Demagnetization Studies on Permanent Magnets : Comparing FEM Simulations with Experiments

In a world where money often is the main controlling factor, everything that can be tends to be more and more optimized. Regarding electrical machines, developers have always had the goal to make them better. The latest trend is to make machines as efficient as possible, which calls for accurate simulation models where different designs can be tested and evaluated. The finite element method is probably the most popular approach since it makes it possible to, in an easy and accurate way, get numerical solutions to a variety of physics problems with complex geometries and non-linear materials. This licentiate thesis includes two different projects in which finite element methods have had a central roll. In the first project, the goal was to develop a simulation model to be able to predict demagnetization of permanent magnets. It is of great importance to be able to predict if a permanent magnet will be demagnetized or not in a certain situation. In the worst case, the permanent magnets will be completely destroyed and the machine will be completely useless. However, it is more probable that the permanent magnets will not be completely destroyed and that the machine still will be functional but not as good as before. In a time where money is more important than ever, the utilization has to be as high as possible. In this study the demagnetization risk for different rotor geometries in a 12 kW direct driven permanent magnet synchronous generator was studied with a proprietary finite element method simulation model. The demagnetization study of the different rotor geometries and magnet grades showed that here is no risk for the permanent magnets in the rotor as it is designed today to be demagnetized. The project also included experimental verification of the simulation model. The simulation model was compared with experiments and the results showed good agreement. The second project treated the redesign of the rotor in the generator previously mentioned. The goal was to redesign the surface mounted NdFeB rotor to use a field concentrating design with ferrite permanent magnets instead. The motivation was that the price on NdFeB magnets has fluctuated a lot the last few years as well as to see if it was physically possible to fit a ferrite rotor in the same space as the NdFeB rotor. A new rotor design with ferrite permanent magnets was presented together with an electromagnetic and a mechanical design