Abstract
We propose to compute the thermodynamic properties of many-body systems using a path-sampling Monte Carlo scheme implemented in a generalized path ensemble. Trial paths are generated through an expanded ensemble using a reversible discretization of Langevin's equation of motion. We also show how the systematic errors resulting from the use of a finite time step can rigorously be taken into account in the path-sampling scheme. We find that the degree of convergence of the estimated thermodynamic quantity towards the exact value correlates with the mean acceptance rates of the path-sampling scheme. An application of the path method for simulating glassy systems is finally suggested.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
