Least Cost Tolerance Allocation for Systems with Time-variant Deviations

This paper presents a methodology for the least cost tolerance allocation of systems with time-variant deviations, such as deformation, thermal expansion, mobility of parts and wear. By means of the approach, the product developer can identify the system's optimal tolerance design against the backdrop of the diverging requirements: wide tolerances to reduce costs vs. narrow tolerances to ensure functionality. Therefore, Particle Swarm Optimization – a global statistical optimization technique – is applied to the time-variant tolerance-optimization problem to determine the tolerance range as well as a certain mean shift for each non-ideal dimension.The practical use of the methodology is illustrated for a modified tolerance stack-up problem. We provide a comprehensive walkthrough of the methodology's application to the stack-up problem. Therefore, the essential mathematical background, the required functional dependencies and the considered parameters are detailed in a step-by-step procedure. So, we aim to establish the application of the statistical tolerance-cost-optimization in academia and industry and to motivate students and young graduates to apply this methodology by themselves.