[Wien] Large forces at the end of minimization!

[Stefaan Cottenier]

A guess of what might have happened: you have displaced the atoms 
manually and search for the minimum of the total energy, and then you 
look at the resulting forces? If so, did you change TOT to FOR in the 
top line of case.in2? Only when FOR is set (needed in the last iteration 
only), the full forces are calculated. The values obtained with TOT are 
meaningless, and can indeed be largely non-zero even when your structure 
is in the correct minimum.

Stefaan

>If by minimization you mean at the end of mini, do a "grep -e :M
>case.outputM" and attach the results (change "case" of course).
>
>If you mean the forces are high at the end of run_xyz, then you have
>to do a minimization! Don't expect the bond length to be close to
>correct (experiment).
>
>On 12/13/06, Hojjat Gholizadeh <gholizadeh.hojjat at gmail.com> wrote:
>  
>
>>Dear all
>>
>> We wanted to achieve CO bond length with PBE, a GGA functional. Wien2k
>>gives correct bond (bond length is about 1.136 A which is comparable with
>>result in literature) after minimization but forces are very large(~15
>>mRy/au).
>>
>> We have checked out many things including parameters such as RMT, lattice
>>parameter for unit cell, functional type (GGA and meta-GAA), Potential
>>cutoff and wave function cutoff (RKmax), SO effect,  and also we've used
>>different Wien2k version(2005-2006), precompiled version and source code.
>>We've used Fedora Core 4 as OS and our machine was intel em64t.
>>
>> we would like to know, is our calculation safe?
>>
>> Best regards
>>
>> Hojjat Gholizadeh
>>
>>
>>_______________________________________________
>>Wien mailing list
>>Wien at zeus.theochem.tuwien.ac.at
>>http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>>
>>
>>    
>>
>
>
>  
>


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm