Framework for integrated mechanical design automation☆
Introduction
Engineering design is an iterative process encompassing requirement definition, concept design, detailed design, and design validation/analysis. Fundamental to this design process is the need for consistent management and propagation of constraints, relations, associations, dependencies and domain knowledge associated with the product. Constraint-based design provides a unified framework to meet these critical needs.
However, constraint management and solving in three-dimensional (3D) space is a very challenging task due to the characteristics of the problem. Recognized issues include non-linearity, large problem size, complexity of user comprehension, multiplicity of solutions, ill-conditioning, poor initial conditions, and large perturbations. Through a decade of research, we have developed Extended Variational Design Technology (VGX) using graph theory and numerical solution techniques in an attempt to address the above issues.
To improve user comprehension, an innovative approach using a Drag and Drop user interaction paradigm is proposed. In this virtual environment, the user can set up constraints, modify dimensions, change geometry size and location, and change dimensioning schemes by directly dragging the desired entity. The intuitiveness and simplicity of this user interaction paradigm greatly improves the usability of a 3D constraint-based design system.
VGX technology overcomes the temporal barrier in history processing. This breakthrough enables the support of 3D flexible constraining/dimensioning that is independent of model creation history. Furthermore, the complete mathematical representation of the VGX-based design achieves true integrated design automation. Mathematical analyses, such as mechanism/tolerance validation, can be performed without having to rebuild separate models. The technology has been implemented in a commercial Mechanical Design Automation (MDA) system and the initial results have proved its promising potentials.
This paper is organized as follows: Section 2 presents the modeling role of MDA systems in the current design process and a brief review of related research. Section 3 gives an overview of the technical foundation of VGX. 4 History-independent design, 5 Drag and drop user interaction paradigm present the history-independent part/assembly design and drag and drop user-interaction paradigm, respectively. Section 6 illustrates the VGX-based downstream applications. Section 7 presents the conclusions and future research areas.
Section snippets
Background—MDA systems and the engineering design process
Solid-based MDA systems have received rapid acceptance over the past decade due to significant productivity gains from new capabilities in rapid prototyping, interference checking, mechanism animation, and better interfaces with analysis programs. The MDA industry also benefits from continued advances in computing technology and mathematical/geometric algorithms. All systems can now model complex parts and assemblies with complex operations such as filleting/blending, shelling,
Technical foundation
A geometric constraint network consists of geometry, constraints (e.g. parallel constraints, coincident constraints, tangent constraints, and perpendicular constraints), dimensions (e.g. linear dimensions, angular dimensions, and radial dimensions), and engineering relations that can be explicitly expressed as numerical formulations to fulfill design intent as illustrated in Fig. 3 [19].
Constraint management deals with constraint validation, constraint degree-of-freedom (DOF) analysis, and
History-independent design
Breaking the dimensionality barriers between 2D wire-frame sketches and 3D parts and also the temporal barriers in construction history requires a new perspective to the geometric representation of parts/assemblies. The part and assembly can be treated conceptually as a group of half-space entities such as points, lines and planes. Constraints and dimensions between geometric entities control the sizing and positioning of features in the model. Part instances of an assembly or features of a
Drag and drop user interaction paradigm
Interactive modeling has existed for years, beginning with Ivan Sutherland's PhD work on the Sketchpad drawing system [21]. However, its use was limited to a simple system that involved no constraint and history. Bier [22] presented an alternative called “dynamic snapping” for 2D and 3D design schemes which provided a means for the user to align objects interactively. These early systems, however, lacked sophisticated mechanisms for preserving the user's intent during dragging.
The power of VGX
Integrated applications
Downstream applications such as tolerance and mechanism analyses are very useful to validate or refine a product design. The lack of a complete mathematical description in a conventional system severely restricts its ability to use a common model for design and analyses, which necessitates special expertise to rebuild analysis-specific models. The model-rebuilding process is very time-consuming and error-prone.
Since a complete mathematical description of a design is available in the VGX-based
Conclusion and future work
In this paper, we have presented VGX as a unified framework towards supporting a more complete design process: from design specification, conceptual design, detailed design, to design validation. VGX technology provides a complete mathematical description of the design model which allows downstream tasks such as mechanism analysis, tolerance analysis and design optimization to be performed without having to rebuild analysis-specific models. In addition, the drag and drop user interaction
Acknowledgements
We wish to express our appreciation to all members of the VGX team at SDRC for their contribution in the development and implementation of VGX technology. We also would like to thank Naomi Lokay for reading and editing the manuscript and making many helpful suggestions.
Jack C.H. Chung is an SDRC Fellow at SDRC. He received his MS and PhD in Mechanical Engineering from the University of Texas—Arlington (1982) and Purdue University (1984), respectively. Prior to SDRC, he was a Project Leader in AI Group at BP-America Corporate Research Center. His research interests include constraint-based modeling, intelligent CAD/CAM systems, virtual assembly and manufacturing, design optimization and engineering animation.
References (23)
- et al.
Analysis of dimension and tolerances in computer-aided mechanical design
Computer-Aided Design
(1978) An approach to computer aided parametric design
Computer-Aided Design
(1991)- et al.
Modification of geometric models through variational geometry
Computer-Aided Design
(1982) Variation of geometry based on geometric reasoning method
Computer-Aided Design
(1988)- et al.
Assembly modeling by geometric constraint satisfaction
Computer-Aided Design
(1996) - et al.
Variational geometry based on logical constraints and its applications to product modeling
Annals of the CIRP
(1987) - et al.
Parametric design in medusa system
- Chung JCH, Klahs JW, Cook RL, Sluiter T. Implementation issues in variational geometry and constraint management. Third...
- et al.
Variational geometry in computer-aided design
Computer Graphics
(1981) - Serrano D, Gossard DC. Combining mathematical models with geometric models in CAE systems. Proceedings of the ASME...
Numerical methods
Cited by (33)
Variational Direct Modeling: A Framework Towards Integration of Parametric Modeling and Direct Modeling in CAD
2023, CAD Computer Aided DesignCitation Excerpt :Clearly, this causes unnecessary loss of parametrics. ( Note that there was a time when synchronous technology also referred to 3D variational modeling [31,32], but this interpretation appears not have been implemented in the current Siemens NX software package.) Integration Method IV: Operation Translating This approach, proposed by both academia and industry (Autodesk), translates direct edits into operations of parameter tuning and/or order rearrangement of the features already presented in the model’s construction history [13,14].
A framework for collaborative top-down assembly design
2013, Computers in IndustryCitation Excerpt :In this way, the design time of a complex product can be greatly shortened [2]. The current computer aided design (CAD) systems are able to model parts and assemblies very well but not to design [3]. They provide only limited support for top-down design and collaborative design although they are good at supporting geometric modeling at the detailed design stage.
Kinds of seeing and spatial reasoning: Examining user participation at an architectural design event
2012, Design StudiesCitation Excerpt :There is a trajectory of thought that associates design with problem solving. On this track there has been a progression in understanding that not all design problems follow a logico-deductive pattern, and also an acknowledgement that not all problems are wicked (Coyne, 2004; Cross, 2001), as in some situations parameters and constraints can be defined for problem solving to be automated (see for example, Chung et al., 2000). ‘Problems’, as way of describing the design task at hand are more than a rhetorical trope in some situations.
The 3D Product Model Research Evolution and Future Trends: A Systematic Literature Review
2022, Applied System Innovation
Jack C.H. Chung is an SDRC Fellow at SDRC. He received his MS and PhD in Mechanical Engineering from the University of Texas—Arlington (1982) and Purdue University (1984), respectively. Prior to SDRC, he was a Project Leader in AI Group at BP-America Corporate Research Center. His research interests include constraint-based modeling, intelligent CAD/CAM systems, virtual assembly and manufacturing, design optimization and engineering animation.
Teng-Shang Hwang is a Principal Software Engineer at SDRC. He obtained his MS and PhD in 1988 and 1990 from Oregon State University and his BS in 1978 from National Taiwan University of Science & Technology, all in Mechanical Engineering. Dr Hwang's technical interest is in mechanical design automation.
Chien-Tai Wu is a Principal Software Engineer in the CAD/CAM department of SDRC, where he is involved with the development and implementation of the VGX technology. Before that he was working for CDI primarily on the technical development of the free form surface modeling. He earned his MS and PhD degrees from the University of Michigan at Ann Arbor in 1983 and 1985, respectively.
Yu Jiang is a Principal Software Engineer in the CAD/CAM department of SDRC. He earned his MS degree in 1984 from Beijing University of Aeronautics and Astronautics, China, joined SDRC in 1993. Prior to that, he did research work in CAD/CAM areas as a visiting scholar at ASU and Cornell University, and worked as a Software Engineer for a CAE software company. He is currently interested in constraint-based modeling and its applications.
Jia-Yi Wang is a Senior Software Engineer at SDRC. Dr Wang received his BS degree from the National Taiwan University in 1983, his MS degree in Civil Engineering and PhD degree in Mechanical Engineering from the University of Iowa in 1988 and 1992, respectively. He worked as an associate research scientist at the Center for Computer-Aided Design at the University of Iowa from 1992 to 1996. Dr Wang's areas of interest are mechanism, computational methods in multi-body dynamics, mechanical design automation system, and CAD/CAE integration for concurrent engineering.
Yong Bai is a Senior Software Engineer at SDRC, where he is involved with the development and implementation of constraint based solid modeling (VGX) technology. He earned a BS in Mechanical Engineering in 1984 and an MS in Manufacturing Engineering in 1987 from Tsinghua University in China. He also earned an MS in Computer Science in 1995 from Southern Illinois University at Carbondale. Before joining SDRC, he worked for the National Center for Computer-Integrated Manufacturing Systems in Tsinghua University in Beijing from 1987 to 1992. His research interests include intelligent CAD/CAM systems, constraint based solid modeling. He has authored and co-authored over 20 publications in scientific and engineering journals and conferences.
Hongliu (Denise) Zou is a Software Engineer II in the CAD/CAM department at SDRC. She received her BS in Precision Engineering and MS in Mechanical Engineering from Shanghai JiaoTong University in 1991 and 1994, respectively. She also earned her MS from the University of Iowa in Mechanical Engineering in 1996. Her research interests are in kinematics, dynamics, mechanism, and CAD/CAM systems.
- ☆
This article is an updated version of the paper “Extended Variational Design Technology—Foundation for Integrated Design Automation” formerly published in the ACM Solid Modeling ‘99 Proceedings and is included here by permission of ACM.