2412 Analysis building blocks: a rich information model context for knowledge-based ﬁnite element analysis Sai Zeng a , Russell S. Peak b,* , Miyako W. Wilson b , Ryuichi Matsuki d , Angran Xiao c a Engineering Information Systems Lab, Georgia Institute of Technology, Atlanta, GA 30332-0560, USA b Manufacturing Research Center, Georgia Institute of Technology, Atlanta, GA 30332-0560, USA c System Realization Laboratory, Georgia Institute of Technology, Atlanta, GA 30332-0560, USA d Shinko Electric Industries Co., Ltd., Advanced Product Design & Development Division, Nagano, Japan Abstract In a product design and analysis process, engineers have different usage views towards product information models. The heterogeneous transformation problem has been presented to characterize the resulting gap between design models and analysis models. The multi-representation architecture (MRA) has been developed to realize this transformation through four stepping-stone information representations, including analyzable product models, context-based analysis models, analysis building blocks (ABBs), and solution method models (SMMs). In this paper, our primary focus is on ABBs for solid mechanics and thermal systems that generate FEA SMMs to obtain their results. ABBs, which represent the analytical usage view for analysis engineers, are constructed using an object- oriented knowledge representation known as constrained objects (COBs). ABBs represent product-independent analysis concepts such as continuum mechanics bodies and idealized interconnections as semantically rich, reusable, modular, and tool-independent objects. To demonstrate the efﬁcacy of the ABB model, an electronic chip package thermomechanical analysis test case is overviewed. This extended ABB approach provides an effective way to capture engineering knowledge and decrease FEA modeling time. Keywords: Heterogeneous transformation; Multi-representation architecture (MRA); Analysis building block (ABB); Solution method model (SMM); Constrained object (COB); Finite element analysis (FEA) 1. Introduction Targeting the needs of design and analysis integra- tion, Peak and colleagues [1–3] have proposed a gen- eral methodology for automating ubiquitous analysis to support product design. In this methodology, the multi- representation architecture (MRA) is presented to facilitate heterogeneous transformations by explicitly representing design-analysis associativity and supporting numerous di- verse analysis models for each product type. The MRA consists of four stepping-stone information representa- tions, i.e. analyzable product models, context-based analy- sis models, analysis building blocks (ABBs), and solution method models (SMMs). ABBs and SMMs are product- independent models that facilitate generalized mappings between a single product model and diverse analysis mod- ∗ Corresponding author. Tel.: +1 (404) 894-7572; Fax: +1 (404) 894-9342; E-mail: russell.peak@marc.gatech.edu els. ABBs describe the theoretic physical systems, such as continuum mechanics systems, while SMMs represent ABBs in relatively low-level solution technique form, such as ﬁnite element analysis models. In this paper, we primarily focus on ABB models for solid mechanics and thermal systems that utilize FEA-based SMMs. In the next section, we brieﬂy introduce the MRA as the context for ABB models. In Section 3, the concept and architecture of the ABB representation is overviewed. In the context of solid mechanics and thermal system, we discuss the information needed to develop an ABB model and transform it into a corresponding SMM that uses FEA as the solution technique. Then, we show how to implement ABBs using an object-oriented knowledge representation known as constrained objects (COBs). A COB diagram is presented in Section 4 to illustrate the key attributes of ABB objects and their relationships. Finally, Section 5 presents an electronic chip package thermomechanical analysis scenario to show the efﬁcacy of the ABB model.  2003 Elsevier Science Ltd. All rights reserved. Computational Fluid and Solid Mechanics 2003 K.J. Bathe (Editor)