Of Physics, Arizona State University
P.O. Box 871504
Tempe, Arizona 85287-1504
Phone: 480-965-6327
Fax: 480-965-7954
E-mail: dong.su at asu.edu