Experimental investigation of helical gear tooth crack location and depth detection using moving average method on transmission error

Abstract

Gear systems are the most useful and essential power transmission systems in the high-speed industry due to their accuracy. It is necessary to make sure that these systems work without defects such as tooth cracks. Therefore, detecting the location and depth of cracks in gear systems is very important. In this research, a new approach is proposed to detect the crack location, and accordingly, some statistical indicators are used to estimate the crack depth in the helical gear tooth. To this end, after explaining the helical gear mesh stiffness and tooth-root crack modeling, the helical gear pair dynamic is modeled. Then, the vibration data of a helical gear system is obtained by an experimental test rig, and the moving average method is undertaken to precisely detect the crack location. The crack depth ratio is estimated using the crest factor, impulse factor, clearance factor, and $S_r$ and $S_{\alpha }$ which are applied to the simulation results and the experimental signal. According to these results, the crest factor, impulse factor, and clearance factor calculated the crack depth ratio with a good agreement, and the indicators $S_r$ and $S_{\alpha }$ estimated it with a more significant error. Also, the average of estimated values is calculated, indicating a better result than each indicator alone.