[Wien] Problem with geometry minimzation

[Saeed Bahramy]

Dear Laurence,

Thanks  a lot. That was indeed the reason why I couldn't optimize the  
atomic positions. Reducing (!)  R0 to 0.00001, I could successfully  
minimize all the forces. Now they're all well below 1.0 mRy/a.u..

By the way, you said that my initial R0 was too small, but I had to  
further reduce it in order to get min_lapw working!  Am I picking up  
wrong R0 for Te!? (UG says it should be between 0.0005 to 0.00005)

Thanks again,
Saeed Bahramy


On Nov 3, 2010, at 3:08 AM, Laurence Marks wrote:

> Te, R0 is too small.
>
> 2010/11/2 Saeed Bahramy <bahramy at riken.jp>:
>>
>> Ok. From this you can clearly see that E-tot and FGL does not fit  
>> together.
>>
>> Smaller forces give less negative E-tot. (At least if forces on  
>> other atoms
>> do not destroy this simple logic) and since PORT minimizes E-tot, it
>> finishes,
>> although it has non-zero forces.
>>
>> Well, for those steps the total forces on other atoms are large and  
>> that's
>> the reason why E-tot is less negative. For example in the second  
>> step, the
>> total forces on all atoms are as follows:
>>        :FGL001:   1.ATOM                 0.000000000     0.000000000
>> -16.042000000 total forces
>>        :FGL002:   2.ATOM                 0.000000000     0.000000000
>>  41.642000000 total forces
>>        :FGL003:   3.ATOM                 0.000000000     0.000000000
>>  10.710000000 total forces
>> and accordingly in this steps :ENE is the highest.
>>
>> I'm quite sure that my struct file is correct. The initail  
>> coordinates (as
>> well as lattice parameters) have been taken from experiment. Before
>> restarting the min_lapw calculation,
>>
>> This does not say anything. Is R0 ok ? RMTs set properly ?
>>
>> R0=0.0001 and RMT=2.5 (for all three atoms). Once again, there is no
>> touching between the MT's
>>
>> I do remove histories (*broyd* and *tmpM files). Below, you can  
>> find the
>> summary of (one of) my min_lapw calculation:
>>
>> Is it always like this, or other runs are different ?
>>
>> I always remove the histories before restarting min_lapw with a new
>> case.inM
>>
>>
>> If yes, the problem must be somewhere else and again, without testing
>> myself,
>> I cannot give more than the "standard hints", although the problem  
>> may stem
>> from something completely different ( Are you using the latest WIEN2k
>> version
>> and have followed all bug-fixes discussed on the mailing list ?)
>>
>> My current version of WIEN2K is 10.
>> Anyway, using this version of WIEN2K, I have been able to do geometry
>> optimization for other systems without any problem.
>>
>> This almost always means that the user (you) has set the IDFT problem
>>
>> up incorrectly -- GIGO.
>>
>> Have you read the optimization notes at
>>
>> http://www.wien2k.at/reg_user/textbooks/ ? Do you have almost  
>> touching
>>
>> spheres, bad RKMAX, k-points etc?
>>
>> --
>>
>>                                      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/
>> --------------------------------------------------------------------------
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>>
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>>
>
>
>
> -- 
> Laurence Marks
> Department of Materials Science and Engineering
> MSE Rm 2036 Cook Hall
> 2220 N Campus Drive
> Northwestern University
> Evanston, IL 60208, USA
> Tel: (847) 491-3996 Fax: (847) 491-7820
> email: L-marks at northwestern dot edu
> Web: www.numis.northwestern.edu
> Chair, Commission on Electron Crystallography of IUCR
> www.numis.northwestern.edu/
> Electron crystallography is the branch of science that uses electron
> scattering and imaging to study the structure of matter.
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien