Multidisciplinary High Fidelity Analysis And Optimization Of Aerospace Vehicles

Author : Julien Marie Anne Jean De Muelenaere
Publisher :
ISBN 13 :
Total Pages : pages
Book Rating : 4.:/5 (17 download)

DOWNLOAD NOW!

Download or read book High-fidelity Trajectory-based Multidisciplinary Design Optimization for the Conceptual Design of a Reusable Air-launched Spaceplane written by Julien Marie Anne Jean De Muelenaere and published by . This book was released on 2018 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: This dissertation presents techniques to enable high-fidelity trajectory-based Multidisciplinary Design Optimization (MDO) for the conceptual design of a reusable air-launched winged space vehicle, or spaceplane. Low-cost access to space is one of the greatest challenges for today's aerospace sector. Prohibitive costs of current systems motivate the search for unconventional approaches and emphasize the importance of reusability. Notable examples of Reusable Launch Vehicles (RLV) are the NASA Space Shuttle and the Boeing X-37B. More recently, in September 2018, SpaceX introduced the latest design of its Big Falcon Rocket (BFR). The second stage features steerable canards and radially adjustable wings. Even though the company tried to avoid these additions, aerodynamic deceleration and aerodynamic pitch and roll control during atmospheric reentry are required. These allow avoiding the dangerous high-speed region at low altitude, and safely re-entering in a controlled way, for a wide range of payloads and atmospheric densities. The BFR second stage is essentially an air-launched spaceplane. Because the design of an RLV involves multiple disciplines (structures, aerodynamics, trajectory, stability, etc.) that interact with one another, it requires multidisciplinary analysis and optimization. High-fidelity MDO in a conceptual design setting, for the rapid and efficient preliminary design of such vehicles, is investigated. The contributions in the thesis address the shortcomings from the literature by developing a new high-fidelity framework and formulate new methodologies to tackle the challenges arising from high-fidelity. A Non-Linear-Programming-based All-At-Once MDO Architecture in the context of high-fidelity is proposed. Also, a methodology for combining sizing load cases with response surfaces in the context of structural analysis, in order to reduce the dimensionality of the required response surfaces, is developed. The methodology is one of the techniques that make the use of All-At-Once MDO architecture possible. The optimization of the spaceplane geometry is coupled with the optimization of its trajectory (ascent and reentry), using SNOPT and GPOPS-ii. A parametric geometry for the spaceplane is generated using GeoMACH. Aerodynamic coefficients are computed using the Reynolds-Averaged Navier-Stokes equations implemented in SU2 and weight estimates for the spaceplane are generated by performing structural optimizations for a variety of sizing load cases using TACS. Kriging response surfaces are used as surrogates for the computationally intensive disciplines. Most importantly, flight mechanics constraints, such as trim and stability, are imposed in order to properly size the planform of the wing as well as the control surfaces. This research covers two novel case studies, that could not have been solved with the frameworks available in the literature. In the first one, we demonstrate the importance of covering RLV's rigid body characteristics during reentry, like trim and stability, even at conceptual design level. In the second one, the optimizer converges to an interesting skip reentry in order to extend the covered downrange and to lower the mechanical and aerothermal loads on the vehicle by extending the reentry time.