ONTOLOGY-BASEDAPPROACHFORASSEMBLYVARIANT DESIGNUSINGLIAISONS

Abstract

Assembly variant design is an interesting way to produce a variety of assembly designs by applying several design suggestions. The design suggestions are used to satisfy various customer requirements (also called design requirements). Various design suggestions are changing different variant parameters of an existing product, changing their modules, product architecture, or mixing and matching of the components. These design suggestions make the variant design a complex problem for the designer. During assembly variant design, identifying the impact of each design suggestion on assembly joint or liaison and the changes of its corresponding assembly plan becomes important. This thesis presents an ontology-based framework to address the issues related to assembly variant design using liaisons. Liaisons available in the literature are primarily developed for the riveting and welding process. The liaison has been extended to represent, extract assembly joints available in adhesive bonding called adhesively bonded assembly features (ABAFs). The adhesively bonded assembly features (ABAFs) is represented as a set of faces with a characteristic arrangement among the faces among parts in proximity suit able for adhesive bonding. Several ABAF attributes like relative overlap, relative orientation, the relative location of the part, joint feature, etc. are represented in the ontology to capture ABAF types (like tongue and groove, dovetail, scarf feature, etc.). These ABAFs are not explicitly available in the current commercial CAD system. Therefore, in this thesis, algorithms are developed to extract the attributes of ABAFs and identify ABAFs from the CAD model. CAD assembly models from the automotive and aerospace domain are used to validate the algorithms. This extracted liaison information is further used for generating a variant design by developing an ontology-based decision support system. In this system, a knowledge base is built by the development of an ontology to formally represent the taxon omy, properties, and causal relationships of/among core concepts involved in the variant design. Second, a five-step sequential procedure is established to facilitate the designer to make automatic decisions in variant design. The process takes the extracted liaison information from the CAD model of an existing product as the input, automatically infers the design suggestions, the variant design type required for each design suggestion, and its effect on joint information through SWRL rule-based reasoning. Finally, the efficacy of the ontology-based decision support system is i evaluated using case studies. Next, the extracted liaison information is used for assembly joining process selection using an ontology-based knowledge framework. This framework helps in identifying the modified liaison information due to variant design and selecting the suitable joining process for a particular joint based on this modified liaison information. So, a joining process selection (JPS) ontology is developed, which represents various core concepts like feature, material, product, joint requirement, and joining processes required for the selection of joining process at the design stage. In this ontology, the core concepts and relationships involved in all aspects of joining process selection are analyzed in detail. A knowledge retrieval and reasoning approach integrating ontology concepts, instances, knowledge rules, and semantic queries encoded with Query-enhanced Web Rule Language (SQWRL) are implemented. The efficacy of the ontology-based knowledge framework for joining process selection is evaluated using case studies from the automobile and aircraft industry. Finally, the thesis concludes by summarizing the main contributions and outlining the scope for future work.