Meaning of SHADOW_GAMMA

33 views
Skip to first unread message

Amit Gupta

unread,
Jan 2, 2023, 5:41:04 PM1/2/23
to cp2k
Following the conversation of this thread: https://groups.google.com/g/cp2k/c/fmZX3_2kivc

I did a parameter sweep of my system using FIST dynamics, and determined that
GAMMA = 0.35
NOISY_GAMMA=1e-5 

yields best thermostat stability in my system. The paper mentioned in above conversation however uses slightly different terminology of Gamma_L (Langevin friction) and Gamma_D(intrinsic friction). And their values are rather low (1e-8 for Gamma_D and 0 for Gamma_L ). I would like to understand

1. How are GAMMA, NOISY_GAMMA, and SHADOW_GAMMA are related to Gamma_L and Gamma_D?
2. Does the value I got (0.3 and 1e-5) pass the initial smell test? Or do they look non-physical (I mostly use Nose thermostat but needs to use Langevin for this study for replicating a paper)? 

Thank you

Amit Gupta

unread,
Jan 2, 2023, 6:02:18 PM1/2/23
to cp2k
Added note:
I just saw another conversation which is similar: https://groups.google.com/g/cp2k/c/EMxxtX0sFK4
I missed it initially as it is too new, and does not come up on google.

It is stated that correct Gamma would result in stable energy partition. For determining the parameters, I was simulating the metal system at 1000K as a simple LJ liquid.
I observed in output file that K.E. is stable, however P.E. was drifting in the time scale that I simulated it for (512 atoms, 2 ns), which might be due to melting lattice. Is that ok?

Thomas Kühne

unread,
Jan 5, 2023, 4:11:37 PM1/5/23
to cp...@googlegroups.com
Dear Amit, 

Gamma_L and Gamma_D of the paper refers to GAMMA and NOISY_GAMMA within the CP2K code. 
SHADOW_GAMMA is the opposite of NOISY_GAMMA and induces a Langevin friction without adding 
the corresponding noise according to the fluctuation-dissipation theorem. This can be used, (i) to determine 
NOISY_GAMMA by reproducing the dissipation as a function of SHADOW_GAMMA, or (ii) to use the 
Langevin equation at the presence of intrinsic noise in the forces e.g. due to the usage of low-precision 
arithmetic. 

Best, 
Thomas Kühne

--
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...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/42ba9996-c1ed-4876-9b70-74049c47dc70n%40googlegroups.com.



==============================
Thomas D. Kühne
Dynamics of Condensed Matter
Chair of Theoretical Chemistry
University of Paderborn
Warburger Str. 100
D-33098 Paderborn
Germany

Reply all
Reply to author
Forward
0 new messages