A Research On Nonlinear And Multi-variable Lubrication System Based On Modeling Simulation And Experiment Method

A combination of experiment, modeling, algorithm research and computer simulation technique is used for the first time to analyze and design automatic chain lubrication system. The research results have provided its reference value for promotion and application of the system and its further structure optimization. The topic of the research is from a cooperative program with a university in UK. The object of research is automatic chain lubrication system. The research is conducted with two main objectives. Firstly, with ensured lubricating effects, the research aims at making the relation clear among system structure, types of lubrication, and lubricating effects, and providing users with a reference under the best working condition. The second purpose is to resolve the problem of system structure optimization under its best lubricating effects. The significance of the research lies in shortening the research and development period of products and promoting their efficiency of use and development. In addition, the improved lubrication system could also be applied to precision instrument which needs continuous lubrication, and mini or micro machine. Therefore, the research has theoretical significance as well as application value.Main researches and achievements:1. Experimental research: Experimental data of the system was acquired under different experimental working conditions, and main factors that influence the system performance were analyzed and discussed. The experimental research provides foundation for system modeling, algorithm research, and simulation analysis.2. Modeling analysis: Navier-Stocks equations were used to establish nonlinear partial differential equations of lubrication flow of the system, and finite volume method was employed to discretize the equations. A mathematical model and the boundary conditions of lubrication motion of the system were obtained. 3. Algorithm research: A proposal of using the method of computational fluid dynamics (CFD) and neural network to analyze the system was brought forward. The algorithms methods were compared, and the rationality of the system mathematical model established was verified.4. Simulation analysis: Based on the algorithm research and the experimental data, it was put forward to build the CFD, BP and RBP neural network system simulation models by using FLUENT and MATLAB. The models were used to simulate and predict the working characteristics of the system. The drippage speed of lubrication was acquired under different working conditions of the system. The query relation was established under the best lubrication condition. A reference was provided for system structure optimization, and the correctness of system models was confirmed through application in engineering.5. Effects and conclusion: By comparing the simulation effects of the CFD method and the neural network method, it reveals that the CFD simulation model could represent the system for lubrication flow by dynamic picture display, which is a vivid and real-time simulation. However, the speed of this kind of simulation is low. The advantage of the BP and RBP neural network simulation model lies in its high speed and accuracy, which makes up the weakness of CFD simulation model. Therefore, it could be concluded that a combination of these two models could improve the performance of simulation system. The research results are also valuable for theory studies on application and further improvement of the system.By investigation and research, it could be seen that the analysis of liquid flow in small-diameter pipeline, the methods employed and the research results are still at the forefront both domestically and internationally. At present, the research results have been implemented in manufacturing and improving the design of motorbike and bicycle chain lubrication system at a British manufacturer. The research has also provided basis for theory studies and simulation analysis on replacement of time-consttming and costly experiment equipment, and has made positive contributions to the shortening of products research and development period, the promotion of their usage efficiency, and the extending of their application.