Pareto approach in multi-objective optimal design of single-row cylindrical rolling bearings

Abstract

Even if it seems simple, the design of the rolling bearings is a difficult task for mechanical engineering. Their specific structure and the involved tribological phenomena in their operation lead to many and complex calculations of rolling bearings. Consequently, to find the optimal design of the rolling bearings the Multi-Objective Optimization should be used. In this paper, two objective functions (the basic dynamic radial load rating of the bearing and the minimum thickness of the elastohydrodynamic film between rollers and raceways) are optimized using multi-objective Evolutionary Algorithms (EAs) in order to obtain the optimal design of a single-row cylindrical roller bearing. Three design variables (the complex factor c, the roller diameter, and the roller length) were used to fully, completely, and uniquely describe the bearing geometry and to construct the objective functions and the constraints. To solve this optimal design problem the authors used Cambrian optimization platform.