[CP2K-user] Thoughts on RDFs for NPT ensemble simulations?

[Travis]

Hi,

You can change the effective atom size in TRAVIS to skip molecule 
recognition (say you don't agree to the molecules recognized, follow 
prompts to change atom size to 0.001), so you would compute a RDF between 
atom type A and atom type B. 

-T


On Tuesday, August 27, 2019 at 8:16:02 AM UTC-3, Ant wrote:
>
> Update:
>
> It would be great if someone could provide some guidance on how to use 
> CP2K to plot RDFs for existing NPT simulations (where the cell 
> volume/parameter varies at each timestep).  VMD accepts only a single 
> volume value.
>
> Travis, who commented earlier, has provided part of a solution (using 
> REFTAJ to read a trajectory back into CP2K and, I think, to modify it so it 
> can be read into VMD), but unfortunately, as a near-newbie, I don't know 
> what the next step would be.  I know I would need to feed in the changing 
> call parameters at each timestep from my *.cell files, but don't know how 
> to do this and also don't know what I would need to do in VMD to get 
> accurate RDFs from it (I'd still need to enter a volume value in the RDF 
> GUI and don't know what that should be).  There doesn't appear to be any 
> documentation on this online, but I suspect it must exist because I expect 
> that other people must have wanted to plot NPT RDFs.  If anyone knows of 
> any, I would be grateful to know where to find it.
>
> Travis made a helpful suggestion about using the travis code.  I have run 
> some examples, but it doesn't seem to do what I need it to.  I have run ab 
> initio simulations and TRAVIS seems only to be able to calculate RDFs when 
> all possible complexes are identified.  This would take a very long time 
> for my system and for all of my simulations.  In fact I want RDFs and 
> running coordination numbers in part to identify the chemical complexes.
>
> If anyone has any hints at all, I would be willing to pay at this point 
> for advice if anyone knows of anyone who deals with small jobs.  I 
> understand from Travis that it is possible to create output for NPT 
> simulations that is readable by NPT for making RDFs, but unfortunately, I 
> can't rerun my simulations.  Thank you very much.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190827/3714274a/attachment.htm>