Thermal Based Optimum Design of Tapered Roller Bearing Through Evolutionary Algorithm

In high-speed applications the maximum temperature in bearings are a crucial concern. In some applications the bearing is the prime source of heat, the temperature at which a bearing operates dictates the type and amount of lubricant and the material for the fabrication of the bearing components. In the present work a thermal based optimum design of tapered roller bearings has been presented. Internal geometry of the bearing has been optimized based by evolutionary algorithm. Constraints are geometrical, kinematical, strength and thermal in nature. Optimum designs have been found to have better performance parameters. Artificial bee colony algorithm has been used for the present optimization problem, for solving constrained non-linear optimization formulations. A total of nine design variables corresponding to the bearing geometry and constraint factors have been considered. A convergence study has been carried and optimum designs based on temperature is compared with the optimized values based on dynamic capacity, both using artificial bee colony algorithm. There is an excellent improvement found in the optimized bearing designs based on temperature when compared with the optimized results based on dynamic capacity in respect of the maximum temperature in the bearing with the artificial bee colony algorithm.