Person:

Robertson, Bradford E.

Permanent Link

https://hdl.handle.net/1853/71635

Associated Organization(s)

Organizational Unit

Daniel Guggenheim School of Aerospace Engineering

The Daniel Guggenheim School of Aeronautics was established in 1931, with a name change in 1962 to the School of Aerospace Engineering

Full item page

Publication Search Results

Now showing 1 - 10 of 12

Uncertain Reduced Order Model Predictions of an Unsteady Field

(Georgia Institute of Technology, 2025-01) Willier, Brenton ; Perron, Christian ; Robertson, Bradford E. ; Mavris, Dimitri N.

The design of blunt-body entry vehicles balances atmospheric heating and drag to ensure payload safety during entry, descent, and landing. Their blunt shape creates turbulent, recirculating wakes, leading to uncertain flight paths and challenging mitigation strategies. Traditionally, uncertainties are managed using conservative scalars and multipliers, resulting in over-engineered designs with reduced payload capacity and less accurate landings. While advances in computational fluid dynamics (CFD) enable high-fidelity analysis, the cost of extensive simulations remains prohibitive. Surrogate models, such as reduced-order models (ROMs), offer a faster alternative but must address the uncertainty introduced by unsteady aerodynamic training data. This paper presents a methodology to capture, encode, and propagate uncertainty in unsteady high-dimensional fields using parametric ROMs. The approach employs a Lorenz model to emulate unsteady fields and applies replication sampling to capture full-order model (FOM) nonparametric uncertainty. Proper orthogonal decomposition (POD) reduces dimensionality, and sparse Kriging regression predicts latent space mean and variance. A linear covariance back-mapping technique is applied to propagate uncertainty from the latent coordinates to integrated scalar coefficients. Results demonstrate the ROM accurately predicts scalar uncertainties consistent with FOM validation, supporting future application to more complex problems, such as entry vehicles in unsteady free-flight simulations.
Sensitivity Analysis of Lunar PNT for Surface Users with Nonuniform Ranging Error

(Georgia Institute of Technology, 2025-01) Gabhart, Austin ; Broering, Alex ; Robertson, Bradford E. ; Mavris, Dimitri N.

As operations in cislunar space and the lunar economy develop, there is a pressing need for accurate and reliable Position, Navigation, and Timing (PNT) support on and around the Moon. This paper seeks to provide insight into the primary contributions to positioning error on the lunar surface. To improve the estimations of error, an alternative metric, user equivalent range error weighted geometric dilution of precision (KDOP), is presented without the assumption that ranging error is constant across systems. Parameters related to user error, onboard clock stability, and orbit determination performance are varied to find and quantify monotonic relationships between these design parameters and performance metrics. It is found that KDOP is affected by both clock stability and orbit determination support performance. This is similar to the results for user equivalent range error except, in some cases, it is of the opposite sign.
A Comparative Analysis of Lunar PNT+C Concepts

(Georgia Institute of Technology, 2025-01) Gabhart, Austin ; Drosendahl, Madilyn ; Robertson, Bradford E. ; Mavris, Dimitri N.

With the expected increase in cislunar operations, there are many proposed architectures to provide these missions with Positioning, Navigation, Timing, and Communication (PNT+C) support. While most of the architectures are frequently similar in concept, there is significant variation in how they are evaluated. This paper provides an overview of system architecture alternatives presented in literature to develop requirements for an evaluation environment. To provide insight into the relative performance between architectures, an evaluation environment is described in detail to establish the basis for comparison. The evaluation environment includes orbit determination and time synchronization modeling, an implementation of more accurate dilution of precision metrics, and orbit maintenance maneuver modeling and optimization. The configurations for comparison are selected from literature. It is found that constellations made of highly stable elliptical orbits still require significant ΔV to maintain their orbits. It is also found that increasing the variety of orbits and number of satellites significantly improves the performance of the lunar PNT architecture.
Trade Studies for Evaluating Lunar Surface Power Architectures

(Georgia Institute of Technology, 2025-01) Thakar, Bhargavi ; McNabb, Jeffrey ; Robertson, Bradford ; Mavris, Dimitri N.

It is necessary to have a flexible power architecture for the implementation of a sustainable lunar base at the lunar South Pole. Several power architectures alternatives consisting of several technology options have been proposed. However, there is a lack of a comprehensive trade analysis of comparative sizing for those options. This paper presents a comparative evaluation of surface and elevated solar panels for power generation, DC cables, AC cables, and power beaming for power distribution, and batteries and fuel cells for energy storage, which helps understand the configuration for power architectures that provide the most optimal solution with regards to landed mass. The comparison is done by using illumination profiles of specific lunar sites and evaluating that at differing heights above the ground. The method of combining time-dependent illumination data with static system sizing is detailed in the paper. The case study compares the different architectures that are possible using all combinations of the technologies that are discussed. This study evaluated each architecture across a wide range of customer power demands, transmission distances, and transmission voltages. Elevated solar panels and regenerative fuel cells were found to be the best option across all cases. The case study determined that there is no best-in-class power distribution system and that each one has portion of the design space in which they are most mass efficient.
Trade Studies on Lunar Navigation Satellite System Architectures

(Georgia Institute of Technology, 2025-01) da Silva, Ricardo F. ; Robertson, Bradford E. ; Mavris, Dimitri N.

The global priority for space exploration is shifting towards lunar exploration, driven by ambitious efforts from both government agencies and commercial ventures. The Artemis program, led by NASA, represents a significant step in this direction, aiming to re-establish human presence on the Moon and pave the way for future missions to Mars through collaboration with international partners. The European Space Agency (ESA) is a key partner in the Artemis program, and it is also actively engaged in its own lunar exploration initiatives, including the Terrae Novae program and the Moonlight Programme, this last one focusing on innovative Communication and Navigation services near the Moon. Additionally, there is a growing number of planned lunar missions, making Position, Navigation, and Timing (PNT) services critical for mission success. This study's objective is to expand on PNT solutions for lunar navigation, review in detail the PNT service requirements, and conduct trade studies considering the navigation performance of lunar navigation satellite systems for users on the surface of the Moon, users in orbit (including Low Lunar Orbit and Near-Rectilinear Halo Orbit) and landing on the Moon, to assess the compatibility of these systems with NASA's requirements, outlining the design process and methodology for achieving this goal. The results allow for the identification of the most sensitive elements that influence the design process, helping to find the most appropriate solution to meet the requirements of PNT services.
Modelling the Geometric Configuration of a Spacecraft in Dyreqt

(Georgia Institute of Technology, 2025-01) Robertson, Bradford E. ; Coleman, Justin P. ; Mavris, Dimitri N.

The Dyreqt framework is a Python tool for the conceptual design of spacecraft. It allows a user to combine custom models of spacecraft subsystems with a mission in order to rapidly size a vehicle and perform trade studies. However, Dyreqt currently has no built-in method for representing the geometric configuration of a spacecraft, limiting its ability to consider this in analysis. This paper introduces a framework that allows for the geometric modelling of an arbitrary spacecraft within Dyreqt. The framework utilizes the methods of primitive instancing and constructive solid geometry to map each Dyreqt subelement to a physical body. This representation can then be passed to a set of models which calculate geometric properties, allowing new information to be computed within Dyreqt, including bounding dimensions, moments of inertia, and projected area. These in turn enable Dyreqt to utilize higher fidelity subsystem models, to include geometric constraints in optimization, and to produce three-dimensional visuals of its results. These capabilities were demonstrated by using the framework to compare different potential configurations for a lunar cargo lander.
A Reduced Order Modeling Approach to Blunt-Body Aerodynamic Modeling

(Georgia Institute of Technology, 2024-01) Dean, Hayden V. ; Decker, Kenneth ; Robertson, Bradford E. ; Mavris, Dimitri N.

Blunt-body entry vehicles display complex flow phenomena that results in dynamic instabilities in the low supersonic to transonic flight regime. Dynamic stability coefficients are typically calculated through parameter identification and trajectory regression techniques using both physical test data and Computational Fluid Dynamics (CFD) simulations. This methodology can generate dynamic stability coefficients, but the resulting data points are limited, and have high degrees of uncertainty due to the nature of data reduction methods. With increased computational capabilities, new methods for dynamic stability quantification have been explored that seek to leverage the high-dimensional aerodynamic data produced from CFD simulations to compute dynamic stability behavior and address the limitations of linearized aerodynamics. The objective of this work is to advance the quantification of dynamic stability behavior of blunt-body entry vehicles by leveraging high-fidelity CFD data through Reduced Order Modeling (ROM). ROMs are capable of leveraging high-fidelity aerodynamic data in a cost effective manner by finding a low-dimensional representation of the Full Order Model (FOM). ROMs based on Proper Orthogonal Decomposition (POD) have shown success in recreating CFD analyses of parametric ROM applications and time-varying ROM applications. Results of this research demonstrated success in constructing two ROMs of a notional blunt-body entry vehicle to recreate heatshield and backshell pressure distributions from forced oscillation trajectories. The ROM was more successful at reconstructing the heatshield pressure distribution, with challenges arising in predicting the chaotic response of backshell latent coordinates.
Optimal Deployment Strategies for Cislunar PNT+C Architectures

(Georgia Institute of Technology, 2024-01) Gabhart, Austin ; Drosendahl, Madilyn ; Robertson, Bradford E. ; Steffens, Michael J. ; Mavris, Dimitri N.

Cislunar operations are expected to rise dramatically within the next decade, requiring a comparable increase in PNT and communications services. However, current PNT systems are at capacity and need to be augmented to serve a cislunar space domain, specifically in the form of novel cislunar PNT architectures. This paper studies the problem of the deployment of PNT and communications satellites, specifically, the problem of deployment strategies spanning multiple stages over extended periods of time. A set of stage definitions will be determined along with areas of potential user activity. A novel application of the hidden gene genetic algorithm to the constellation optimization problem is presented. A design space exploration is presented with comparisons of circular and elliptical constellations. Optimization results from the first stage are also provided. It is shown that acceptable performance can be achieved with a low number of deployed satellites and that strong trade-offs exist between performance and stability.
Unsteady Aerodynamic Uncertainty Quantification of a Blunt-Body Entry Vehicle in Free-Flight

(Georgia Institute of Technology, 2024-01) Willier, Brenton J. ; Hickey, Alexandra M. ; Robertson, Bradford E. ; Mavris, Dimitri N.

The design process of blunt-body entry vehicles balances atmospheric heating and drag to ensure crucial payloads can safely traverse through entry, descent, and landing. However, the blunt shape leads to a chaotic recirculating wake. Currently, uncertainties in the vehicle design process are captured through scalars and multipliers, and these conservative estimations lead to over-engineered vehicles, reduced payload capacity, and less accurate landings. To supplement the data gathered through physical testing, CFD-in-the-loop free-flight trajectories can be simulated throughout the flight regime. While CFD performance has improved significantly, the number of cases required to produce a meaningful sample for an uncertainty analysis remains computationally intense. Parametric uncertainty can be captured with traditional uncertainty methods like Monte Carlo analysis. However, the non-parametric uncertainty due to the unsteady nature of the chaotic wake has yet to be studied for free-flight analysis. This paper presents and implements an ensemble sampling initialization approach to determine the impact of unsteady wake structures imparted on CFD-in-the-loop data produced using replicated trajectory simulations. To enable this data generation, the Genesis vehicle gridding process is detailed, along with an overview of the free-flight CFD simulation setup for a supersonic flight regime. While running a static unsteady simulation, ten flow fields were saved at various times to capture different instantaneous structures in the wake. After initializing identical free-flight simulations with the ten different flow fields, results of vehicle aerodynamic angles, aerodynamic force and moment coefficients, inertial velocity, and vehicle trajectory in multiple reference frames showed identifiable trends with diverging behavior. The uncertainty on these variables due to unsteady flow is also quantified throughout the motion. It is concluded that this aspect of uncertainty must be carefully considered when CFD-in-the-loop is used to model the flight characteristics of a blunt-body vehicle.
Analysis of Infrastructure to Support a Future Space Economy

(Georgia Institute of Technology, 2024-01) Roohi, Zayn A. ; Robertson, Bradford E. ; Mavris, Dimitri N.

Beginning with the Artemis-I mission in late 2022, NASA is embarking upon a series of increasingly complex missions to establish a permanent presence on the surface of the Moon, potentially leading to manned Mars missions within the next few decades. Several private companies have also announced that they have begun work on space tourism projects with the goal of launching within this same time-frame. Supporting this expansion will require advanced space logistics and the development of dedicated space-based supply chains in order to reduce cost and increase resiliency. Previous research has focused on studying the impact that a specific technology, vehicle, or type of infrastructure has on supporting a single space campaign or mission; this paper takes a wider view by examining the impact that several types of infrastructure concepts together will have on the entire set of operations that could take place within the next decade. Lunar in-situ resource utilization, space depots, and space tugs are considered as infrastructure concepts, and a Lunar space station, Lunar habitat, Earth space stations, and Mars missions are considered as the operations to support. A time expanded mixed-integer nonlinear programming model is used to solve traditional network flow and supply chain problems, the results of which are used to propose future resupply missions and supply chain architectures.