Assembly accuracy prediction based on CAD model

Analysis of assembly tolerance is an important method to evaluate the reasonableness of tolerance allocation in product design. For the purpose of convenient applications, the method to assembly accuracy prediction presented in the paper is to extract the related assembly information from the CAD assembly model directly, which is managed in the way of data structures with inter-relationship and accessed with database technology. From the extracted assembly information, 3-D assembly dimension chains was automatically generated through searching the assembly constraints and their dependence and related dimension positions in the database. Then, the dimension chain is optimized by using ant colony algorithm to create an optimal one. Based on the input tolerance or actual dimension for each dimension/constraint, the assembly accuracy of the product can be predicted by the dimension chain and the accuracy prediction model. Taking a machine tool as an example, the accuracy prediction result shows that the predicted assembly accuracy meets the design specifications.