[gmx-users] Re: Calculating Temperature from Velocity

Andy Shelley robert.shelley at gmail.com
Mon Sep 22 19:58:52 CEST 2008

A few more questions added to my first response.  When calculating the
Temperature should I use the mass of the molecule or of the atoms times the
velocity of the atoms or velocity of the molecule? Also, when looking at the
output of g_traj if I create a small group of atoms close to each other, say
10-12 atoms, the temperature of these atoms varies alot while the
temperature the whole system does not, why is this?  Is there a way to
accurately determine the temperature of a small number of atoms?

Thanks again,

On Mon, Sep 22, 2008 at 10:42 AM, Andy Shelley <robert.shelley at gmail.com>wrote:

> I am trying to calculate the temperature from the final velocity posted in
> the confout.gro file after a simulation runs. I have implemented a berendsen
> thermostat to control the temperature to 300K and ran it for 1ns. Looking at
> g_energy the final temperature of the system is 299.981.
> Statistics over 10001 steps [ 995.0001 thru 1000.0001 ps ], 1 data sets
> Energy                      Average       RMSD     Fluct.      Drift
> Tot-Drift
> -------------------------------------------------------------------------------
> T-System                    299.981    1.36143    1.35806  0.0663083
> 0.331574
> From the manual section 3.4.3 the temperature can be calculated from T=
> sum[mass_i v_i^2] / (#degrees of freedom * boltzmann constant) where number
> of degrees of freedom=3N-N_c-N_com. I used N_com =3 because the center of
> mass velocities are constants of the motion. If I only control the
> temperature does this effect the number of constrants on the system? When I
> calculate the temperature by hand with # degrees of freedom =3N-3 I get a
> temperature of 317.5. Any ideas why I can not calculate the same temperature
> g_energy does?
> Thanks,
> Andy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20080922/c6163812/attachment.html>

More information about the gromacs.org_gmx-users mailing list