Purpose – To discuss the effects of metal matrix composite (MMC) journal structure on the pressure distribution and, consequently, on the load‐carrying capacity of the bearing are predicted using feed forward architecture of neurons. Design/methodology/approach – The inputs to the networks are the collection of experimental data. These data are used to train the network using the Batch Back‐prop, Online Back‐prop and Quickprop algorithms. Findings – The neural network (NN) model outperforms the available experimental model in predicting the pressure as well as the load‐carrying capacity. Research limitations/implications – The experiment specimens used in this study have been made of MMC with aluminum based reinforced with SiC ceramic particles, using the stir casting technique. Various composite journal structures can be investigated. Practical implications – The simulation results suggest that the neural predictor would be used as a predictor for possible experimental applications on modelling journal bearing system. Originality/value – This paper discusses a new modelling scheme known as artificial NNs. An experimental and a NN approach have been employed for analysing MMC journal structure for hydrodynamic journal bearings and their effects on the system performance.
Industrial Lubrication and Tribology – Emerald Publishing
Published: Mar 1, 2006
Keywords: Neural nets; Metals; Composite materials; Algorithmic languages