Factors influence the lubrication characteristics investigation and optimization of bearing based on neural network

Abstract

Purpose This paper aims to give the guidance for the design of the bearing. Design/methodology/approach The finite element method, the multi-body dynamics method, the finite difference method and the tribology are combined to analyze the lubrication. Findings The performance parameters of crankshaft-bearing system such as the misalignment, the oil filling ratio and the oil groove are also investigated. Misalignment causes the pressure to incline on one side and the pressure increases obviously. Filling ratio has great relationship with pressure distribution; the factors influencing the filling ratio are also analyzed. Different oil groove models are investigated, as it can provide the theory for oil groove design, and three factors above are always combined to influence the lubrication characteristics. Originality/value The optimization of bearing system is conducted by orthogonal test and neural network, unlike the linear optimization theory. Neural network uses the nonlinear theory to optimize crankshaft-bearing system.