A Comprehensive Analysis of Lunar Surface System Architectures
- Paper number
IAC-09.A3.2INT.23
- Author
Mr. Arthur Guest, Massachussets Institute of Technology (MIT), United States
- Year
2009
- Abstract
I. Introduction In the 2004 Vision for Space Exploration, President George W. Bush laid out a new direction for U.S. space exploration that included returning humans to the Moon by 2020 [1]. To this end, NASA began work towards establishing a program to develop a sustained human presence on the lunar surface. The reasoning for this program, as laid out in the Global Exploration Strategy [2], includes objectives ranging from exploration preparation and scientific knowledge to economic expansion and public engagement. The architecture for the lunar exploration program will require two interconnected systems to be developed: the lunar transportation system, which delivers crew and cargo to the Moon and returns crews back to the Earth, and the lunar surface system, which includes all assets required to allow sustained crewed operations on the surface. To ensure that the program provides robust return across the set of objectives and for a range of scenarios, it is necessary to perform a comprehensive analysis of the architecture space for the lunar surface systems. II. Architecting the Lunar Surface System This paper applies a framework for representing the feasible architectures of the lunar surface system known as the Architecture Decision Graph (ADG), which effectively represents a systems architecture search space as a set of interconnected decision variables [3]. In ADG, system architectures are viewed as a decision problem, which is represented as a collection of two types of nodes: system variables and relations. System variables consist of decision variables, which are controlled by the decision-maker, and property variables, which are derived from the calculation of metrics. The domain of a decision variable is a finite set of alternatives that can be assigned to that variable. The relations between system variables are either logical constraints, which are propositional statements that specify the feasible assignments to two or more decision variables, or property functions that specify a metric function to calculate property variables. The architecture search space for the lunar surface system is represented by a set of forty-two decision variables grouped into six classes of decisions related to specific functions required for the system. The six classes of decisions are: • Human habitation (pressurized volume, material type, number of elements) • Crewed surface mobility (type of mobility elements, pressurized and unpressurized) • Power Generation and Storage (stationary and mobile power generation and storage assets) • Communications (choice of communications assets) • Logistics (resupply choice, container size, ISRU) • Infrastructure Construction (offloading, transporting, and assembling) The decision variables are connected to each other through logical constraints developed through assessing the physical limitations of the architectures, logical reasoning about the architectures, and through external requirements placed on the system. The connections between the decision variables and logical constraints can be seen in the constraint graph shown in Figure 1. In choosing the decision variables, it is important to recognize the boundaries of the system. For this analysis, both the lunar transportation system and the exploration/scientific payloads are considered outside of the system boundary. The lunar transportation system places constraints and limitations on the lunar surface system based on its capabilities, which are assumed to be consistent with NASA’s overview given at the 3rd Exploration Conference [4]. The exploration/scientific payloads are placed outside the boundaries of the system in order not to constrain the objectives of the overall exploration architecture. Instead, one of the metrics for the lunar surface system is maximizing the amount of available space for payload mass by minimizing the mass of the lunar surface system. Another important aspect of developing the architecture for the lunar surface system is incorporating the dynamic nature of the system. In order to account for this, the architecture is developed as a series of phases and the ADG framework is applied to each phase until an entire scenario is developed. Each phase is located at a specific location on the lunar surface and consists of a deployment sub-phase and a utilization sub-phase. Follow-on phases can exist at locations after the utilization of a preliminary phase. These phases will build on and utilize the assets already deployed at that location. The property variables have been chosen to capture both the performance and cost of the lunar surface system. Because of the challenge of obtaining accurate cost-models, the cost of the system has been captured by a set of six property variables meant to act as proximate metrics for the development, production, and deployment costs. These variables focus on the number of unique elements developments, the number of total elements produced, and the number of launches required. To capture the performance of the system, property variables were chosen to capture mission duration (both cumulative and maximum single crew duration), exploration range number of locations visited, and the available mass capacity for exploration/scientific payloads. III. Analysis of the Lunar Surface System The results from the ADG analysis of the lunar surface system are viewed using two types of figures. Pareto Front Views, similar to the one shown in Figure 2, allow an architect to view the dominant and non-dominant architectures. Initial results have shown that along with NASA’s currently envisioned single outpost located at the South Pole, other interesting architectures include deploying several shorter duration outposts at multiple locations or developing outposts that do not include assembling outposts from multiple habitation modules. The second type of figure created using the ADG framework is the Decision Space View. This view allows architects to determine which decision variables strongly impact the property variables and which decision variables are highly connected throughout the architecture. For example, the choice of total habitable pressurized volume is highly connected in the architecture and greatly impacts the cost and performance property variables. System architects will want to address these highly connected, highly influential decisions first and focus their analysis capabilities on these decisions. IV. Conclusions Performing a comprehensive analysis of the lunar surface system architecture space is necessary in order to choose robust lunar exploration architectures. The ADG framework is used to represent the search space as a decision problem consisting of 42 decision variables and ten property variables. By developing phases, consisting of deployment and utilization, a comparison of architectures with single and multiple outpost locations was completed. The results of the analysis are presented using Pareto Front Views and Decision Space Views that allow a system architect to not only determine the dominant architectures, but also identify the most important decisions related to the lunar surface system. V. References [1] Bush, G.W. “A Renewed Spirit of Discovery: The President’s Vision for U.S. Space Exploration”. January 14, 2004. Washington, DC [2] NASA et al. “Global Exploration Strategy: The Framework for Coordination” May 31, 2007 [3] Simmons, W.L. “ A Framework for Decision Support in Systems Architecting”. Massachusetts Institute of Technology. June 2008. Cambridge, MA [4] Dorris, C. “Altair Project”. 3rd Space Exploration Conference. February 27, 2008. Denver, CO.
- Abstract document
- Manuscript document
(absent)