Use of Langevin Thermostat to correctly sample boltzmann distribution

mercredi 17 septembre 2014

In semiclassical and classical molecular dynamics it appears common to use thermostatted molecular dynamics in order to predict e.g. static equilibrium properties.



In all the papers I have read they merely state that attaching a thermostat (specifically Langevin) to the dynamics is a stationary process, conserving the equilibrium distribution of the non thermostatted system.



Whilst I can see intuitively that this makes sense I was hoping for help on how one proves that attaching a Langevin Thermostat does indeed rigorously generate the correct canonical ensemble.



many thanks





0 commentaires:

Enregistrer un commentaire