Object-oriented propagation modeling to support CAD/CAM and software engineering

Date of Completion

January 1997


Information Science|Computer Science




Representing dynamic interdependencies between design objects is an essential part of modeling the critical software communications found in today's increasingly complex software systems. Current software development methodologies, software development environments (SDEs), and computer aided design (CAD) systems lack the ability to represent and manipulate these interdependencies, even though the specific information needing representation is often well understood by the designer. Representing dynamic interdependencies in an object-oriented design environment focuses on the ability to specify relationships between objects that are not ancestrally related. For example, a CAD designer altering the pitch of a gas turbine engine's fan blade may need to consider the stress where the blade is mounted to the turbine disk. Changes in the stress (caused by the change in fan blade pitch) are said to propagate to the attachment of the fan blade to the disk. Propagation modeling seeks to represent such interdependency knowledge as an important part of the design process. From a software engineering perspective, propagation modeling, as applied to the software design process, holds potential as a means to more precisely define and then consistently maintain design interdependencies. Our research pursues an increase in the engineering rigor of the software design and development process by expanding the object-oriented paradigm to include the mechanisms necessary to support application-wide information consistency. The general context of our effort is the development of a framework for supporting the design, analysis, and development of propagation modeling capabilities in an object-oriented environment. Our framework offers propagation modeling constructs for both design and development, with support for dynamic, designer-defined propagation modeling choices. Included in our framework are constructs that promote information consistency through automatic and on-demand design analyses, which alert the designer to potential design flaws and inconsistencies, in support of the detailed evaluation of a design. ^