[CP2K:191] Re: metadynamic

[Teodoro Laino]

Velocity rescaling is a quite well established technique (please  
refer to the thousands of paper discussing merits and drawbacks of it)..
if you want the kinetic energy of a set of variable to be equal to a  
certain value within a tol (with velocity rescalinf) this is the  
correct equation to use:

>  IF(ABS(ekin_w-meta_env%ekin_s)> tol_ekin)THEN
>

With this one :
> should be changed into (the following is my own change, not from CVS)
>   IF((-ekin_w+meta_env%ekin_s)> tol_ekin)THEN
>

the temperature of your collective variable could be much lower than  
the one at which you may want to sample the free energy..
No doubt the temperature of the collective variables should be equal  
to the temperature at which you want to compute the free energy.

> As I said yesterday, my understanding is that only the high  
> velocity should be avoided.  If the low velocity is also scaled to  
> high velocity, variable will always decreas or increase linearly.

Teo

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20070714/fc944721/attachment.htm>