The paper’s focus is on the module division dynamically in the virtual assembly sequencing process for the product family, which will affect the module variant quality seriously. Most research merely divides the module on their functions, without paying the attention on which the module division is beneficial for the latter assembly process. Considered assembly relationships as a key constraint, a new variant module division measure based on Extension Logic is presented. Since the tolerances existing everywhere in the assembly process, the connection analysis is employed to build a constraint function on the tolerance. Physical constraints contain mating and combination constraints. After the dynamical position for different parts, the mating constraint is defined as the movement opposition between two mechanical elements including face alignment mating, shaft and hole mating or face against mating. Then the extension radiation model could be generated. According the average value of the relationship function to realize the module division, the product structure is divided into several sub-assemblies. As the main problem in virtual assembly is process collisions, a modified integer coded Particle Swarm Optimization Algorithm is used. Here every part is mapped to corresponding particle. The result is performed in the virtual environment of software DELMIA. At last the extension cluster method in this paper is compared with the direct dynamic cluster method. A case study is performed on virtual assembly of high power Light Emitting Diodes automatic sorting machine manufactured by a factory.

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