Issues in industrial multidisciplinary optimization
J. Bennett, Peter A. Fenyes, W. Haering, M Neal · 7th AIAA/USAF/NASA/ISSMO Symposium on Multidisciplinary Analysis and Optimization · 1998
Several mathematically based multidisciplinary design strategies are illustrated with an exploratory multidisciplinary analysis and optimization package on a simple example problem. These examples are used to motivate a discussion of required data handling and processing modules. These requirements envision a situation where some disciplines may have computationally expensive analysis capabilities and where not all disciplines have easily available approximations for all required quantities and Optimization Conferences. The second is a more ad hoc approach which is evolving from the traditional design and analysis communities and is typified more by a Multidisciplinary Analysis capability that is evolving in the commercial CAD and CAE environments. These environments envision a common parametric description of the artifact and an ability to generate input information for several disciplines from this format. Then analyses will be performed using complex commercial or proprietary codes and decisions made on how to modify the initial design. This process is usually characterized by significant human interaction to develop the artifact model, generate the analysis models, execute the analysis models and finally to examine the output and make decisions. The formal mathematical approach tends to use much simpler and easily modified local analysis methods (that execute on the order of minutes or seconds) and to concentrate on multidisciplinary design algorithms which interact with the analysis methods in an almost automatic fashion.