Normalization of gearbox vibration signal for tracking tooth crack severity progression under time-varying operating conditions

It is a challenge to track gear tooth crack severity progression under time-varying operating conditions since time-varying operating conditions induce amplitude modulation (AM) and frequency modulation (FM) effects into gearbox vibration signals. The AM and FM effects induced by time-varying operating conditions can mask tooth crack signature, which results in difficulty in distinguishing between changes of tooth crack severity and changes in gearbox working conditions. The reason for this difficulty is that due to the AM and FM effects, changes in condition indicators (CIs) may indicate changes of tooth crack severity, changes in gearbox operating conditions, or both. To overcome this challenge, AM and FM effects induced by time-varying operating conditions need to be removed. For FM effect, it can be removed using order analysis. In this study, a novel normalization method is proposed to conduct removal of the AM effect. The effectiveness of the proposed normalization method for removing AM effect is verified using simulated gearbox vibration signals, which benefits tracking tooth crack severity progression under time-varying operating conditions.