Ascent dynamics: an efficient algorithm probing long timescale dynamics

MD simulations is a powerful tool of studying the atomistic behavior of disordered and amorphous materials, but the covered timescales are 7-8 orders of magnitude shorter than dynamics scale when system approaches to glass state. Here, we present a new accelerated simulation method, ascent dynamics, which allows the system to escape deep energy minima through crossing saddle points with given index explicitly at finite temperatures. Master equation approach is then used to analyze the trajectory data to compute viscosity and relaxation time. Using this new algorithm, relaxation time and viscosity are not only in well agreement with traditional molecular dynamics simulation at high temperature, but they can be probed reliably at the low temperature regime. Our results show that we achieve over 10 orders of magnitude gain in the equilibration time scale compared to conventional methods, thus paving the road to computational studies of long timescale phenomenon.