Applications of micromagnetics

Publisher Website
Google Scholar
Abstract
The main idea in the original development of the theory of micromagnetics was to avoid the crude approximations and the arbitrary assumptions and constraints used in the domain theory. It was hoped that a theory could be built which would describe rigorously all the magnetization processes, from the nucleation of a wall surrounding a reversed domain, through the motion of this wall in the crystal, to its final disappearance in the saturated sample, including high-frequency phenomena like ferromagnetic resonance. In principle this goal can be achieved, at least for ideal crystals, by solving a certain set of non-linear partial differential equations. The solution is definitely not easy, but it could be done for relatively simple cases with modern computers. Choosing, say, a spherical single crystal of an appropriate material, should lead to computations which are very lengthly, but not prohibitively so. The reason this has never been done is that it is known in advance that a solution for such an idealized case cannot have much physical significance in the application to any real sample. This is known from studies of the nucleation of magnetization reversal in a previously saturated crystal for several simple geometries, in which either an analytic solution or upper and lower bounds could be given, as discussed in a previous review.1 In all these cases, the calculated nucleation field approached the experimental values only for experiments done under conditions approaching the idealized assumptions, but was very different from the results of the more usual experiments. The difference was also qualitatively understood1 as being mostly due to surface roughness in soft magnetic materials, and due to lattice defects (probably dislocations) in hard magnetic materials.