Re: [Wien] Fwd: Lapw2_1.def ERROR

[Laurence Marks]

Looking at your structure, you have four oxygens (O 17 to O21) which are
not really bonded to anything. For instance, the closest atoms to O17 are
K4 at 2.87 Angers and O4 at 2.84 Angers. The BVS [1] for these atoms are
too small, 0.2-0.4. I strongly suspect that these are really water
molecules where the H could not be refined (probably no neutron diffraction
was done).

I suggest that you go back to the original CIF, check the source paper and
use one of the many viewers to ensure that your starting structure is
appropriate. In general I expect only small (~0.05 Angers) changes in
positions between DFT refined and good XRD refinements.

I also suggest that you update to 19.1. For your small cluster the use of
openmp threading would be a significant improvement.

[1] You can see the BVS via " x nn ; grep Bond case.outputnn". They are a
standard sanity check, for instance as a validation of CIF files in some
cases.

On Tue, Dec 3, 2019 at 1:00 AM Ashwani Kumar  wrote:

> Thank you Dr. Marks for the comments. Case.struct file is attached.
>
> 1) With a largish unit cell like this you only need 1 k-point for optimization
>
> Noted. i will repeat calc. with 1 k-point for optimization.
>
> 2) An RKMAX of 3.5 is absurdly small. You probably misinterpreted the CIF 
> file, which I suspect contained partial occupancies.
>
> Ok. i will repeat calculation with Rkmax=5. Please see case.struct, no 
> partial occupancies.
>
> 3) It appears that your cell contains Hf/K/C/O. You probably should have RMTs 
> for C and O near to 1.0, which means that you need an RKMAX of about 5
> for a decent calculation.
>
> Yes. RMT for C:1.15 & O:1.04 is defined.
>
> 4) Why are you using Wien2k 16.1? Is this a legal copy? If it is legal then 
> you should update. If you "acquired" 16.1 from somewhere then I doubt that 
> anyone should help you.
>
> i purchased wien2k in 2016. Installed at 6 core PC, works fine. For supercell 
> calc., installed at cluster but could not make it to run for 
> mpi_parallelization.
> Lot of time has gone in debuggin. I dont want to disturb w2k_16 installation 
> at cluster since we spent lot of time on this.
> But Mpi-parallelization still could not be done.
>
> thanks,
>
> A. Kumar
>
>
>
> ___
> Wien mailing list
> Wien@zeus.theochem.tuwien.ac.at
>
> https://urldefense.proofpoint.com/v2/url?u=http-3A__zeus.theochem.tuwien.ac.at_mailman_listinfo_wien=DwICAg=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0=D-G4QPvClPSD-tTOr4kCA2tEKfwpOVg-TomMZw_LSVY=DUvtxtPFGluXUAMnoF62rSDRaxfANK8PprSInkW5FyY=
> SEARCH the MAILING-LIST at:
> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.mail-2Darchive.com_wien-40zeus.theochem.tuwien.ac.at_index.html=DwICAg=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0=D-G4QPvClPSD-tTOr4kCA2tEKfwpOVg-TomMZw_LSVY=Fyq_P9yySsOwmmEEcShhXsi1tzb29PxafNcr29EiQXY=
>


-- 
Professor Laurence Marks
Department of Materials Science and Engineering
Northwestern University
www.numis.northwestern.edu
Corrosion in 4D: www.numis.northwestern.edu/MURI
Co-Editor, Acta Cryst A
"Research is to see what everybody else has seen, and to think what nobody
else has thought"
Albert Szent-Gyorgi
___
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

Re: [Wien] Fwd: Lapw2_1.def ERROR

[Laurence Marks]

You provide everything except the most relevant file -- case.struct

Comments:
1) With a largish unit cell like this you only need 1 k-point for
optimization
2) An RKMAX of 3.5 is absurdly small. You probably misinterpreted the CIF
file, which I suspect contained partial occupancies.
3) It appears that your cell contains Hf/K/C/O. You probably should have
RMTs for C and O near to 1.0, which means that you need an RKMAX of about 5
for a decent calculation.
4) Why are you using Wien2k 16.1? Is this a legal copy? If it is legal then
you should update. If you "acquired" 16.1 from somewhere then I doubt that
anyone should help you.

On Mon, Dec 2, 2019 at 1:34 PM Ashwani Kumar  wrote:

> Lapw2 error file generated with message "Error in Lapw2" only. Other
> *.error file (~lapw1_x.error) created empty. Case.dayfile is attached which
> shows last job id issued for 16 cores. You are right, Dr. Lyudmila.
> Warnings are present in case.dayfile related to inaccurate EF (appeared 16
> times in 22 iterations). 23rd iteration never completed. At cluster, it was
> not running when i checked today, but w2k interface status showed running.
> Neither JOB nor error output file created (surprise for me). Case.scf0,
> case.scf1 files are shared through Google Drive.
>
>
> https://drive.google.com/open?id=0B97owdomubQ7WlFRT3F0TnZfTWhyY2VQU0ZQS0VaVnlPZlc0
> 
>
> https://drive.google.com/open?id=0B97owdomubQ7N1VOMlJvbkMyVU5ZZzlvVkZHeXVDS3VsU1Vj
> 
>
> https://drive.google.com/open?id=0B97owdomubQ7ZFVsTGp2VmE0U0YxdVhYZHBEWFlxbVV6cVJF
> 
>
> https://drive.google.com/open?id=0B97owdomubQ7andEZGNuWHREMXFGN0MyS1hsc21DbDdwcTJV
> 
>
> https://drive.google.com/open?id=0B97owdomubQ7QWplS3IzeEx6QjlmMFo2bzY1QXFHSXhwUW44
> 
>
> thanks,
> A. kumar
>
> On Mon, Dec 2, 2019 at 2:43 PM Ashwani Kumar 
> wrote:
>
>>  Hi,
>>i want to calculate EFG for a oxalate compound. During structure
>> relaxation calculation (min_lapw), lapw2 failed after 32 iterations. Error
>> file is attached with the mail.
>> Calculation steps:
>> 1. Charge and energy convergence -successfully terminated
>> 2. Force convergence (run_lapw -p -fc 5.0) -successfully terminated
>> 3. min_lapw (PORT, tolf=1.0) , calculation crashed after 32 iterations.
>> (using WIEN2K 16.1 version)
>> parameters: ecut -5.5, rkmax: 3.5 (kept 3.5 as calculation was too
>> computationally expensive at more than 4 rkmax and despite trying very hard
>> i could not make wien2k run parallel using mpi, only k-point
>> parallelization is used at single node, 12 cores),  G-max:12, gmin: 6.7,
>> :klist =100.
>>
>> Most of the threads on wien2k forum relates similar kind of error to
>> inaccurate structure information (for which i m not convinced since cif
>> file from CSSD used). Help me in understanding what went wrong.
>>
>
>
>>
>> thanks,
>> A. kumar
>>
> ___
> Wien mailing list
> Wien@zeus.theochem.tuwien.ac.at
>
> https://urldefense.proofpoint.com/v2/url?u=http-3A__zeus.theochem.tuwien.ac.at_mailman_listinfo_wien=DwICAg=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0=uSOIrm4du1hr0XrC0kFthl1ZlUm8GpSakrgkLYDzIgA=-VUPOm7Rc38DGP1K9-CFKIQroug2t-RniB78keigw7s=
> SEARCH the MAILING-LIST at:
>