Research On Property Modeling Method Of Self-lubricating Composites And Simulation Of Wear Process

Pantograph slider is an important part of locomotive on high-speed railway.And the abrasion between the wire and the pantograph results in significant loss and safety risk.The mechanical and physical properties of composites have important effects on wear performance.Besides,the properties of composites depend on the development of materials recipe,which is guided by the relative theoretical models.In order to improve the wear performance of slider,amounts of researches concentrate on recipes of composites.However,the study on property models and dynamic simulation of wear process should be concerned more about.In this dissertation,with self-lubricating composites as the study object,the Gene Expression Programming(GEP)and Cellular Automaton Model are used to establish models between ratios of recipe and properties of composites,analyse covariant properties,optimize recipe of composites and simulate wear process.The main research contents and achievements are as follows:(1)Composites are tested and analyzed to design a kind of conductive,wear-resistant self-lubricating composites for pantograph slider.The 1:0.8 graphite/pitch material is chosen as matrix and adhesive after comparing properties of graphite/pitch,graphite/polyvinyl butyral(PVB)and graphite/polyvinyl formal acetal(PVFE)composites,from 1:1 to 1:0.3.Then carbon fiber/wollastonite(2:1)and copper powder are added to enhance the performance of the materials.Finally,basic data of resistivity,bending strength,compressive strength,wear amount,hardness and density are collected by performance tests for modeling study.Compared with matrix composites,properties of composites have been significantly enhanced.The bending strength increases from 1.13-6.20 MPa to 6.98-13.32 MPa,the compressive strength increases from 5.6-8.4 MPa to 8.26-15.49 MPa and the average wear amount decreases from12.67 mg to 5.9 mg.(2)To select basic models for modeling the properties of composites is difficult.In order to solve this problem,the concept of model search space is proposed in this study.The concept and GEP method are verified with polynomial model and Sunspot problem.The method is also valid in finding empirical models of mechanics problem.With the basic data of self-lubricating composites,model search spaces are optimized by choosing the basic functions.Then,models of resistivity,bending strength,compressive strength,wear amount,hardness and density have been founded for recipe optimization.Results show that all the R~2 are greater than 0.7,the maximum absolute error is 2.4,and the average relative error is less than 5.58%.Compared with polynomial models,those models have strong robustness,stable generalization performance and restrict the Runge phenomenon.(3)As covariant properties of self-lubricating composites have similar basic functions and surfaces,this dissertation presents a method of recognizing covariant relation among properties.It is proved that resistivity of composites varies exponentially with compressive strength.Then,the covariant model of wear amount,bending strength and hardness is founded with this method.Besides,properties of composites are mutually restricted,a comprehensive optimization model is established to cope with this problem.And the optimum proportion of recipe is obtained by solving this model.The performance of the optimum sample is verified by experiment,and properties of composites have been balanced.(4)An improved multi-time scale cellular automata model is established to simulate friction process of graphite-asphalt composites.The enhanced tridimensional cellular automaton model firstly integrate the interfacial wear process,thermal process and morphological process as a whole.The model overcomes the multi-time scale problem of different processes.It realizes the real-time calculation of friction coefficient of lubrication structure,the dynamic updating of topography of friction interface,as well as the simulation of related friction quantity,which is conducive to further research of self-lubricating composites.As pore defects commonly exist in composites,a cellular automaton with defective modules has been established to simulate the friction process of defective lubrication structure.Results show that pore defects have significantly raised COF and the local temperature has been raised by 20percent.