MS thesis abstract - Kimbrel, Scott
| Author: | Kimbrel, Scott |
| Degree: | Masters of Science |
| SERC #: | 20-02 |
| File type: | PDF, 1255 kB |
 Download file (opens in new window) | |
Optimization of Electric Propulsion Orbit Raising
The increasing power levels now available on geo-synchronous satellites have made it feasible to use electric propulsion engines to perform orbit raising from transfer orbits to GEO. Electric thrusters have very low thrust but are highly efficient, so transfers require the thruster to fire almost continuously for weeks or even months, but also provide significant savings in propellant mass compared to all chemical missions. The complicated nature of the transfer and almost continual firing of the thruster require the thrust angles to be calculated and optimized for the entire transfer time. It is also important to optimize the transition point between the chemical and electric transfers, however the available low-thrust optimization tools are not rapid and flexible enough to allow a broad survey of possible strategies. For this reason, highly analytic derivations have been completed and new optimization software (called MITEOR - MIT Electric Orbit Raising) has been developed in Matlab to optimize thrust angles for constant-low-thrust transfers with no plane changes (2D), as well as for transfers with plane changes (3D) that are restricted to not rotating the argument of perigee or longitude of the ascending node. The 2D version of MITEOR is robust, converges well, and can optimize for transfers with specific initial conditions or display multiple transfer optimizations at once and view trends between transfers. Derivations have also been completed for both 2D and 3D transfers that optimize both thrust angles and thrust magnitude. These variable thrust derivations have been found to be completely analytic and require no additional numerical routines. The results of the 2D and 3D variable thrust transfers are typically 5-10% more fuel-efficient than constant thrust, and can be used to easily calculate first cut approximation to the constant thrust cases, providing an optimum upper bound. This project has been completed with promising results and a strong understanding of the analysis. Continued work and improvements on the 3D analysis and code will provide more realistic optimizations and should allow Space Systems / Loral to directly apply MITEOR to the development of their next-generation GEO satellites.