Integrated parameter and tolerance design for multiple response optimization

Abstract

Robust parameter design and tolerance design are effective approaches for continuous product quality improvement. However, the traditional two-stage design approach of parameter design followed by tolerance design is time-consuming. The tolerance levels determined in the tolerance design stage typically affect quality loss; therefore, the parameter levels obtained in the parameter design may be suboptimal. In this paper, we propose an integrated parameter and tolerance design model, and then a criteria importance through intercriteria correlation-based grey relational analysis technique is adopted to obtain the optimal parameter and tolerance levels. A case study is conducted to validate the proposed method. Comparison results show that the parameter levels and tolerance levels of controllable factors obtained using the proposed method are more reliable and enable a better tradeoff among quality loss, tolerance cost and predicted loss.