The Construction And Application Of The Platform For MDO Based On DAKOTA

Multi-disciplinary design optimization (MDO) is a field of engineeringthat uses optimization methods to solve design problems incorporating anumber of disciplines. The multidisciplinary design optimization framework,is refers to the construction of a design and implementation method of theoptimization of multiple disciplines computing environment. It can calculatethe integrated operation of different disciplines and achieve automation ofthe design, analysis and optimization process.With engineering problems become more and more complicated,Companies and research groups are eager to shorten the time, improveefficiency, increase technology content and adapt to the fierce marketcompetition with the help of the multidisciplinary optimization framework.Under this background, the paper establishes the multi-objectiveoptimization design platform. It select open source optimization frameworkDAKOTA[1]as the algorithm library and the Graphical User Interfacesoftware Orange to achieve visual programming, integration and automationof the optimization process, optimization analysis, data mining and otherfunctions.The platform consists of six modules, the master, the model generator,the optimizer, the monitor, the data manager and the postprocessor. The master module is responsible for the overall control of all optimization tasks.The model generator is for generating problem-dependent analysis andoptimization codes. The optimizer is for performing the subsystem andcoordination optimization tasks. The monitor is for viewing optimizationresults and steering iteration processes. The data manager is for themanagement of all data produced during analysis and optimization processes.The postprocessor is designed for visualizing and printing computationresults after completion of the optimization procedure.Through the research of automatically parametric technology onplatform of Pro/E, NX, ANSYS and their respective secondary developmentinterface, this paper develop a multidisciplinary optimization frameworkwhich integrates CAD, CAE components.Two examples of multidisciplinary optimization problems are issued.After the developing and testing, the system effectively improves theefficiency of the rotor disk design and Water assisted injection molding,shortens product development time. The platform has considerablecompetitive strength and practical value in fields of efficiency, open source,scalability and cost, which is quite suitable for small and medium-sizedenterprises and defense unit.