Early Detection of Mechanical Faults in Loads Coupled with Asynchronous Electric Motor

Optimising methods for diagnosing and monitoring asynchronous electric motors (AEM) requires a methodical and integrated approach. This experimental work proposes to adopt the spectral analysis of vibrations based on the Fast Fourier Transform (FFT). It is obvious that the FFT fails to provide temporal information and to manage the identification of several faults with their time of appearance, however it remains the most widely used signal processing technique that provides useful information on frequency content, and it is often adopted in industries with a large number of vibration monitoring points, such as heavy industry plants. This is because FFT can identify and characterise the frequency signatures associated with mechanical faults, such as imbalance and misalignment, providing an in-depth understanding of structural anomalies. By performing a preliminary spectral analysis, the specific frequency characteristics of mechanical defects become clear, enabling early and accurate identification. This approach has several advantages, including greater sensitivity to weak signals and the ability to distinguish mechanical anomalies from normal fluctuations. The adoption of the FFT of the vibration signals from the AEM, in the presence and absence of mechanical faults for an 11 kw machine as part of this experimental work, clearly demonstrates the value of the frequency analysis of vibration signals using the FFT.