Dr. Richard F L Evans, lecturer in Condensed Matter Theory in the physics department at the University of York, gave a stimulating seminar exploring the current state-of-the-art in atomistic modelling of magnetic materials and presenting calculations of two industrially important magnetic materials: NdFeB permanent magnets and IrMn/CoFe bilayers for spin valve sensors.
The first part of seminar consisted in an introduction to atomistic spin models and spin dynamics. In this part, he gave an explanation of some basic concepts as ferro- and antiferro-magnetism as well as exchange interaction concept. He also emphasized how magnetic materials are essential to the operation of a plethora of technological devices, from hard disk drives to hybrid electric vehicles and wind power generators. Then, he introduced the stochastic Landau-Lifshitz-Gilbert equation and spin dynamics. Next, he lingered on Vampire: an open source software simulation package for atomistic simulation of magnetic materials (Dr. Evans is the lead developer of this software). He highlighted the fact that Vampire employs a massively parallel code that can make spin models a lot bigger than they would ordinarily. He corroborated these claims by showing runtime vs number of CPUs plots.
In the second part he explained properties of Nd2Fe14B permanent magnets with respect to the temperature. He introduced the concept of Curie temperature and showed how magnetic properties are affected by the temperature. He also explained the concept of magnetic anisotropy and hysteresis effect and how to account these features in simulations. Finally, he showed results of simulations of IrMn3 related to pinned interfacial spins in exchange bias, underlying that the structure at atomic and granular length scales determines overall magnetic properties.
Throughout the seminar, Dr. Evans emphasized the importance of running simulations at different scales to account for particular features of the physical system in the simulation models. He constantly highlighted the confrontation of simulation results with analytical models. He concluded by saying that future technological improvements rely on a detailed understanding of how these complex materials behave and how their properties can be further optimized.