Analysis of the influence of different manufacturing methods on the thermal aspects of permanent magnet electrical machines

Abstract

Accurate thermal modelling of permanent magnet (PM) electrical machines is of primary importance for improving the overall design in terms of efficiency, torque/power density and over-load time capability. The investigation of crucial thermal parameters leads to achieve a more homogenous temperature distribution and avoid hot spots in electrical machines. This doctoral research employs different manufacturing methods for improving machine thermal performance such as coil pressing and stator end windings encapsulation for PM machines. A number of thermal modelling approaches including analytical and computational finite element (FEA) methods are utilised to identify the machine thermal parameters within an acceptable accuracy and the results are validated experimentally. The effect of uncertainty and sensitivity of important thermal parameters have also been studied thoroughly. Coil pressing improves the thermal characteristic of PM machines by packing the conductors in a smaller slot volume but could lead to a degradation of winding insulation, affecting machine reliability. A first step for investigating the coil pressing is to ensure that the proposed method does not deteriorate the mechanical integrity and electrical durability. Then, attention is focused on the estimation of the effective thermal conductivity of compressed coils. Lastly, a set of accelerated life tests, when the acceleration variable is temperature have also been conducted to investigate the life expectancy of the on-tooth compressed stator windings. Encapsulation of the stator end windings by adding thermal paste into the machine end region is another method having potential for the thermal improvement of PM machines. This method aims to enhance the cooling capability of the machines, which do not have dedicated cooling instruments. A permanent magnet alternator (PMA) with and without thermal paste for an aerospace application has been investigated thermally in this case. Analytical and computational temperature estimation techniques have been employed simultaneously to analyse the thermal improvement of the machine quantitatively. The mechanism of heat flow through a flange mounted PMA is also described by developing an equivalent thermal lumped parameter model.