Hysteresis multicycles in nanomagnet arrays.

We predict two physical effects in arrays of single-domain nanomagnets by performing simulations using a realistic model Hamiltonian and physical parameters. First, we find hysteretic multicycles for such nanomagnets. The simulation uses continuous spin dynamics through the Landau-Lifshitz-Gilbert (LLG) equation. In some regions of parameter space, the probability of finding a multicycle is as high as approximately 0.6 . We find that systems with larger and more anisotropic nanomagnets tend to display more multicycles. Our results also demonstrate the importance of disorder and frustration for multicycle behavior. Second, we show that there is a fundamental difference between the more realistic vector LLG equation and scalar models of hysteresis, such as Ising models. In the latter case spin and external field inversion symmetry is obeyed, but in the former it is destroyed by the dynamics, with important experimental implications.