Research On Friction Performance And Prediction Of Dynamic Friction Coefficient Of Copper-graphite Embedded Spherical Plain Bearings

Self-lubricating spherical plain bearings have the characteristics of self-alighting,impact resistance,maintenance free or less maintenance,etc.,and are widely used in various construction machinery,water conservancy and hydropower facilities and military machinery.Among them,spherical plain bearings in the field of nuclear power have long been in service in the environment with high temperature and irradiation.In this paper,a 40Cr/ copper-inlaid graphite type self-lubricating spherical plain bearing developed for nuclear hydraulic damper in an ocean working condition is tested.The main research contents are as follows:(1)Firstly,in order to truly simulate the influence of the hydraulic damper’s longterm intermittent swing condition on the spherical plain bearings,the whole life intermittent friction and wear test of the bearing was carried out under the load of 6700 N,the swing Angle of ±5° and the swing frequency of 1 Hz by using the independently developed spherical plain bearings load simulator.Using collected during a test analysis of the friction temperature,abrasion and friction torque of spherical plain bearing tribology performance change,finally by means of scanning electron microscopy(SEM)and energy spectrometer(EDS)was carried out on the friction surface microstructure analysis and detection feature elements together to explore the wear mechanism of self-lubricating spherical plain bearing and failure reasons.(2)Secondly,in order to timely judge the service state of the self-lubricating spherical plain bearing for maintenance and avoid unnecessary losses,it is necessary to predict the short-term change trend of the friction performance of the spherical plain bearing.In view of the discrete data of dynamic friction coefficient recorded in the test process,this paper uses a stage robust method based on the improved Huber M estimation method to conduct robust processing on the data,and obtains robust friction coefficient data,which provides reliable and accurate data for the next prediction.(3)Finally,the mutual information method and Cao’s method were used to calculate the parameters needed for phase space reconstruction.The maximum Lyapunov exponent was calculated by the small-data method and the time series of the friction coefficient of the spherical plain bearing was determined to be chaotic.Two prediction models of chaotic time series were built by using the obtained delay time and embedding dimension,and the prediction of the last 30 steps(representing the time of 4h)was carried out.The maximum relative error of the prediction results was15.49%,and the established prediction model met the demand of actual engineering prediction.