Monitoring of gear wear progressions based on a modulation signal bispectrum analysis of vibration

Gear tooth wear is a common fault mode inevitably happened in gear transmission system, which could result in many failures such as surface pitting, scuffing, and tooth breakage. Moreover, the failures developed from tooth wear may cause catastrophic economic loss and safety hazard. Therefore, condition monitoring of gear tooth wear is critically important to ensure safe and reliable operations of gearbox systems. With the decades of development, many methods have been applied to the condition monitoring of gear transmission system, while among which vibration analysis are the most widely used and generally accepted method due to its high reliability and easy installation of sensors compared with other methods. Although vibration analysis has been developed for many decades, the techniques for gear wear monitoring have not been well-established. In order to strengthen research in this respect, this study will conduct vibration-based condition monitoring of gear tooth wear based on a run-to-failure test. Modulation Signal Bispectrum (MSB) allows an accurate diagnostic feature to be extracted due to its significant performances of noise suppression and sparse representation. Based on the run-to-failure data, it shows that MSB based monitoring trend shows better trending of the fault progression in terms of both diagnostic information completeness and computing efficiency, compared with well-known time synchronous averaging (TSA) based approaches.