Dimensional variation analysis and optimal process design for non-rigid sheet metal assemblies.

摘要

Non-rigid sheet metal assembly is widely used in manufacturing industries, such as aerospace and automotive industries. Improving product quality and reducing the cost are main concerned issues for a manufacturing company to achieve higher product competitiveness in current global market. The dimensional quality of a non-rigid sheet metal assembly is a crucial and yet challenging quality indicator due to the non-rigidity of the sheet metal components. Although the product dimensional variation analysis and process design for rigid assembly have been studied for many years, such study for non-rigid assemblies is emerging, and also challenging. There remain many unrecognized and/or unsolved issues in the study of non-rigid assemblies. This thesis presents a number of new, systematical, and generally applicable methods for analyzing and minimizing the non-rigid sheet metal assembly variations.; Firstly, a novel fractal-based method for sheet metal assembly variation analysis is developed to deal with the fractal variations of parts (i.e., component of an assembly). The surface microstructure of part variation is modeled by fractal geometry and its influence on the final assembly variation is studied by modeling the sheet metal assembly process.; Next, a new methodology based on wavelet transform is proposed for analyzing the contribution of variation components with various scales to the final assembly dimensional variation, considering possible sources of variation from both parts and the assembly process. It is more general and advantageous than the approach based on the fractal geometry. The integrated procedure of wavelet transform and Finite Element Method (FEM) for non-rigid assembly variation analysis is developed and implemented. Its effectiveness is demonstrated via an application example.; Thirdly, a simultaneous optimization method for fixture layout and joint positions is developed. The optimization variables from both the product design (assembly joint positions) and the production plan (the fixture layout) are included in the mathematical model. The mode-pursuing sampling method (MPS) is modified and employed to search for the global optimal solution.; Finally, the elastic contact phenomenon in the sheet metal assembly process is studied. A non-linear assembly dimensional variation analysis method is developed by establishing the elastic contact model between the assembly surfaces. The assembly dimensional variation analysis with and without contact modeling is respectively conducted. The corresponding physical experiments are also carried out and used to validate the contact FEM models.; The work enables us to gain more in-depth understanding on the characteristics of the non-rigid sheet metal assembly dimensional variation. It provides not only the fundamental analysis and modeling methodologies, but also the corresponding software tools that can be easily integrated with most current general-purpose commercial FEA packages (such as ANSYS and CATIA). The developed approaches, technologies and tools presented in this thesis can benefit both the academic research and industrial applications on the design and manufacturing of non-rigid sheet metal assemblies.