The analysis of effects of shaft surface porosity on journal bearing using experimental and neural network approach

Purpose - The purpose of this paper is to investigate pressure distribution of the journal bearings with aluminium shafts with varying surface porosity in varying revolutions using experimental and neural network approach. Design/methodology/approach - The collected experimental data such as pressure variations is employed as training and testing data for an artificial neural network (ANN). Back propagation algorithm is used to update the weight of the network during the training. Findings - Neural network predictor has superior performance for modelling journal bearing systems with shafts of different surface porosities. Research limitations/implications - Back propagation algorithm is used training algorithm for proposed neural networks. Various training algorithms can be used to train proposed network. The spectrum of the journal surface porosity can be enlarged. Practical implications - From the experimental and simulation results, neural network exactly follows the experimental results. Because of that, this kind of neural network predictors can be applied on journal bearing systems in practice applications. Originality/value - This paper discusses a new modelling scheme known as ANNs. A neural network predictor has been employed to analyze of the effects of shaft surface porosity in hydrodynamic lubrication of journal bearing.