Purpose – The purpose of this paper is to investigate and discuss the influence of the pattern, size and orientation of textures on journal bearing load carriage capacity. An important development in load carriage capacity of journal bearings can be obtained by forming regular surface structure in the form of threaded on their shaft surfaces. This is performed both theoretically and experimentally using shafts with textured (threaded) and untextured surfaces. Each screw thread can serve either as a micro‐hydrodynamic bearing in cases of full or mixed lubrication or as a micro reservoir for lubricant in cases of starved lubrication conditions. Design/methodology/approach – The pressure distribution and the load‐carrying capacity are predicted using feed forward architecture of neurons. The inputs to the neurons are a collection of experimental data. These data are used to train the network using the delta‐bar‐delta, batch‐backprop, backprop, and backprop‐rand algorithms. The proposed neural model outperforms the available experimental system in predicting the pressure as well as load‐carrying capacity. Findings – Theoretical models are developed using a neural network approach, and tests are performed, to investigate the potential of threaded textured surfaces in tribological components like mechanical seals, piston rings and journal bearings. In these tests, load carriage capacity is significantly increased with threaded textured shaft surfaces to the shafts with non‐textured surfaces. Originality/value – The paper discusses a new modelling scheme known as artificial neural networks. A neural network predictor has been employed to analyze the effects of shaft surface profiles in hydrodynamic lubrication of journal bearing.
Industrial Lubrication and Tribology – Emerald Publishing
Published: Aug 14, 2009
Keywords: Surface texture; Bearings; Neural nets; Load capacity