[CP2K-user] Setting NOISY_GAMMA for different atom kinds in SGCPMD

[Marcella Iannuzzi]

Hi Matt, 

You should be able to use different gammas by activating thermal regions 
addressing specific subsystems. 
You need a LANGEVIN section with GAMMA and a THERMAL_REGION section with 
DEFINE_REGION subsections, 
each of them containing a NOISY_GAMMA_REGION keyword and a LIST of 
particles. 

Kind regards
Marcella

On Wednesday, July 14, 2021 at 5:41:18 PM UTC+2 Matt D wrote:

> Hi all,
>
> I'm trying to run second generation Car-Parrinello MD (i.e. Thomas Kühne's 
> method in Kuhne, TD; Krack, M; Mohamed, FR; Parrinello, M., PHYSICAL REVIEW 
> LETTERS, 98 (6), 066401 (2007) ) in CP2K but using different values of 
> NOISY_GAMMA for different species.
>
> In Kühne's paper it suggests this is possible, but I can't seem to find 
> the relevant keyword in the manual. I've also seen different values of 
> NOISY_GAMMA being used in recent papers using SGCPMD, for example in work 
> from Marcella Ianuzzi and Jun Cheng's groups (JACS Au 2021, 1, 569−577) so 
> I believe it is possible.
>
> So far I have tried:
>
> NOISY_GAMMA {value1} {value2} {value 2}
> and 
> GAMMA_D {value1} {value2} {value3}
>
> but CP2K version 7.1 doesn't recognise GAMMA_D as a parameter (I 
> informally heard it used to used in earlier implementations) and 
> NOISY_GAMMA is expected to be a single value, rather than multiple values.
>
> If anybody can share/suggest what might be the correct syntax I'd greatly 
> appreciate it.
>
> Thanks.
>
> Best wishes,
> Matt
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210715/dddd749f/attachment.htm>