A parallel finite element micromagnetics package has been implemented, which is highly scalable, easily portable and combines different solvers for the micromagnetic equations. The implementation is based on the standard Galerkin discretization on tetrahedral meshes with linear basis functions. A static energy minimization, a dynamic time integration and the nudged elastic band method have been implemented.
The static energy minimization method is used for the investigation of domain wall pinning processes in SmCo permanent magnets. The pinning of magnetic domain walls on the precipitation structure and the influence of material parameters, cell structure, and cell geometry are studied in detail. The thickness of the coherent precipitation plays an important role, since it has to be thicker than the domain wall width for effective pinning, but it must not be too thick, which would allow the reversal of the whole intercellular structure. Nucleation and magnetization reversal processes in FePt nanoparticles are investigated for particles with single and multiple easy axes. The results show a strong reduction of coercivity if more misaligned easy axes within a particle are assumed and the particle size is reduced.
The static as well as the dynamic properties of the magnetic vortex state of soft magnetic nanodots are studied using the time integration of the dynamic Landau-Lifshitz-Gilbert equation. A phase diagram of the magnetic ground state of magnetic nanodots has been obtained and a comparison with an analytical vortex model and experimental results is given. Vortex precession and radial excitation modes are calculated and their eigenfrequencies measured. Finally, the properties of elliptical and rectangular permalloy nanoparticles are studied. The shape and the demagnetizing field play an important role in the magnetization reversal process. In chains of these particles magnetostatic coupling leads to stable magnetization configurations with antiparallel magnetization in neighboring particles.