Abstract:
Objective of this research is to study how to use acoustic emission in condition monitoring of spur gear defect diagnosis which was calculated by using the statistical parameter analysis technique. Acoustic emission (AE) is defined as defect propagation rate in the microscopic and macroscopic of rotating machinery which has spur gear as the main part of power transmission system. This experiment was set up based on motorcycle spur gear with torque loading in 6 conditions such as defect free, a 25% crack tooth, a 50% crack tooth, a 50% broken tooth, a 100% broken tooth, one gear has a 50% broken tooth and a 100% broken tooth. AE sensors were placed on 3 position of the bearing casing of the wheel shaft and the shaft axis was the best signal. Used 40 sets for AE signal analysis obtained from Energy, Root mean square, Crest factor, Mean, Kurtosis, Variance and Skewness and AE analysis with two statistical parameters amount 5 series between Kurtosis and Root mean square, Kurtosis and Energy, Variance and Mean, Variance and Root mean square, Energy and Root mean square for plot boundary conditions and Accuracy test of boundary used 10 sets in other groups. This experiment proved that defect could not be clearly identified by using acoustic emission signal analysis with time-frequency domain and one statistical parameter, whereas, conducting analysis with two statistical parameters between Kurtosis and Root mean square can be evidently used for defect identification.