SUMMARY

Advancements in manufacturing technology significantly impact the design process. The ability to manufacture assembly components with specific tolerances has increased the need for tolerance allocation. Tolerance allocation has significant impact on both the needs of the designer and the manufacturer. Engineering designers are concerned with the impact of tolerances on the uncertainty of the output, while manufacturers are more concerned with the cost of fitting the parts. Those factors that affect the manufacturing cost are ‘fuzzy’ and cannot be evaluated numerically. Traditional tolerance control methods do not take into account both these needs. This thesis proposes a framework that overcomes the drawbacks of the traditional tolerance control methods, and reduces subjectivity via fuzzy set theory and decision support systems (DSS). Fuzzy comprehensive evaluation (FCE) is utilized in this thesis as a method of quantifying the fuzzy factors, but there is a need for decision support in the evaluation of importance of each factor. The combination of FCE and DSS, in the form of Conjoint Analysis (CA), is used to reduce subjectivity in the manufacturing cost. Taguchi’s quality loss function is considered in this framework to improve the robustness of the output. The application of the framework is demonstrated with three practical engineering applications. Tolerances are allocated for assemblies of a friction clutch, an accumulator O-ring seal and a Power Generating Shock Absorber (PGSA) using the proposed framework. Each example is validated for improved robustness in performance. The use of CA is also validated by checking the efficiency of final tolerances with and without use of CA.