Hi sorry I was unclear. I am running a cluster that was originally
part of a bulk system. Since the waters were originally surrounded
by a bulk solution, I ran the cluster for 50 ps of NPT to try and
generate an initial set of coordinates that were more reasonable.
Then, I ran NVE using TEMPI to assign the initial velocities.
I've also tried just minimizing the initial coordinates and then
running the cluster in NVE and get the same results. There still
appears to be a collective rotation about the center of mass. After
minimization here is the output from the dynamics run (NVE) related
to the COM correction for the first step:
check COM velocity, temp: 0.029286 1.39(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
...
Here is the output for the 100th timestep:
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
check COM velocity, temp: 0.000000 0.00(Removed)
...
Do you have another recommended way to start running an isolated
water cluster to avoid this problem? Or, is this something I
shouldn't worry about unless it actually results in an unphysical
cluster geometry due to the rotations?
With thanks,
Jennie
On May 14, 2007, at 2:37 PM, David A. Case wrote:
> On Mon, May 14, 2007, Jennie Thomas wrote:
>
>>
>> The temperature fluctuates around 250 K, which seems normal. I've
>> equilibrated the system in NPT and experienced the same issue with
>> lack of removal of rotations.
>
> Now I am confused: NPT simulation only makes sense for periodic
> simulations.
> What is the relation to your previous question about simulating "an
> isolated
> small water cluster using Amber 8 (NVE ensemble)"? Do you have
> periodic
> boundary conditions turned on? If so, no rotation would be
> removed, since
> that doesn't make sense for periodic simulations. (See the
> discussion of the
> NSCM variable in the manual).
>
> ...dac
>
> ----------------------------------------------------------------------
> -
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed May 16 2007 - 06:07:22 PDT
