Fault diagnose of an engine's main bearing based on time-frequency analysis and order tracking

In an engine's main bearing fault diagnose, the vibration signal is nonstationary and the law of frequency changing with rotating speed is not clear. So the time-frequency analysis and order tracking were put forward to diagnose its fault. Under acceleration and 1800r/min uniform speed conditions, after setting the bearing clearance of 0.08, 0.2 and 0.4mm, the vibration signal's time-frequency distribution and order were analyzed combined with speed and oil pressure signals, the 108 order component and 2600~3100Hz frequency components were found in the vibration signal. Under acceleration condition, with increase in the main bearing clearance, the 108 order component rose and the 2600~3100Hz frequency components weakened. After getting parameters of these two parts of signal components under acceleration condition, it was found that the parameters for the main bearing's normal, gentle and serious wear conditions distribute near 0.0027, 0.0100 and 0.0350, respectively, the parameters extracted can efficiently distinguish between these conditions.