Reduction of Tolerance Stack-up by Grouped Random Assembly for Components with Uniform Distributions

Tolerances of components can accumulate to result in quality variations when these components are assembled. One way to reduce the assembled variation is to tighten tolerance specifications of each component that, however, increases product cost. This paper investigates the effect of grouping for components with uniform and normal distributions by the developed “grouped random assembly” method. It is a method that first sorts and divides components into several groups and then assemble each group with corresponding group in order to reducing assembly tolerance. Distribution of resultant dimension based on the “grouped random assembly” approach is then analyzed. The results showed that, without changing components’ tolerance specifications, assembly tolerance depends on the number of grouping. Tolerance stack-up can be dramatically reduced with a suitable grouping strategy. The merit of this research is to develop a theoretical foundation for the grouped random assembly method. The results lead to a design for assembly strategy that assembly tolerance can be effectively reduced without tightening components’ tolerances.