https://ae.gatech.edu/event/phd-defense-tristan-sarton-du-jonchay

Abstract

From telecommunications to weather monitoring, Geosynchronous (GEO) satellites represent a critical infrastructure supporting a multitude of terrestrial markets. This, however, comes at the cost of large capital expenditures to manufacture, insure, and launch these large spacecrafts to their remote orbits. Until recently, the traditional paradigm to maintain and upgrade this infrastructure consisted of replacing the outdated satellites with new assets designed to last 15 to 20 years. However, with the advent of On-Orbit Servicing (OOS), this paradigm will soon change.
OOS is a nascent space-based industry aimed at making the operations and management of Earth-orbiting satellites sustainable. Two space systems central to OOS infrastructures are servicers and orbital depots. Servicers are robotic spacecraft providing services (e.g., refueling) to client satellites. Orbital depots are in-space warehouses storing commodities (e.g., spares, propellant) to support the long-term operations of the servicers and client satellites. With a growing interest in designing and deploying OOS infrastructures comes the need to plan their long-term operations as well as the supply chain of commodities needed to support them.
This dissertation presents a set of methods that model and simulate the operations of large OOS infrastructures dedicated to the servicing of GEO satellites with uncertain needs. The second chapter demonstrates how Discrete Event Simulation (DES) can be used to explore simple operational concepts with relatively small design tradespaces. For larger ones, however, a more automated design method is needed. To fulfill this need, the third chapter introduces a novel simulation framework that simultaneously makes decisions regarding the operations (e.g., servicers’ routing and refueling) and the supply chain (e.g., re-supply of depots) of OOS infrastructures. The first contribution presented in the third chapter is to develop a Time-Expanded Generalized Commodity Network Flow (TE-GMCNF) model with the inherent capacity to allocate service tasks optimally. Furthermore, the third chapter leverages the Rolling Horizon (RH) procedure to embed service demand uncertainties within the TE-GMCNF model. The fourth chapter generalizes the framework introduced in the third chapter by modeling the relative dynamics of the nodes of the OOS network and different propulsive options for the servicers, such as impulsive thrust, low thrust, or both.

Committee
• Dr Koki Ho – School of Aerospace Engineering (advisor)
• Dr Glenn Lightsey – School of Aerospace Engineering
• Dr Sandra Magnus – School of Aerospace Engineering
• Dr Mark Whorton – School of Aerospace Engineering
• Dr Paul Grogan – School of Systems and Enterprises, Stevens Institute of Technology