Metadynamics Questions: COLVARs yielding unexpected values.

[Rolf David]

Hi,

First, it's in a.u, so bohr radius (0.53 to get Å)

First CV1 ranges from 4 to 5.5 Å 
CV2 ranges from 1.7 to 3.1 Å

What do you mean by "should be" ?
It's the value form your input (the CV1 start at 1.7Å) or the one you want 
(at CV1 = 1.7Å my reaction is "finished") ?

If it's the first case, that's wierd. 

If it's the second  either you started from 4Å and it's easier to explore 
longer distance (so you'll have to put a wall to force the exploration to 
go the other way), if you've started from 5.5Å, you can restart. 

For the CN, for a *very simple *example you ideally go for 0 (nothing) to 1 
(you put one atom near the other) 

Metadynamics can be restarted: either without new hills, or with new hills 
(the same, like 1e-3/15 steps or you can change), but if i'm correct you 
cannot add new colvar.
Restarting is the same as a standard MD (and with RESTART_METADYNAMICS)



On Tuesday, April 28, 2015 at 2:10:13 AM UTC+2, Christopher O'Brien wrote:
>
> CP2K users,
> I am obtaining very unexpected results from the COLVARS of the 
> METADYNAMICS module. I had to piece together an input from the mail list 
> and what I could learn from the manual, but I think I have all of the 
> relevant additional input required for METADYNAMICS. Please tell me if I'm 
> wrong. 
>
> I am obtaining very large COLVAR distances for variable 1 and entirely too 
> small ones for variable 2. COLVAR 1 should be around 1.7A, but it ranges 
> from 7.99 to 11.17 (Bohr?). COLVAR 2 should be about 1.5A, but it appears 
> to vary from 3.621-6.252 (Bohr?). (Coordination, COLVAR 3, varies from -0.2 
> to 0.3, but I have not idea what it should be). The range of distances were 
> reported in by the fes.sopt code which I ran with the options:
> fes.sopt -ndim 3 -find-minima -cp2k -file xxx.restart
>
> I am also interested to learn if METADYNAMICS can be restarted. There are 
> a lot of unexplained keywords that have something to do with restarting 
> METADYNAMICS in the manual, but I don't know if we have to use them. 
>
> I have included the relevant section from the &MOTION input section and 
> the &FORCE_EVAL/&SUBSYS input section below:
>
> &MOTION  
>    &FREE_ENERGY
>
>    METHOD METADYN 
>
>    &METADYN 
>
>     DO_HILLS .TRUE. 
>
>     WW [hartree] 1.0E-3 
>
>     NT_HILLS 15 
>
>
>     &METAVAR 
>
>      COLVAR 1 
>
>      SCALE 0.08 
>
>     &END METAVAR 
>
>
>     &METAVAR 
>
>      COLVAR 2 
>
>      SCALE 0.08 
>
>     &END METAVAR 
>
>     &METAVAR 
>
>      COLVAR 3 
>
>      SCALE 0.08 
>
>     &END METAVAR 
>
>     &PRINT 
>
>     &COLVAR MEDIUM 
>       &EACH 
>
>         METADYNAMICS 1 
>
>         MD 15 
>
>       &END EACH 
>
>       FILENAME xxx_colvar 
>
>      &END COLVAR 
>
>     &END PRINT 
>
>    &END METADYN 
>
>    &FREE_ENERGY_INFO MEDIUM 
>
>      &EACH 
>
>        METADYNAMICS 1 
>
>        MD 15 
>
>      &END EACH 
>
>      FILENAME xxx_freeenergy 
>
>    &END FREE_ENERGY_INFO 
>
>   &END FREE_ENERGY
>
> &END MOTION
>
> And the &SUBSYS section containing the COLVARS:
>   &SUBSYS
>
>     &CELL 
>
>       ABC 12.97658 12.90794 12.97688 
>
>       PERIODIC XYZ 
>
>     &END CELL 
>
>     &COLVAR 
>
>      !P to F
>
>      &DISTANCE 
>
>       ATOMS 11 17 
>
>      &END DISTANCE 
>
>     &END COLVAR 
>
>     &COLVAR 
>
>      !P to O
>
>      &DISTANCE 
>
>       ATOMS 11 9 
>
>      &END DISTANCE 
>
>     &END COLVAR 
>
>
>     &COLVAR 
>
>      !see 10.1021/jp301490k for ND NN R0 parameter meaning 
>
>      ! P to O in water 
>
>      &COORDINATION 
>
>       ND 12 
>
>       NN 6 
>
>       R0 [bohr] 3.0 
>
>       KINDS_FROM P 
>
>       KINDS_TO Ow 
>
>      &END COORDINATION 
>
>     &END COLVAR
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20150428/7393f324/attachment.htm>