Distributed Advanced Multidisciplinary Algorithms for Genetic Evolution (DAMAGE)

The D. N. American research team proposes a novel approach to improving the process of collaborative design optimization for composite rocket motors. In Phase I, we will investigate the feasibility of our Distributed Advanced Multidisciplinary Algorithmsfor Genetic Evolution (DAMAGE) approach to the problem. The proposed system leverages the team's experience--in composite rocket motor design, in developing parallel Genetic Algorithms for composite material design optimization, and in distributed Windowsnetwork programming--to design a state of the art web-enabled Windows 2000 distributed Genetic Algorithm system to facilitate the design process. The hallmark of our idea is an object-oriented design incorporating parallel Genetic Algorithm COM+ objects,facile network access to distributed CAD and analysis software via Web Services, and a means of accessing these functionalities via a robust Application Programming Interface (API). The Phase I work effort will yield a proposed design to be implemented inPhase II, with immediate benefits to the military in terms of creating optimal designs in the face of multiple participants and parameters. There are many potential commercial extension of the DAMAGE system. These include other organizations involved indesign optimization, as well as commercial entities that will benefit from the underlying distributed GA technology. This tool is expected to have widespread use among entities that design and manufacture rocket motor cases as well as other compositestructures. Beyond these applications, this tool can be used by any entity that needs to optimize an outcome in the face of numerous, conflicting constraints and variables. Thus, multidisciplinary teams can also use this tool to design aircraft,automobiles, buildings, consumer products, financial management products, etc. The beauty of D.N. American's tool is that it identifies optimum outcomes, independent of human bias, in the midst of numerous conflicting inputs.