[CP2K-user] Is there a need to keep the Outer SCF loop?
Kavita
bulmave... at gmail.com
Fri Dec 21 09:36:15 UTC 2018
Dear Anton,
Thanks a lot, it was really helpful.
I have one more silly question which I would like to ask, I'm a bit
confused whether I have used PBC in my calculations or not, so could you
please look at my input file and tell me whether I have implemented PBC or
not. I saw two section's for PBC one in DFT POISSON and other in SUBSYS CELL
which caused the confusion.
Thanks and Regards,
Kavita
On Friday, December 21, 2018 at 12:02:11 AM UTC+5:30, Anton Kudelin wrote:
>
> Dear Kavita,
>
> There is no need in the outer loop if SCF converges quickly. Feel free to
> experiment with your input, try, for instance, direct mixing or Broyden's
> scheme, and you may achieve even faster convergence.
>
> Happy calculations,
> Anton K.
>
> четверг, 20 декабря 2018 г., 19:44:38 UTC+3 пользователь Kavita написал:
>>
>> Dear all,
>>
>> I'm using the following SCF setting's
>>
>> &SCF
>> MAX_SCF 30
>> EPS_SCF 1.0E-6
>> SCF_GUESS ATOMIC
>> &OT
>> MINIMIZER DIIS
>> PRECONDITIONER FULL_SINGLE_INVERSE
>> &END
>> &OUTER_SCF
>> EPS_SCF 1.0E-6
>> MAX_SCF 10
>> &END
>> &PRINT
>> &RESTART OFF
>> &END
>> &RESTART_HISTORY OFF
>> &END
>> &END
>>
>> My question is If my inner SCF loop is converging, then do I need to keep
>> that OUTER_SCF section in my input file? What if I erase this section and
>> run my calculation's without it? Does it's presence have any effect on the
>> results that I will get? My INNER SCF is converging very quickly though.
>>
>> Best Regards,
>> Kavita
>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20181221/30667dd6/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.inp
Type: chemical/x-gamess-input
Size: 5441 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20181221/30667dd6/attachment.inp>
More information about the CP2K-user
mailing list