Modeling and diagnosis of dimensional variation for assembly systems with compliant parts

Sheet metal assembly is widely used in the fabrication of automotive body structures, aircraft fuselages, office furniture and home appliances. One of the most important challenges for sheet metal assembly is the dimensional variation. Dimensional variation can stem from both the design and manufacture of a product and may affect the final product functionality and process performance. Therefore, dimensional variation reduction in assembly processes is necessary to improve the final product quality.; The purpose of this research is to develop a variation propagation model for multi-station assembly system with compliant sheet metal parts. The model is used to study how the variation propagates from the components and tooling to the final product. In addition, a diagnostic tool to isolate multiple fixture fault occurrences for compliant parts is developed. The proposed methodologies are applied to automotive body assembly.; The main contribution of this dissertation can be summarized as: (1)  Multi-station sheet metal assembly model or compliant parts. A methodology for variation propagation analysis in compliant sheet metal assembly is presented. The model uses a state space representation and extended the method of influence coefficients to a multi-station system. The multi-station model estimates how variation propagates from the components and tooling to the final product during the assembly process. (2) Sheet metal assembly modeling using geometric covariance. A new method for variation analysis was developed using the components covariance matrix. The method, called “variation vectors”, replaces the method of influence coefficients. The method combines principal component analysis (PCA) with finite element methods to estimate the effect of components variation on assembly variation. This methodology can significantly reduce the computation effort required for variation analysis in sheet metal assembly. (3) Fixture location impact on dimensional variation for sheet metal assembly. The proposed methodology focuses on the impact of fixture position on the dimensional quality of sheet metal assemblies, considering part and tooling variation and assembly springback. An optimization algorithm is presented that combines finite element analysis and nonlinear programming methods to find the optimal fixture position such that the assembly variation is minimized. (4) Multiple fixture fault diagnosis considering a N-2-1 locating scheme. Using designated component analysis, it is possible to successfully isolate multiple fixtures faults in compliant sheet metal locating systems. DCA extracts the significance contribution of specific designated pattern to the total variation of the production measurement data.