Vibroacoustic Diagnostics of Rolling Bearings - Test Rig for Research Dataset Acquisition and Didactic Applications

The chapter begins by emphasizing the importance of rolling bearings in various industries and the critical need for effective diagnostics to prevent failures that can lead to significant downtime and safety hazards. It explores the generation of vibroacoustic signals by bearing faults, which can be analyzed to gain insights into the condition of the bearing assembly. The text delves into the various sources of machine spindle vibrations, including manufacturing defects, misalignments, and damage to the rolling elements. It discusses the natural frequencies generated by rolling element bearings and the impact events that occur due to surface anomalies, which can excite the bearing assembly and machinery. The chapter presents a novel test rig designed for research and educational purposes, capable of acquiring vibration signals over a wide range of rotational speeds. It details the experimental setup and the analysis of recorded data, demonstrating the identification of bearing damage through signal spectrum analysis and envelope detection. The text also highlights advanced signal processing techniques, such as spectral kurtosis and fast kurtogram algorithms, which enhance the diagnostic capabilities. Furthermore, it explores the application of machine learning techniques for bearing condition monitoring and the prediction of bearing wear using state space models and deep learning approaches. The chapter concludes by discussing the versatility of the test rig for research and educational applications, emphasizing its role in acquiring datasets for machine learning models and artificial intelligence algorithms.