Preliminary design of a crewed Mars flyby Solar Electric Propulsion mission

Preliminary design of a crewed Mars flyby Solar Electric Propulsion mission

De Smet, S.

Noomen, R. (mentor)

Aerospace Engineering

Space Exploration

2014-12-10

The goal of this thesis research has been the improvement of a validated, robust low-thrust optimization tool for interplanetary trajectory design, written during the author's internship. This improved tool has then been applied to design crewed Martian flyby missions. Such a mission is a crucial step within the flexible path scenario for human and robotic space exploration, formulated ``in support of the Exploration Beyond LEO committee of the Review of US Human Space Flight Plans Committee commissioned by President Obama''. This mission would provide critical experience in preparation for a Mars landing without the actual risk of the landing itself. Several technologies could be validated on this kind of mission such as new countermeasures for radiation shielding, better regenerative life support systems, further investigation of the effect of deep-space isolation on the human psyche, etc. This thesis was the follow-up of previously conducted research during the author's internship for which a basic Sims-Flanagan based low-thrust optimization tool was written. The first goal of this thesis was the improvement of this existing code. Therefore, the code has been extensively profiled and analyzed to identify and remove bottlenecks. Furthermore, advantage has been taken from the sparsity of the Jacobian to further decrease the run time. Depending on the scenario, gains in run time of up to a factor 10 have been observed. Additionally, different representations of the Sims-Flanagan transcription have been investigated. It was found that compared to a throttled representation, the classical thrust representation is slower, but more robust. The second goal of this thesis was the addition of time-optimization capabilities. Therefore, methods to analytically derive the Jacobian elements with respect to time have been established. However, numerical difficulties arose from this method. To circumvent this problem, a forward finite-difference method has been written. Furthermore, several representations to couple ephemeris to time have been set up and compared. These time-optimization capabilities have been tested on Earth-Mars-Earth flyby missions launching in 2018. Based on previous results, those time-optimization capabilities could be validated for two different objective functions: minimized launch mass and maximized final mass. Finally, the latter objective function has been selected for this research. The third goal of this thesis was the automation of the addition and optimization of additional legs. Therefore, automation algorithms have been set up throughout the program. These and the previously established time-optimization capabilities have been tested on Earth-Venus-Mars-Earth flyby missions. During these tests, issues with local optima arose. Therefore, a multi-start method has been implemented and tested. This multi-start method circumvents the majority of those local optima issues. Using the added and validated capabilities, several launch windows for crewed Martian flyby missions have been identified for different SEP power levels, different launcher configurations and different payload masses in 2018, 2019 and 2021. In addition, an opportunity for a crewed Venus and Martian flyby mission has been identified launching in 2021.

Low-thrust
Trajectory Design
Crewed
Flyby mission
Mars mission
Orbital mechanics
Astrodynamics
Solar Electric Propulsion

http://resolver.tudelft.nl/uuid:289554a2-77bf-4692-9e47-39ff75015a19

Student theses

master thesis

(c) 2014 De Smet, S.