[gmx-users] Wrong heat capacity (water)

[Andreas Kring]

Hello all,

I ran a MD simulation of TIP4P water in the NVT ensemble (976 molecules, 
density=0.98 g/cm3, T=300K (Nose-Hoover), SETTLE and PME enabled). This 
works out fine, i.e. I get a correct radial distribution function 
(g_OO), but the heat capacity (C_V) is wrong. I calculated it using the 
command:

$  g_energy -f ener.edr -ndf 6 -nmol 976

and get the value below (which sould have been approximately 75 J/(mol*K)).

Is the command above incorrect?

Kind regards
Andreas



Energy                      Average       RMSD     Fluct.      Drift 
Tot-Drift
-------------------------------------------------------------------------------
Potential                  -41.1524   0.478018          0   -3.69759 
-36.9796
Kinetic En.                 7.48322   0.282052          0   -4.33977 
-43.4021
Total Energy               -33.6692   0.604335          0   -8.03739 
-80.3819
Temperature                 300.006    11.3076    11.2959  -0.178262 
-1.7828
Heat Capacity Cv:     -3.94726 J/mol K (factor = 2.77307)