Numerical evaluation of energy barriers and magnetic relaxation in interacting nanostructured magnetic systems

We discuss a model to quantify long-time thermally induced magnetization reversal in magnetic systems with distributed properties. Two algorithms, based on kinetic and Metropolis Monte Carlo are introduced. While the former requires the constant recalculation of all energy barriers and is useful when the interactions are weak, the latter uses the Metropolis Monte Carlo to estimate the magnetization trajectory and, consequently, only the most probable transition rates are evaluated. The ridge optimization method is used to evaluate the energy barriers in a multidimensional energy landscape. The algorithms are applied to a granular system modeled by means of Voronoi polyhedra and having random in-plane anisotropy. [Copyright &y& Elsevier]