Tolerance synthesis for complex assemblies is formulated as a probabilistic optimization problem. A main challenge in statistical tolerance synthesis for complex assembly system design is the computation intensity in estimating the conformity probability of Key Product Characteristics (KPCs) or yield of assembled products. In assemblies with multiple KPCs, the yield can only be obtained numerically through simulation techniques such as Monte Carlo (MC) algorithm. The existing tolerance synthesis methods require a large number of yield assessments in optimization. A new approach is developed for yield model approximation based on computer experiment, multivariate distribution transformation (MDT) and regression analysis. An explicit yield function can thus be approximated. Therefore, the widely used gradient-based approaches (such as Sequential Quadratic Programming) can be applied and the intensive computation in direct optimization can be avoided. An industrial case study is presented to illustrate and validate the proposed methodology and compared with the existed tolerance synthesis methods.

This content is only available via PDF.
You do not currently have access to this content.