Wavelet-based high order spectrum for local damage diagnosis of gears

Gears play an important role in transmission systems, allowing for high performance in terms of load capacity and efficiency. Common failures, such as spalling and pitting of gear teeth, can occur as a result of contact fatigue, excessive load, or sudden impact. Starting from an initial stage, their steady growth can lead to irreparable damage and unexpected breakdowns. Therefore, condition monitoring of gears is critical for smooth operation of drivelines and transmissions. Local tooth damage produces transient impacts in the vibration signals, exhibiting non-linear characteristics. Taking into account its capacity to characterize the phase coupling between signal components caused by non-linearities, wavelet-based high order spectrum is considered to be effective to attain reliable fault-related features. Among others, wavelet bicoherence has been successfully estimated to detect artificially created gear faults. However, a number of challenges still remain in selecting the most informative bi-frequency bands and extracting instantaneous diagnostic features. Moreover, the adaptability of the technique in identifying natural defects on multiple teeth still needs further investigation. This paper presents a novel strategy for selecting informative bi-frequency bands and extracting instantaneous diagnostic features in the time bi-frequency domain. The performance of the proposed methodology is applied, evaluated and compared with a state-of-the-art method on scenarios that include diagnosis of single-tooth and multiple-teeth damage. To validate its effectiveness, the methodology is tested using three datasets: two featuring artificially induced pitting damage on one gear tooth, and a third one containing naturally developed spalling defects on two gear teeth.