Global multi-disciplinary design optimisation of super-hypersonic components and vehicles
Uloženo v:
| Název: | Global multi-disciplinary design optimisation of super-hypersonic components and vehicles |
|---|---|
| Autoři: | Cusick, Andrew |
| Informace o vydavateli: | University of Glasgow, 2021. |
| Rok vydání: | 2021 |
| Sbírka: | University of Glasgow |
| Témata: | TL Motor vehicles. Aeronautics. Astronautics |
| Popis: | Both high-speed transportation and reusable launch vehicles are once again under constant development, after reaching stagnation points due to immature technologies. This resulted in many programs being shut down prior to launch, while the fortunate few that made it to production, were ultimately cancelled due to their combination of unreliability and high operational costs. Today, national institutions and private organisations alike are competing to revolutionise the super-hypersonic regime, but optimal performance remains as imperative as ever if such concepts are to be realised as long term solutions. The focus of this research is to develop a framework capable of generating, analysing, and optimising such vehicles; producing high performance designs that comply with real-world constraints from a broad spectrum of potential candidates. This contrasts with the majority of design optimisation work found in aerospace engineering literature, which favours improving existing designs through localised alterations. While this is useful in many engineering applications, it means little in the context of high-speed transport and reusable launch vehicles; since reliable designs do not yet exist, and current concepts vary drastically based on a number of factors. The direction of this research is therefore to define optimality on a larger scale, based on performance metrics that are transcendent among all super-hypersonic vehicles. Quantifying the performance of such vehicles in this large scale conceptual design context, in terms of both optimality and feasibility, requires the development of versatile analysis and optimisation tools. A number of vehicle generation and integration tools have been developed, combined with low-fidelity aerodynamic methods, particle swarm optimisation, and robust constraint handling techniques, to provide the basis of this research. The justification, necessary background, and implementation of these methods is discussed in detail, alongside their validation in a large design space context. Further analysis modules and optimisation techniques are employed in addition to these primary methods, which are used to provide a deeper understanding of vehicle performance, optimality, and feasibility. Initial optimisation work is focussed on aerofoils. A number of generative methods are employed and tested for their level of control over such shapes, since fine tuning will be required to achieve globally optimal results. Following this, aerodynamic optimisations are carried out, demonstrating the combined methods and their ability to consistently produce high performance results in both existing design and global optimisation contexts. This work is then extended to multi-fidelity analysis; where open-source high-fidelity programs are employed efficiently through the implementation of dynamic surrogate modelling techniques. The combination of these methods with the existing framework is used to improve the quality of results. The remaining work is based on optimisation of three dimensional bodies; with both wing and full vehicle design spaces employed. A low-fidelity wing-box model is implemented in the former case, adding further design variables and constraints to ensure structural integrity within optimal designs. A prior reusable launch vehicle concept is used to benchmark these results, allowing trim and longitudinal stability characteristics to be determined. An improved parametric fuselage definition is then employed, enabling full vehicles to be designed and optimised from an arbitrary starting point. Uncertainty quantification and improved constraint handling methods are utilised to increase confidence in vehicle performance. Finally, a mass estimation routine is implemented to determine weight, payload and trim characteristics; with surface temperatures, lateral stability, and payload integration defining further feasibility characteristics. |
| Druh dokumentu: | Electronic Thesis or Dissertation |
| Jazyk: | English |
| DOI: | 10.5525/gla.thesis.82486 |
| Přístupová URL adresa: | https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.838817 |
| Přístupové číslo: | edsble.838817 |
| Databáze: | British Library EThOS |
Nájsť tento článok vo Web of Science Buďte první, kdo okomentuje tento záznam!