Investigation of Static and Dynamic Performance Characteristics of Threelobe Gas Lubricated Journal Bearings, Using Artificial Neural Networks

Due to vast use of gas lubricated bearings in various industries and their advantages in specific applications, they have been investigated by many researchers. However, analytical treatments of gas lubrication are tedious due to high non-linearity of the pressure equation as the consequence of lubricant compressibility. Hence, in this paper a feed-forward neural network is employed to investigate the performance of three-lobe gas lubricated bearings. It is believed that neural network can easily compete with theoretical model in predicting the solution of lubrication problems. The performance parameters considered are stability margin, power loss, and attitude angle for various values of bearing compressibility numbers, mount and tilt angles. The results of the neural network are compared to theoretical model (FEM) and it is observed that they are in good agreements. The results also indicate that the above parameters can influence the performance of the bearings.