Rotor eccentricity in electric machines: An analytical framework with experimental validation for fault diagnosis

초록

Rotor eccentricity is a major fault in electric machines, as it distorts the air-gap magnetic field and generates unbalanced electromagnetic forces that can excite vibration and reduce reliability. While numerous studies have explored aspects of eccentricity, most remain fragmented, lacking a unified framework that connects electromagnetic modeling with diagnostic applications. This paper presents an analytical framework for fault diagnosis of rotor eccentricity, integrating mathematical modeling, predictive analysis, and multi-fidelity validation. A rigorous analytical model is first established to quantify electromagnetic force components arising from both static and dynamic eccentricity, clarifying their spectral characteristics. Based on this formulation, a predictive approach is developed to anticipate eccentricity-related fault features in a physically interpretable manner. The framework is validated through finite element analysis and experimental tests on a custom-designed motor with adjustable eccentricity, confirming its accuracy and practical feasibility. The results demonstrate that the proposed approach not only deepens understanding of eccentricity-induced electromechanical interactions but also provides a reliable basis for early fault detection in electric machines.

키워드