Function modeling combined with physics-based reasoning for assessing design options and supporting innovative ideation
- 1 February 2017
- journal article
- research article
- Published by Cambridge University Press (CUP) in Artificial Intelligence for Engineering Design, Analysis and Manufacturing
- Vol. 31 (4), 476-500
- https://doi.org/10.1017/s0890060417000403
Abstract
Functional modeling is an analytical approach to design problems that is widely taught in certain academic communities but not often used by practitioners. This approach can be applied in multiple ways to formalize the understanding of the systems, to support the synthesis of the design in the development of a new product, or to support the analysis and improvement of existing systems incrementally. The type of usage depends on the objectives that are targeted. The objectives can be categorized into two key groups: discovering a totally new solution, or improving an existing one. This article proposes to use the functional modeling approach to achieve three goals: to support the representation of physics-based reasoning, to use this physics-based reasoning to assess design options, and finally to support innovative ideation. The exemplification of the function-based approach is presented via a case study of a glue gun proposed for this Special Issue. A reverse engineering approach is applied, and the authors seek an incremental improvement of the solution. As the physics-based reasoning model presented in this article is heavily dependent on the quality of the functional model, the authors propose a general approach to limit the interpretability of the functional representations by mapping the functional vocabulary with elementary structural blocks derived from bond graph theory. The physics-based reasoning approach is supported by a mathematical framework that is summarized in the article. The physics-based reasoning model is used for discovering the limitations of solutions in the form of internal contradictions and guiding the design ideation effort.Keywords
This publication has 19 references indexed in Scilit:
- Frameworks for organising design performance metricsJournal of Engineering Design, 2016
- Discovering the Real Problematic Situation: The First Aspect of Conceptual DesignINSIGHT, 2014
- Making function modeling practically usableArtificial Intelligence for Engineering Design, Analysis and Manufacturing, 2013
- Automated Mapping of Physical Effects to Functions Using Abstraction Ports Based on Bond GraphsJournal of Mechanical Design, 2013
- Evaluation of the functional basis using an information theoretic approachArtificial Intelligence for Engineering Design, Analysis and Manufacturing, 2010
- Design methodologies: Industrial and educational applicationsCIRP Annals, 2009
- Experimental Studies Assessing the Repeatability of a Functional Modeling Derivation MethodJournal of Mechanical Design, 2003
- Qualitative physics using dimensional analysisArtificial Intelligence, 1990
- Qualitative Physics: Past, Present, and FuturePublished by Elsevier BV ,1988
- State Variables and Pseudo Bond Graphs for Compressible Thermofluid SystemsJournal of Dynamic Systems, Measurement, and Control, 1979