Comprehensive experimental study on a wound-rotor asynchronous machine fault detection through frequency response analysis

Frequency Response Analysis (FRA) is a well-established diagnostic technique for detecting various faults in power transformers, with emerging applications in condition monitoring of electrical machines. However, existing studies on applying FRA to electrical machines often suffer from limited experimental validation, neglect of rotor fault detection, and a lack of systematic evaluation of circuit configurations. Moreover, most prior investigations rely heavily on visual interpretation of FRA signatures, without exploring quantitative sensitivity or the influence of rotor–stator interactions. This paper addresses these gaps through a comprehensive experimental analysis conducted on a three-phase wound-rotor asynchronous machine. The effect of rotor angular position on FRA signatures is examined in healthy and faulty conditions, revealing significant variation even in cylindrical rotors and underscoring the need for rotor fixation to ensure diagnostic reliability. The study identifies best circuit connections, demonstrating that floating the stator terminals during rotor FRA tests and vice versa enhances FRA sensitivity. Various faults including inter-turn short circuits, phase-to-phase short circuits, and insulation failure are successfully detected, confirming the robustness of FRA in this context. Additionally, it is shown that single-phase measurements on the rotor are more sensitive than traditional two-phase series measurements in FRA of the rotor. Notably, rotor faults induce detectable effects on stator FRA responses. This phenomenon poses both challenges and opportunities, highlighting the need for deeper investigation in future studies.