Roadmap for Moderate Fidelity Conceptual Design with Multilevel Analysis and MDO

Abstract

††† * ** The design process for an aircraft or spacecraft can be divided into three sequential stages: Conceptual, Preliminary, and Detail Design. The first design stage represents the evaluation of the initial options to satisfy a given set of performance and cost specifications by low order methods, resulting in consensus on a general design concept. Any changes to this concept become more costly as the design progresses. If more information can be obtained about the design space at the conceptual design stage, more informed decisions can be made early, saving cost and effort, or allowing for more configurations to be evaluated in the same time frame. To this end, there have been recent suggestions to perform MAO studies, normally executed in preliminary design, earlier, in the conceptual design phase. The development and/or enhancement of a low-to-moderate fidelity aerospace vehicle synthesis software system in a parallel and distributed computing environment is proposed here. Its distinguishing feature will be its ability to perform significantly accelerated but robust vehicle design trade and optimization studies for various disciplines and configurations in parallel. The system should feature an object-oriented template-driven framework, written either in C++ or based on a .NET environment, in conjunction with standard FE codes for structural, dynamic, and aeroservoelastic analysis and optimization, and additional templates for performance, costing, etc. The efficiency of the software can be enhanced by an object-oriented interactive model builder, which will allow for the fast creation of many different scalable structural FE design models with the associated aerodynamics and controls input for conventional and unconventional aircraft. Additional improvements in efficiency may come from incorporating a product platform/product family approach in the synthesis system to allow for multi-mission design. Additional sub-objectives include: developing a capability for analyzing non-linearities early in the design process, especially for unconventional aircraft, to avoid costly fixes at the preliminary design stage; identifying parallel computing software and/or schemes for converting complex serial source codes to parallel operation and developing Grid-enabled software for potential efficiency improvements and time savings; demonstrating the capabilities of the resulting design tool to prove its applicability and performance under complex operational conditions. The sensorcraft under investigation at AFRL will provide excellent test objects for all aspects of the proposed development, since they are unconventional due to their joined wing configuration and highly non-linear due to their considerable flexibility.