[adf-list]未能重新启动SCF

josécarlosconesa cegarra ICP.CSIC.ES的JCCONESA
星期五2019年9月27日15:40:25

Dear Pier,

Now I understand quite well. This means that there is no way of 
restarting a calculation using a previous wavefunction/density if the 
functional used is different. A pity, considering for example that one 
might wish to use the result of a GGA calculation to launch one using a 
hybrid functional.

Or is there some way? Can one force that the NAOs of the previous 
calculation are used for the second one?

Best,

José Carlos

El 27/09/2019 a las 14:38, Pier Philipsen escribió:
> Hello,
>
> Band uses numerical atomic basis functions, let us call these NABs. The radial part is a numerical table. The angular part is a spherical harmonic.
>
> The smallest (SZ) basis set consists exclusively of functions that are solutions for the free spherical atom (for the given xc functional). Let us call these functions NAOs.
>
> With bigger basis sets STO functions are added.
>
> When you change the functional you are changing the Numerical Atomic Orbitals. So you have the same number of basis functions, but the radial part of the NAOs has become different. (being checked by band)
>
> This then prevents you from restarting the SCF.
>
> One could think of an I-know-what-i-am-doing option, but this does not exist right now.
>
> Best regards, Pier
>
>
>
>> Op 27 sep. 2019, om 14:24 heeft  josécarlosconesa cegarra  < ICP.CSIC.ES的JCCONESA > het volgende geschreven:
>>
>> Hi,
>>
>> I tried to restart a BAND calculation (in order to use a different functional), but found in the restarted calculation the following error:
>>
>>   Failed to restart the SCF
>>   restart file name: rutil3.results/band.rkf
>>   Reason(s)  of failure:
>>      *Basis set (of the restart file) is not contained in the curent one
>>
>> Then the scf cicle began and reached convergence; but I guess that the results of the previous calculation were not used, and that the new calculation proceeded from scratch. Of course the required rkf file exists, and both the first and second jobs include the specifications
>>
>> in AMS input:
>>
>> Task SinglePoint
>>
>> in BAND input :
>>
>>    Basis
>>      Type TZ2P
>>      Core Medium
>>    End
>>
>> BeckeGrid and Kspace specifications are also the same. The second job also includes
>>
>> in AMS input:
>>
>> LoadSystem
>>    File rutil3.results/ams.rkf
>>    Section InputMolecule
>> End
>>
>> in BAND input:
>>
>>    Restart
>>      File rutil3.results/band.rkf
>>      SCF true
>>    End
>>
>> Which may be the reason for the failure in restarting the scf?
>>
>> Thank you in advance
>>
>> -- 
>> José C. Conesa
>> Research Professor
>> Instituto de Catálisis y Petroleoquímica, CSIC
>> Marie Curie 2, Campus de Cantoblanco
>> Tel. 915854766
>> Madrid, Spain
>> _______________________________________________
>> ADFlist mailing list
>> ADFlist at scm.com
>> //lists.tofoba.com/mailman/listinfo/adflist
> _______________________________________________
> ADFlist mailing list
> ADFlist at scm.com
> //lists.tofoba.com/mailman/listinfo/adflist

-- 
José C. Conesa
Research Professor
Instituto de Catálisis y Petroleoquímica, CSIC
Marie Curie 2, Campus de Cantoblanco
Tel. 915854766
Madrid, Spain


有关Adflist邮件列表的更多信息