CloudCraft: Software for Early-Stage Aircraft Design and Optimization

Abstract

The configuration design of fixed-wing aircraft is a highly-interconnected process; changes in a seemingly insignificant parameter can have a meaningful impact on sys-tem performance, and even validity. This thesis proposes a software platform to streamline much of this design process. An extensive database of existing aircraft parameters was compiled, which forms the basis of the design optimization of simplified aircraft geometry with respect to a variety of optimization criteria, including wing loading, thrust to weight ratio, and takeoff mass. The software, called Cloud-Craft, takes advantage of existing software such as NASA’s Open Vehicle Sketchpad, Xfoil, and Python’s SciPy optimization package to provide the aircraft designer with meaningful and realistic feedback during the design process. It does not remove the designer from the design process; rather, it guides the designer in making informed refinements while automating primary analyses and ensuring self-consistency. CloudCraft was tested on the design of a long range, wide-body jet. Early results showed that global optimization methods are more suitable than local optimization methods, and that utilizing feedback in design parameter choices is essential to producing realistic results. Using a genetic algorithm with parameter feedback, Cloud-Craft succeeded in producing a realistic design in terms of constraint plot analysis and comparison to existing aircraft.