Optimization of needle roller bearing design using novel hybrid methods

A long fatigue life is one of the most important criteria in the optimum design of needle roller bearings (NRBs). In the present work the dynamic capacity of the bearing is optimized. This non-linear optimization formulation has been solved using the artificial bee colony algorithm (ABCA), differential search algorithm (DSA), grid search method (GSM) and hybrid method (HM, a novel approach of combination of the ABCA/DSA and GSM). A total of four design variables corresponding to bearing geometry are considered, which include the roller diameter, roller length, pitch diameter and number of rollers. Apart from these three more constraint parameters have been optimized. The constraint violation study is carried out to prioritize the constraints. The sensitivity analysis is carried out to see the effect of tolerance on design variables on the dynamic capacity. The dynamic capacity of optimized bearings is found to have an upper hand from those specified in bearing catalogs.