Hydrodynamic Analysis and Performance Prediction of Herringbone Grooved Thrust Bearing for Locomotive Turbocharger

Abstract

In this study, herringbone texture over the thrust bearing (TB) has been proposed and optimized with an AI technique to improve the power loss and load-carrying capacity (LC). The herringbone-grooved thrust bearing (HGB) has been designed to withstand a significant amount of thrust load generated due to the pressure difference between the turbine and compressor wheel. The non-linear Reynolds equation for HGB has been formulated and solved numerically with the FDM and SOR algorithm for pressure profile, film thickness, LC, and power loss. The influence of different texture parameters such as helix angle, angular groove width, number of grooves, speed, and fluid film thickness on static characteristics over the bearing surface has been investigated. An artificial neural network (ANN) is used to train the distinctive datasets obtained from the numerical analysis, and its performance is assessed by computing the root mean square error (RMSE) and regression coefficient (R2). Following ANN analysis, an ANFIS surface plot is produced to identify the optimum value of the bearing characteristics for which it has the highest LC and minimum power loss. The results show that the well-designed HGB has the potential to improve the load-carrying capacity. The outcome of this paper highlights the importance of numerical methodology, parametric analysis, and the applicability of artificial intelligence networks for the design of herringbone grooved thrust bearings.