Optimizing the design of mechanical components for reliability and cost

This paper discusses a design optimization algorithm that uses probabilistic methods to predict product performance, reliability and cost. By considering variations in the design parameters, the influence of uncertainty can be determined and the design can be optimized to prevent early failures while minimizing. manufacturing cost. This paper discusses how the sensitivities of the design variables to the statistical distribution parameters are determined. The sensitivities are then used in an optimization model to determine the best combination of design parameters and manufacturing tolerances. A demonstration problem that illustrates the application of the method is provided.