Characterisation of acoustic emissions for the frictional effect in engines using wavelets based multi-resolution analysis

The friction between piston ring-cylinder liner is a major cause of energy losses in internal combustion engines. However, no experimental method is available to measure and analyze the fictional behavior. This paper focuses on the investigation of using acoustic emission (AE) to characterize the friction online. To separate the effect relating to friction sources, wavelets multi-resolution analysis is used to suppress interfering AE events due to valve impacts and combustion progress. Then a wavelet envelope indicator is developed to highlight AE contents from friction induced AE contents. The results show that the AE contents in the middle strokes correlate closely with viscous friction process as their amplitudes exhibit a continuous profile similar to piston speed. Furthermore, the AE envelope indicator proposed can distinguish the differences between two types of lubrication oils, showing superior performance of AE based online lubrication diagnosis.