A fault diagnosis method for loose slipper failure of piston pump in construction machinery under changing load

Abstract

Piston pump is the power source and key component of hydraulic system in construction machinery. Its working status and service performance directly affect the construction quality and operation safety of the construction machinery. However, the existing fault diagnosis methods for piston pump still have some shortcomings, for example, signal processing method do not fully consider the coupling effects on the piston pump between health status and external load, and artificial intelligence-based method require a large amount of samples under different health conditions and different loads. Furthermore, the above two methods lack analysis of the dynamic characteristics of piston pump. In order to solve the above problems, based on the virtual prototype model, the dynamic response of piston pump under different health conditions and different loads is simulated, analyzed and tested. The results show that both external load and structural health condition take effect on the dynamic response of the piston pump. It is also found the gradients of the trend lines of the root mean square (RMS) of axial vibration is a sensitive feature reflecting the failure of loose slipper under changing load. So a fault diagnosis method for loose slipper failure of piston pump under changing load is proposed. The proposed method first collects the axial vibration signal of the piston pump, then divides the vibration signal into equal parts, then calculates the RMS value of each segment signal and calculates the gradient of the RMS trend line, and finally the loose slipper fault is detected based on the gradient of the trend line. The effectiveness of this method is verified through experiments.