Publication

Abstract : Roller bearings are used to support the driveshaft of a rotating machinery and are suitable for handling the radial load. In this paper, multi objective design optimization of roller bearing configurations namely Spherical Roller Bearing and Needle Roller Bearing have been carried out. The two optimization objectives are chosen in this study to enhance the working life of the bearings. The objective functions considered are Dynamic Capacity (DC) and Elasto-Hydrodynamic Lubrication Minimum Film Thickness (EHLMFT) of the bearing. These two objectives are conflicting in a real-life scenario and the trade-off between the two objectives has been obtained using the Multi-Objective Optimization approaches. Traditional weighted sum and pareto optimal methods are implemented in this study to obtain optimal design parameters. Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) algorithms are used to arrive at optimum solutions for the problem formulated using the Weighted Sum Method (WSM). Two nature-inspired meta-heuristic approaches namely NSGA II and Multi-Objective Feasibility Enhanced PSO (MOFEPSO) are used in this study to obtain non-dominated solutions. The results show that MOFEPSO efficiently solves the problems of both Spherical and Needle roller bearing design.