[CP2K-user] [CP2K:20985] ENERGY_FORCE calculation not converging on same ENERGY runs

[Hezhou Zhang]

Dear Jürg,

Thanks for the response and sorry for the ambiguity. The 4 energies are 
from four different frames. 
The problem is that, for the first frame on the right panel (with 
ENERGY_FORCE), the same condition gives a non-convergence and wrong energy, 
where only running ENERGY gives a converged result.
The EPS convergence threshold does not concern me for now.

Regards,
Hezhou
On Wednesday, December 18, 2024 at 3:37:08 AM UTC-6 Jürg Hutter wrote:

> Hi
>
> after rereading your mail, I might have misunderstood your question.
> What are the energies 1-4 ? Where are they coming from?
>
> regards
> JH
>
> ________________________________________
> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of Jürg 
> Hutter <hut... at chem.uzh.ch>
> Sent: Wednesday, December 18, 2024 10:34 AM
> To: cp... at googlegroups.com
> Subject: Re: [CP2K:20980] ENERGY_FORCE calculation not converging on same 
> ENERGY runs
>
> Hi
>
> you are restarting both calculations. From the initial energy it seems for 
> energy_force this was
> not optimal. However, this would not lead to different results if your 
> EPS_SCF would be smaller.
> With the value you chose 10^-5 SCF stops rather early and energy is only 
> correct to about 5 digits.
>
> regards
> JH
>
> ________________________________________
> From: cp... at googlegroups.com <cp... at googlegroups.com> on behalf of Hezhou 
> Zhang <lzha... at uchicago.edu>
> Sent: Tuesday, December 17, 2024 12:06 AM
> To: cp2k
> Subject: [CP2K:20978] ENERGY_FORCE calculation not converging on same 
> ENERGY runs
>
> Dear CP2K community,
>
> I'm benchmarking a UKS calculation on a small system. By only changing 
> GLOBAL RUNTYPE from ENERGY to ENERGY_FORCE and printing FORCE in 
> FORCE_EVAL, the calculations that successfully converged on ENERGY 
> calculations failed to converge in ENERGY_FORCE, and gave a drastically 
> wrong energy. This occurred in versions 7.1 and 2023.2 seemingly randomly, 
> but I haven't tried others. Any idea why this is happening? Thank you all!
>
> Regards,
> Hezhou
>
> [Screenshot 2024-12-16 170212.png]
>
> --
> You received this message because you are subscribed to the Google Groups 
> "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to cp2k+uns... at googlegroups.com<mailto:cp2k+uns... at googlegroups.com
> >.
> To view this discussion visit 
> https://groups.google.com/d/msgid/cp2k/2d457180-b004-4e01-8928-a74c0f52df31n%40googlegroups.com
> <
> https://groups.google.com/d/msgid/cp2k/2d457180-b004-4e01-8928-a74c0f52df31n%40googlegroups.com?utm_medium=email&utm_source=footer
> >.
>
> --
> You received this message because you are subscribed to the Google Groups 
> "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to cp2k+uns... at googlegroups.com.
> To view this discussion visit 
> https://groups.google.com/d/msgid/cp2k/ZR0P278MB0759485EB6766D4CC476F1AA9F052%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM
> .
>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/ae29b4cb-e2fe-4c84-a430-b4ea20c943fcn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20241218/a3693faa/attachment.htm>