Magazine | Issue 1 2025 The thesis is aimed at developing objective evaluation criteria for identifying damaged windings. These evaluation criteria were based on the algorithms already present in the frequency response analysis software and a separate assessment method. The frequency ranges must be slightly modified and adapted to the rotor equipment for the evaluation algorithms already present in the software. Based on the equipment we investigated for the thesis, we were able to show that SFRA is more sensitive than the RSO method our company has been using. Additionally, we were able to establish objective evaluation criteria for the rotors. You mentioned that the objects you investigated revealed the new method to be more sensitive than the RSO. Is this something you can quantify? Pascal Fröhlich: During the thesis, we looked at seven rotors in detail. We investigated criteria such as reproducibility, differences in the rotor types (non-salient pole or salient pole), measurements when the rotor was pulled or in the stator, and other parameters. Coincidently none of the seven rotors investigated had interturn short circuits. To achieve a comparison between the two methods and a conclusion which is the most suitable way we recreated measurements with various artificial faults. The faults created should be small, which means a short-circuit of one or just a few turns to simulate the reality as close as possible and to create a challenging situation for the measurement methods Such shorts could be the result of the rotor’s designs or combined stresses during operation. Frequency response analysis detected all the faults, but RSO only recognized the most significant one, meaning single interturn faults would go undetected with RSO. What are the limitations of frequency response analysis? Pascal Fröhlich: In principle, a fault located exactly in the winding’s center won’t be detected, but the same is true of an RSO measurement. These interturn short circuits could only be detected when compared with a fingerprint reference measurement. I’d like to refer to the rotor winding’s electrical equivalent circuit diagram, which can be divided into infinite consecutive individual parts depending on your point of view. So, a situation where a fault occurs in the absolute center is highly unlikely and doesn’t constitute a limitation in practice. Additionally, the evaluation criteria were derived from seven rotors. Although our findings have been verified several times since the thesis, I could still imagine further investigations being carried out. We’d like to thank both of you for taking the time to speak with us. 39
RkJQdWJsaXNoZXIy NTkxNzY=