[gmx-users] Microcanonical

[A B]

I am curious about the reason for the heating that allegedly occurs with 
cutoffs in NVE simulations ("The use of (electrostatic) cutoffs is well 
known to cause heating in NVE simulations."). This should not, and does not, 
happen if an appropriate smoothing function is applied. Does this heating 
mean that cutoffs are implemented in a sub-optimal fashion in GROMACS, or is 
it just a bug? Adding a heat bath would just hide the symptom, since there 
would still be an energy flow from the heat source, so simulations in other 
ensembles would also be affected?

_________________________________________________________________
Lev livet lite sundare. MSN Hälsa. http://www.msn.se/halsa/