[CP2K-user] [CP2K:19349] Re: ETRS and CN propagators lead to completely different Ehrenfest dynamics

[Natalia K]

UPD on this issue: apparently, the Cranc-Nicolson is the less stable one. I 
did some additional tests with a very small time-step (0.0005 fs) and the 
CN propagator converged to the ETRS result, the same one that ETRS gives 
even with dt = 0.006 fs. So, the answer is, one should use the ETRS 
propagator to be able to use larger time step.

Natalia

On Monday, October 9, 2023 at 2:06:33 PM UTC+2 Natalia K wrote:

> Dear CP2K team,
>
> I performed simple tests for H2 with cp2k-9.1 using an input file from 
> *cp2k/tests/QS/regtest-rtp-2* and modifying the external field (the input 
> file is attached). I then performed 4 tests, using ETRS or CN propagators 
> and two different time steps, dt=0.25 a.u. and 0.125 a.u. With the ETRS 
> propagator, the energies as well as the H atom dynamics are the same for 
> both dt values. However, the results with the Crank-Nicolson propagator 
> (CN) are not only different for two time steps, but also are completely 
> different from when using ETRS. I attach the results here. I tried refining 
> some other parameters, but the two propagators always lead to different EMD.
>
> My question: when performing Ehrenfest dynamics with applied external 
> field, which propagator should I trust? Why they give such different 
> results?
>
> Best regards,
>
> Natalia
>
>

-- 
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 on the web visit https://groups.google.com/d/msgid/cp2k/45a76931-2f65-4883-81b0-4951c93e9028n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20231011/a42103fa/attachment.htm>