Robust design of ball bearings for an improved performance using genetic algorithm

Abstract

The design of rolling bearings must be optimized to get an improved fatigue life. As the fatigue life is directly affected by the dynamic capacity, therefore the later is considered as an objective function, which is to be maximized using a robust design optimization. Such design procedures enhance the qualities and reliability of products by ensuring that they are insensitive to the source of variabilities or uncertainties present in them, without eliminating these sources and at the same time maintaining the design feasibility with probabilistic constraints. In the present work, deep groove ball bearings (DGBBs) are designed in such a way that it will have minimum variation in the objective function due to the variation in design variables. A nonlinear constrained optimization problem is formulated with one objective function, ten design variables and a set of total thirty-five constraints. The real-coded genetic algorithm (GA) is used as the optimization tool. The result obtained shows the optimized bearings have improved fatigue life. A convergence study is performed to ensure global optima in the design space and further a sensitivity analysis is performed. The analysis shows that the dynamic capacity is the most sensitive to the variation in the inner groove curvature radius and least sensitive to the bearing pitch diameter. This paper allows the designers to design the rolling bearings by optimizing the fatigue life along with the minimization of their variations, which are due to variations or having tolerances in input design variables during its manufacturing. Moreover, the proposed design methodology is not limited to the bearings only, but can also be enforced in the design of any machine components (like clutches, gears, cams, levers, springs, etc.) having substantial uncertainties in their geometry or material, such as the residual strain, modulus, thickness, density, etc.