Abstract:It is very difficult that to measure the working parameters reflecting end face friction state (such as end face opening time, film thickness, etc.) of the mechanical liquid seal when it in the operation. As the Acoustic Emission(AE) signals generated by seal end face have the characteristics of time varying nonlinear and strong impulsiveness, the method of mechanical fluid seal end face friction condition monitoring was proposed based on the Acoustic Emission signals,and the Empirical Mode Decomposition (EMD) was introduced into the mechanical fluid seal work condition monitoring analysis. EMD is equivalent to a set of adaptive filter. The signal can be decomposed into a series of independent characteristics with different time scales, different frequency band of the Intrinsic Mode Function (IMF). And then according to the characteristics of energy distribution to eliminate false components, got the proximal source acoustic emission signals. It’s features was extracted and the mechanical liquid seal end face contact state was identified accurately using Laplace wavelet correlation coefficient method. By mechanical oil seal testing proved that this method could accurately identify the contact state of dynamic and static rings for the mechanical fluid seal, and it could be used widely in industrial field.