Abstract :
The paper discusses existing practice of specifying temperature-rise of electrical machinery by thermometers, and points out that the B.E.S.A. Rules are more satisfactory but are not generally adopted, probably on account of the higher limits. The relations between surface temperatures and internal or hot-spot temperatures are discussed, together with the influence of the load on the temperatures of different parts of a machine. Location of hot spots in the axial direction is considered, as well as the heat flow through the insulation. The influence of the thickness of insulation on the relation between internal and surface temperatures is dealt with, and a diagram showing the radial temperature distribution is given. The importance of considering the voltage of a machine in connection with temperature-rise on account of the greater temperature gradient through the insulation is pointed out. Consideration is given to the question of operating large turbo plants on ultimate temperature rather than on temperature-rise, on account of the increased load obtainable at low air temperatures. A plea is made to operating engineers to call for the fitting of embedded temperature detectors, although temperature guarantees based on this method may not be obtainable. A series of tests on mica-insulated bars is described, including heating runs on bars insulated with mica wraps containing varying percentages of mica, and it is shown that the percentage of mica has a direct bearing on the breakdown voltage which the bars will stand after heating. The results of bending tests on bars before and after heating are shown. The conclusion arrived at is that mica insulation, as used in modern turbo-alternators, retains its mechanical and electrical properties at temperatures up to 200° C. It is suggested that the limit of the ultimate temperature should be raised to 160° C. instead of the usual value of 125°C.
