Optimization of hydrodynamic journal bearing: MOGA approach

Abstract

The design and optimization of hydrodynamic journal bearing is most required because of their usage supports machineries which rotate at high speeds such as compressors and turbines. It is difficult to optimize the bearing parameters using conventional algorithms as they require non-linear optimization with constraints. Genetic Algorithms (GAs) is a general purpose algorithm that could achieve the process parameters with all the available constraints. In this work, the geometric parameters and their range such as variance ratio (0 - 1), journal radius (25 - 65), radial clearance (30 - 60), dynamic viscosity (0.026 - 0.046), surface pattern parameters (1/6) and surface roughness parameter (15 - 30) were used as an input for training the ANN model and to evaluate the performance parameters that would result in optimal value of minimum fluid film thickness, frictional torque and critical journal mass of journal bearing. This has been separately carried out for three bearings having transverse type of roughness patterns using the Multi Objective Genetic algorithm (MOGA) approach. Using Pareto optimal concept, the optimization of design parameters has been evaluated. The designed model using MOGA Approach exhibits the satisfying response compared to experimented data for the roughness pattern.