A Methodology for Aeroelastic Constraint Analysis in a Conceptual Design Environment
The research examines how the Bi-Level Integrated System Synthesis decomposition technique can be adapted to perform as the conceptual aeroelastic design tool. The study describes a comprehensive solution of the aeroelastic coupled problem cast in this decomposition format and implementation in an integrated framework. The method is supported by application details of a proof-of-concept high speed vehicle. Physics-based codes such as finite element and an aerodynamic panel method are used to model the high-definition geometric characteristics of the vehicle. A synthesis and sizing code was added to referee the conflicts that arise between the two disciplines.
This research's novelty lies in four points. First is the use of physics-based tools at the conceptual design phase to calculate the aeroelastic properties. Second is the projection of flutter and divergence velocity constraint lines in a power loading versus wing loading graph. The mapping of such constraints in a designer's familiar format is a valuable tool for fast examination of the design space. Third is the improvement of the aeroelastic assessment given the time allotted. Until recently, because of extensive computational and time requirements, aeroelasticity was only assessed at the preliminary design phase. This research illustrates a scheme whereby, for the first time, aeroelasticity can be assessed at the early design formulation stages. Forth, this assessment allowed to verify the impact of changing velocity, altitude, and angle of attack and identify robust design space with these three mission properties.
The method's application to the quiet supersonic business jet gave a delta shaped wing for the supersonic speed regime. A subsonic case resulted in a high aspect ratio wing. The scaling approach allowed iso-flutter and iso-divergence lines to be plotted. The main effects of velocity, altitude, and angle of attack on these iso-lines were also discussed, as was the identification of robust design space. The response surface surrogate models allowed convergence of the system optimization but questions were posed as to the accuracy of these quadratic models. Other future improvements include the addition of more disciplines and more detailed models.
Advisor:Daniella E. Raveh; Rudolph N. Yurkovich; Jaroslaw Sobieszczanski-Sobieski; Daniel P. Schrage; Dewey H. Hodges; Dimitri N. Mavris
School:Georgia Institute of Technology
School Location:USA - Georgia
Source Type:Master's Thesis
Date of Publication:04/12/2004