Developing ways to measure and use eddy current and hysteresis loss data in permanent magnets used in dynamic motor environments Copy Link

Hirst Magnetic Instruments have been providing solutions for 60 years in magnetics and magnetic measurement. Hirst manufactures precision hand-held Gaussmeters, Fluxmeters, bench top & workstation industrial magnetisers & demagnetisers, industrial production-line magnetisers and pulsed field magnetometers (PFMs) for developing / characterising magnetic materials. These instruments are used in multiple industries including aerospace, automotive, electronics and medical equipment to support the productions of a range of different end use products.  
 
Currently Hirst Magnetics carry out magnetic characterisation measurements using a PFM. This collects a range of different magnetic parameters that are both static and dynamic in nature. The measurements made on a PFM are used to generate static data sheet parameters for high performance permanent magnets – the dynamic information is currently not exposed or shared outside of the PFM but it offers useful information that we can address. 
 
Current models and simulations use static measurements and magnet parameters to predict the performance of the motors. These usually have a significant error in predicting performance and an experience factor tends to be incorporated into product design, the experience factor is not linear, and increases with magnet size and motor power. 
 
In the design of permanent magnet synchronous machines, the analysis of losses usually focuses on the properties of the stator laminations. The losses in the rotor are often considered negligible: Since the rotor moves synchronously with the rotating stator field, the fundamental oscillations of the stator field are not seen as time-varying fields in the rotating components. Flux changes at higher harmonic frequencies, however, will give rise to eddy currents in the electrically conductive rare earth rotor magnets. These harmonics occur due to modulation of the main rotating field by the stator teeth (ripple) and due to the power supplies.  
In cases where the eddy current losses in the rotor magnets become too significant, the magnets are segmented into thin, electrically insulated slices. The eddy current losses and their reduction by segmentation are predictable (provided conductivity with frequency is known, but it is not clear these are predictions are accurate and PFMs can in theory directly measure this parameter), using numerical methods or analytical equations. It thus appears straight forward for the designer of a machine to determine the optimum degree of segmentation for a given machine—balancing the reduction of eddy current losses against the loss of magnet volume to the glue lines and against the considerable manufacturing cost for segmenting the magnets. Other sources of losses, in particular, magnet hysteresis losses, are not considered. The hysteresis in the permanent magnets is not however negligible, as their coercive field will obviously be much higher than the amplitude of any field oscillations. However, magnets typically contain regions of lower coercive fields. In particular, the surfaces of rare earth magnets are known to have considerably reduced coercive fields and thus potentially cause hysteresis losses. The creation of additional magnet surfaces by segmenting is therefore not without risk. Determining the degree of segmentation without considering possible surface losses can potentially lead to incorrect decisions during the design of an electrical machine. Thus, there is a need to measure these parameters more accurately. Pulsed Field Magnetometers such as the instruments manufactured by Hirst use pulses of differing lengths to generate a static measurement of magnetic properties, but it is clear the information contained in these pulses is dynamic and could allow these eddy current and hysteresis losses to be measured at the same time as the static data sheet parameters.   
 
There is a need to develop a way to measure and use eddy current and hysteresis loss data within models that operate dynamically. The ability to measure these losses and used them in modelling software would allow for better motor design leading to a more robust and complete design and perhaps reduced cooling requirements or reduce the grade and cost of magnets needed. This potential magnet material saving could reduce the required quantity of critical materials found in magnetic products and reducing the reliance on insecure supply chains.