Study On Fault Diagnosis Technique To Reciprocating Membrane Pump Based On The Analysis Of Fault Mechanism

With the development of industry and science, the modern machine is more and more complex, higher and higer efficient. So the reliability and the safty of the machinery have been more and more important. Now the research on failure machinery and diagnosis technology of the modern complex equipment have become the hot spots of modern diagnosis technology. The article studies the fault technique to the reciprocating membrane pump based on the analysis of fault mechanism.After the discussion of the vibration mechanism and the types of the faults, the article focuses the wear fault. The wear fault can be looked as the clearance between the pairs. Different kinds of dynamatic models are set up according to the separation-contact state contact force model. In the models, many working situations are considered,such as the impluse of the fluid, the linkage of different compents, the intercoupling of different faults. Based on the simulation numerical solutions, the relations between the dynamatic response and different kinds of the pump's working situtations are researched. All these are beneficial to the realization of the pump's falut mechanism.Modern nonlinear theories such as wavelet, fractal and chaos theories are used in the status analysis of the reciprocating membrane pump.According to the excellent characteristics of wavelet transformation to micro signal in time domain and frequency domain and the ability to catch faint signals, based on the factors of relative energy and peak value, wavelet multi-resolution analysis through db4 wavelet is useful to the diagnosis of the membrane pump.Considering the fuild's impulse, the status of the reciprocating membrane pump' wear fault are all in chaos, which is proved by Poincare sections and the largest Lyapunov exponent. But without the wearing clearance, the status of the pump is quasiperiod. Because the Poincare sections are different concentrated point set, the attractors' structures are diffent with different status. The Lyapunov exponents of different working status and different faults are calculated. It is found that the Lyapunov exponent of different fault varies in the same order of magnitude with different working status. This is showed that the Lyapunove exponent can be used to express different fault of the pump.The chaos of nolinear system has another characteristic, which is the fractal with self-similarity and graduation invariability. The fractal's main parameter is correlation dimensions. The correlation dimensions of different faults and different working status are also calculated. According to the research, the correlation dimensions of the same working status are different, so they can also be used to discribe different fault of the pump. The correlation dimensons are also vary in some degree with different pressure, different rotate speed, different clearance, thus they can be used to monitor the working status. With further study, the article used wavelet fractal technology to study the signals of different clearance. According to the correlation dimensions of different frequency segments which are achieved by wavelet packet decomposition and reconstruction technique, the messages of different situation vibration can be caught.In the study, it is also found that the value of the chaos characteristic parameters can reflect some messages of the system, such as the degree of nonlinear states, the complexion of the states, and the energy dissipation of the system.The article explores the method of status monitoring and fault diagnosis of reciprocating membrane pump based on the anlysis of fault mechanism and its dynamtic response. As an attemption, it pioneers a useful directon for further studies on diagnosis of modern complex or reciprocating machines.