Sponsor: NSF
During the last few decades computer software has become more and more critical in the analysis of engineering and scientific problems. Much of the reason for this change from manual methods has been the advancement of computer techniques developed by the research community and, in particular, universities.
In most engineering disciplines the development cycle for a typical research advancement has previously begun with a graduate student who works on a problem starting with near-zero code (that is, no computer software) and a few function libraries. The student creates not only the actual code of an advancement, but also a huge amount of additional "support" code. After scholarly papers have been generated from this problem-specific code, the code is typically abandoned because it is usually poorly documented, lacks flexibility to solve a broader range of problems, is not error tolerant, and is not easily integrated into any standard framework.
The problem, therefore, is how can such efforts and implementations be captured such that they can be of longer term value and immediately make technical advances available for solving practical engineering problems. It is hoped that by adopting object-oriented paradigms for analysis, design, and programming, the capture of research work will become more practical. The present project's main thrusts to address these advancements are essentially two-pronged, and may be stated as: (1) enhancement of general purpose software development environments through creation of domain-specific object-oriented programming tools that support software reuse, and (2) object-oriented analysis, design, and implementation of critical domain-specific leverage components for engineering computation. The structural engineering domain is targeted as the test case for this work.