Design-Analysis Associativity Technology for PSI, Phase I Report: Pilot Demonstration of STEP-based Stress Templates
R. S. Peak, R. E. Fulton, A. Chandrasekhar, S. Cimtalay, M. A. Hale, D. Koo, L. Ma, A. J. Scholand, D. R. Tamburini, M. W. Wilson (Feb. 2, 1999) Georgia Tech Project E15-647, The Boeing Company Contract W309702.
Aerospace structural analysis integration, Boeing, COB, MRA, CATIA, lug, fitting, Xaitools
As part of PSI, Georgia Tech has contributed an information modeling language, termed constrained objects (COBs), that is aimed at next-generation stress analysis tools. COBs combine object and constraint graph techniques to represent engineering concepts in a flexible, modular manner. COBs form the basis of the extended multi-representation architecture (MRA) for analysis integration, which is targeted at environments with high diversity in parts, analyses, and tools [Peak et al. 1998, 1999]. A key MRA distinctive is the support for explicit design-analysis associativity (for automation and knowledge capture) and multidirectional relations (for both design sizing and design checking). Another MRA characteristic is using COBs to represent and manage complex constraint networks that naturally underlie engineering design analysis.
Using a case study approach, lug and fitting design guides have been recast as example reusable COB libraries. The use of these and other COBs on structural parts relevant to the aerospace industry has been demonstrated. These case studies utilize XaiTools, a toolkit implementation of MRA concepts, which interfaces representative design tools (CATIA CAD, materials and fasteners libraries) and general purpose analysis tools (Mathematica solver, ANSYS FEA).
It is anticipated that COBs and the MRA will contribute key technologies to the overall PSI next-generation analysis tool architecture. The potential impact of explicit design-analysis associativity is significant. Capturing such knowledge, which is largely lost today, enables libraries of highly automated analysis modules and provides a precise reusable record of idealization decisions. User adaptation/creation of existing/new analysis templates is also possible.
Today creating views of analysis results such as internal analysis documentation (strength check notes) and regulatory agency summaries typically requires extensive manual effort. While COBs focus on core associativity and analysis computation relations, their combination with technology like XML should enable interactive "pullable views" to help streamline this analysis task. Other COB applications are anticipated, including upstream sizing and inter-analysis associativity.