Assembly Unit Partitioning Method of Complex Mechanical Products for Error Propagation

Aiming at the difficulty of directly modeling and analyzing assembly errors of complex mechanical products, an assembly unit partitioning method of complex mechanical products for error propagation is proposed. According to the product assembly process information, the assembly relationships among parts and the measurement relationships are taken into account, which obviously affect product assembly quality, then a part-level association network based on complex network is constructed. The method combining fuzzy analytic hierarchy process (FAHP) and expert scoring is introduced to weight the association relationship among parts to obtain a part-level association weighted network. Compared with the traditional method, the objectivity of the association relationship strength is heightened to a certain extent. An improved link communities (LC) algorithm is used for assembly unit partitioning based on the community characteristics of the part-level network. The division of assembly units reveals the hierarchical nature of the parts network and simultaneously digs out overlapping parts belonging to different assembly units, which more accurately coincides with the characteristics of complex product structures. A low-pressure module of a steam turbine is taken as an example to verify the feasibility and effectiveness of this method. The assembly unit in this method can be further extended to an assembly feature-level error propagation network to provide a relatively complete analysis platform for the assembly quality analysis of complex mechanical products. © 2021, Editorial Office of Journal of Xi'an Jiaotong University. All right reserved.