A new paradigm for integrated bi-objective robust-tolerance design modeling and optimization

Robust design (RD) is an enhanced process/product design methodology for determining the best settings of control factors while minimizing process bias and variability. RD directly reflects the design of a bi-objective to minimize process bias and variability of a quality characteristic. Tolerance design (TD) is aimed at determining the best tolerance limits that will minimize the total cost incurred by both the customer and the manufacturer by balancing a quality loss due to variations in product performance and the cost of controlling these variations. Although RD and TD have received much attention from researchers and practitioners for more than two decades, and a number of methodologies for modeling and optimizing the RD and TD process have been studied, there is ample room for improvement. The majority of existing research considers RD and TD as separate research fields.; The primary objective of this research is to integrate joint optimization procedures for RD and TD. First, existing RD methodologies are evaluated, and new RD methodologies are proposed. Second, RD optimization is considered as a bi-objective problem associated with the quality attributes (process bias and variability) of a quality characteristic and the weighted-Tchebycheff method is used to generate efficient solutions. Third, the integration of the Lambert W function is applied to a TD optimization problem using quality and empirical loss functions with the assessment of costs incurred by both the customer and manufacturer. Fourth, an integrated model is presented that considers a detailed treatment of RD and TD using a single sequential optimization model in order to arrive at an optimal trade-off between quality and costs. Fifth, the TD optimization problem is considered as a bi-objective optimization problem with the consideration of the tradeoff between a quality loss and manufacturing cost in order to provide Pareto solutions of product/process tolerances. Sixth, this research develops an integrated bi-objective RD and TD methodology by considering two quality attributes and the tradeoff between quality loss and manufacturing cost. Finally, comparison studies are performed for verification purposes, and further studies are discussed.