The Lubrication Performance Analysis And Optimization Of Main Bearings In High-speed Gasoline Engine

Engine has high load, high speed, high temperature character ist ics in the working process, which d irectly led the cranksha ft- bearing s ystem to ge nerate vibrat ion and noise, ma in components of fat igue fracture, tile burning and cylinder scoring, etc. The lubricat ion perfor mance of the ma in bear ing not only affects the norma l operat ion of the engine, but also affects the economy, reliability and life of the who le machine. Recently, the coupling analys is of mult i- body dyna mics and Elasto- hydrodyna mic lubricat ion method was tradit io nally used by domest ic and overseas scholars to study the main bearing’s lubrication performance. This paper study on the lubrication perfor mance ana lys is and optimizat ion o f main bearings in high- speed gasoline engine by comb ining t he methods of numer ica l s imulat ion and art ific ia l inte lligence technology, its main research and innovative points are as follows:(1) The finite ele ment models was built for ma in bear ing syste m of t he high- speed gaso line engine, and then seat the modal reduct ion to obtain the reduced finite element models, the experimental results verify the validity of this models.(2) The mult i-body dyna mics model was built for cranksha ft-bearing syste m of the high- speed gasoline engine, and obtained the boundary cond it ions of t he model by the bench test. Then, the ma in bearing’s peak oil film pressure, minimum oil film thickness, fr ict io n power loss, lubr icat ion oil consumpt ion rate were obtained by simulat ion calc ulat ion. Through ana lys is of the results s hows that many ma in bearing’s lubr icatio n perfor mance are poor, and the posit ion o f bearing wear by simulat ing is consiste nt w ith w here the actual wear took place, this ver ify the correctness of t he simulat ion model. As the bearing lubr icat ion worst at 4000 rmp conditions, subsequent optimization is carried out according to this work condition.(3) Then, O ptima l Lat in H ypercube was used to conduct Des ign of Exper iment for 4 major bearing shells parameters which affect the lubr icatio n perfor mance of the ma in bearing, then, the approximat ion mode l was built wit h the met hod of Radia l Basis Funct ion neura l network. With the minimum total average fr ict ion power loss and the minimum total average lubricat ion oil consumpt ion rate as the optimizat ion objective functio n, w ith t he minimum film t hickness and the peak oil film pressure as the optimized boundary cond it ions, then conducted the mult i- object ive optimizat ion for approximate model by the met hod of Mult i- is la nd Genet ic Algor ithm, to some extent, it improved the lubrication performance of the main bearing.