Gear Crack Detection Based on Vibration Analysis Techniques and Statistical Process Control Charts (SPCC)

Vibration condition monitoring is a non-devastating technique that can be performed to detect tooth cracks propagating in gear systems. This paper proposes apply new methodology using time-domain analysis, frequency-domain and statistical process control charts (SPCC) for crack detection of 10 DOF dynamic model spiral bevel system (SBGS). The mesh stiffness effect used the has been studied analytically different levels faults. Adding Gaussian white noise discussed as first step simulating initial modeling signals real-world conditions. Second, signal analysis was identify periodic vibration pulses failure components calculate standard deviation (STD) feature fault-sensitive feature. Third, fast Fourier transform (FFT) time variable applied determine frequency sidebands cracks. Fourth, SPCC designed Shewhart X-bar chart an exponentially weighted moving average (EWMA) based on STD healthy gears. Finally, testing stage, are carried out with simulation under faulty conditions results showed EWMA outperformed domain detecting all at early stage.