Research On STEP-based Manufacturing Feature Recognition And 3D Reconstruction Technology

The integrating of CAD/CAPP/CAM systems is key technique for realizing the integration, intellectualization and automatization of the machinery manufacture industry. However, there is a bottleneck problem about sharing and exchanging of data between CAD and CAPP systems, which restricted the application of integrating technique. STEP can resolve this problem between differents CAX systems effectively by providing a neutral mechanism without dependence on other systems and building data model for whole lifecycle of product. The dissertation analyzed and researched the STEP neutral file created by CAD system, extracted and recognised manufacturing features from the data model, built a intellectualized interface between CAD and CAPP system, and reconstructed 3D solid model for the STEP data model. A prototype automated feature recognition system without dependence on any CAD systems was implemented. The feasibility of the neutral mechanism provided by STEP was validated, and the integration of information between CAD and CAPP systems was explored and practised effectively.The research work in this dissertation is mainly focused on the following four areas:(1)The main content of STEP and its detail about exchange technology was researched. The EXPRESS language, the construction and format of STEP neutral file were researched fully. On the basis of the detailed analyse of geometrical and topological information in the STEP data model, and the characteristic of Object Oriented EXPRESS, C++ classes were design for mapping with key entities in the STEP data model. Finally a parsing program for extracting geometrical and topological information in the STEP neutral file was developed.(2)The manufacturing features recognition was using the attributed adjacency graph (AAG) based feature recognition approach. Firstly, the convex-concave decision algorithm for the edge in the STEP data model was proposed, and the AAG for parts was created by using this algorithm. After the AAG was created, it was stored in the computer memory using the extended attribute adjacency matrix (EAAM) method introduced in this dissertation. Secondly, these sub-AAG with the predefined sub-AAG of classical manufacturing features were also stroed in the computer memory using EAAM, and then the algorithm of extracting the sub-AAG of features was designed. Finally, a feature matching algorithm was proposed, manufacturing features were effectively recognized in the feature matching procedure.(3)The integrated construction and function of free and open source Open CASCADE (OCC) 3D modeling platform was researched. The data exchanging module and the visualization module mainly concerned part of OCC were analyzed detaily. The interface function about reading STEP neutral file and reconstructing 3D model were researched. Finally, 3D reconstruction of STEP data model for mechanical part was implemented using the function in dynamic link library provided by OCC.(4)On the basis of the theoretics research, and designs of STEP information extraction module, AAG build module, feature sub-AAM (sub Attribute Adjacency Matrix) extraction module, feature sub-AAM match module and STEP 3D solid model reconstruction module, a prototype system named DNAFRS (Dong Nong Automated Feature Recognition System) was implemented by using Visual C++ development tool.