Study On Technologies Of In-Line Electrostatic Abrasive Monitoring And Distinguishing In Oil System

The technology of abrasive monitoring in lubrication circuit can earlier detect component’s performance degradation and provide real time information for maintenance decisions, compared with vibration, temperature and other conventional methods. Therefore, the study on the abrasive monitoring in lubrication circuit has important significance. In order to improve the ability to distinguish particles of the abrasive monitoring in lubrication circuit, while to provide a theoretical basis for the abrasive monitoring experiments, and improve the design of electrostatic monitoring sensor, in this thesis basic research is carried out concentrating on abrasive monitoring in the lubrication circuit of the aviation engines and wind turbine gearbox, the main contents are as follows:(1) Theoretical research. Based on having improved the physical model of the sensor and the corresponding circuit, the mathematical model of the electrostatic abrasive monitoring system is created utilizing Coulomb’s law and Gauss theorem, which firstly and clearly shows the relationship between the charge particle carried and the output of the monitoring system using explicit mathematical expression and visually indicates the factors affecting the output of the system. Meanwhile, the concept of spatial sensitivity is explicit expressed using mathematical representation, and the factors influencing the spatial sensitivity are quantitatively simulated correctly and firstly.(2) Proposing the identification method for the abrasive electrostatic signal based on the invariant static discrete wavelet transform and kurtosis. To aimed at the noise of the power frequcensy and its harmonic components and random pulse, the comb filter is used to remove the noise of the power frequcensy and its harmonic components, and the method based on the invariant static discrete wavelet transform and kurtosis is proposed to remove the noise of random pulse, and to prove the identification of the induced signal. the validity and the necessity of which are validated through the simulation and the experimental data. And the result shows that the proposed method can improve the recognition ability of abrasive electrostatic signal, and that the technology of the debris electrostatic monitoring can distinguish the metal debris and non-metal debris.(3) Establishing the quantitative relationship between the voltage of the electrostatic sensor and the diameter of the debris. The experimental platform is designed to test the relationship between the output of system and the particle size, and accordingly the platform is used to obtain the experimental data, using which the relationship is established via the method of least–square curve fitting. Meanwhile, the analysis of relationship between the equivalent diameter and the output is conducted under the condition of different test load, test material and flow velocity, which shows that the larger the test load, the larger the output of the same equivalent diameter; that different materials have different fitted-curves; that for the same material and test load, the oil flow rate does not influence the relationship between the system output and the particle size.(4) The research on the induced signal is carried out in the circumstances of multiple particles with different charges along different radial positions simultaneously through the sensing area of the sensor, namely on the spatial resolution of the electrostatic sensor. Having found the outputs of the particle of the same size and charge under this circumstance through experimental studies, the spatial sensitivity is introduced and the analytical function is established. Then the research on the characters of the spatial induced field and of the spatial filtering of the sensor is carried out, and the quantitative relationship between the cutoff frequency of the electrostatic sensor and the peak time difference of the transfer function curve is created, thus the factors influencing the cutoff frequency of the electrostatic sensor are found out and simulated. Meanwhile, the experimental data are used to validate the simulation. On the basis of assuming the probability density functions of the equivalent charge and the equivalent radial position, the probability distribution function is created. According to the actual data obtained, a neighborhood containing the data is selected, and the probability is calculated in the neighborhood. If a greater probability is obtained, the data should be preserved and the probability of the occurrence of both the equivalent radial position in an interval and the equivalent charge in an interval is achieved according to the joint probability density function of the equivalent radial position and the equivalent charge, and thus the probabilistic representation of the spatial resolution of the sensor is obtained.(5) The research on the induced signal is carried out in the circumstance of multiple particles with different charges along different radial positions successively through the sensing area of the sensor,namely on the time resolution of the electrostatic sensor. According to the theoretical graphics of electrostatic abrasive induced signal, the calculative analysis on the axial length of the sensing area is carried out under the condition of different length and radial diameter of the probe, and shows the ratio of the result of the length of the sensing area minus the axial length of probe to the axial length of probe is fixed in the circumstance of different probe size and of the same ratio of length to diameter of the probe. Thus, the mathematic function between the three is established according to the theoretical data. The verification device is designed, and according the test data obtained using the device, the actual length of sensing area is calculated. The result shows the correctness of the relationship established above, and thus the maximum abrasive detection rate of the electrostatic monitoring sensor is defined.