Development of a Tool for Design of Reusable Space Transport Systems

Abstract

Recent developments in reusable launcher technologies performed by SpaceX have caused an increase in interest in this field across the world, with new companies and research institutes undertaking the challenge of designing such a launcher. However, expendable launch vehicles (LV) still remain a sizeable portion of the industry, and the possibility to design one must still be considered. At the beginning of an LV project, a preliminary design is necessary; it serves as a frame on top of which the project is defined more in detail. Thus, a versatile tool for preliminary LV design is needed. To address this challenge, in this thesis, written at the Department of Supersonic and Hypersonic Technologies of the German Aerospace Center (DLR), a programmatic tool in Python for preliminary LV design has been developed. The tool, called AIOLOS, employs a design algorithm with statistical relations to determine the launchers mass composition and geometry, as well as its centre of gravity and moment of inertia. Then it compares the outcome with the initial assumptions in an iterative optimising loop, until it reaches satisfactory convergence. Special emphasis was put on the application of the object-oriented programming paradigm, which improves code organisation and flexibility. The tool has been designed for compatibility with DLR's aerodynamics analysis tool CAC, and the trajectory simulation tool STRATOS, establishing the main frame for a complete launcher design software. Moreover, mathematical model for general optimal staging, alongside several design procedures, has been developed, and new component mass- and dimension-estimating relations have been derived. To assess the tool's correctness, an extensive simulation campaign has been undertaken, in terms of which 8 launchers and advanced launcher concepts have been modelled, and the outcome has been compared with the available data. Three of these launchers have been verified with the aerodynamics and trajectory calculations, to investigate if the vehicle would achieve the aimed orbit. Lastly, an input parameter study has been conducted, to investigate the influence of the initial assumptions on the final LV design. The comparison campaign has shown promising results, with payload mass of the modelled Ariane 5 ECA deviating from the reference by -4.6%, and of RETALT by -1.0%,; these launch vehicles have been tested coupled with CAC and STRATOS, which indicates that the whole design software may produce realistic preliminary designs.