Meta-physical product modeling

This dissertation summarizes research to develop a product modeling framework that can capture functionality, design intent, shape, and allowable shape variations. The proposed framework divides product models into two realms: a physical realm and a meta-physical realm. The physical realm model contains the information commonly associated with physically based models: geometry, topology, dimensions and tolerances, materials, etc. The meta-physical model contains the meta-knowledge of the physical design, i.e., the information that describes the nature, structure, behavior, and reason far existence of entities in the physical model. In this dissertation modeling products in these two realms is termed Meta-Physical Product Modeling.; The meta-physical product modeling paradigm uses Product Definition Units (PDU) at several levels of generality including need, function, physical principle, embodiment, artifact type, and artifact instance. Function, the behavior of an object, is defined as an operation on energy, material, or signal, and is represented as a PDU at the function level of generality. Functions are decomposed into subfunctions and specialized to the artifact instance level of generality. Design intent, defined as the purpose or objective of an object, is represented as a characteristic of a PDU. A generalized feature concept is the key to linking the meta-physical model with the physical model. A feature is the physical realm image of a meta-physical realm PDU.; A prototype system, containing a meta-physical modeler, a feature modeler, a geometric modeler, and a dimension and tolerance modeler, was designed, implemented, and tested. The system demonstrates the ability to capture the design intent and functionality of physical objects and their features, as well as the ability to capture the design intent and functionality underlying the dimensions and tolerances of physical objects.