Organizational Unit:

Aerospace Systems Design Laboratory (ASDL)

Permanent Link

https://hdl.handle.net/1853/70744

Parent Organization

Organizational Unit

Daniel Guggenheim School of Aerospace Engineering

ArchiveSpace Name Record

https://finding-aids.library.gatech.edu/agents/corporate_entities/1135

Full item page

Publication Search Results

Now showing 1 - 10 of 121

Accelerated Simulation-Based Analysis of Emergent and Stochastic Behavior in Military Capability Design

(Georgia Institute of Technology, 2023-07-25) Braafladt, Alexander

In military capability design, the United States Air Force (USAF) is working to modernize to be ready to succeed in future operations. During the process, high-fidelity military simulation is used iteratively to build up understanding of complex military scenarios and consider technology and concept alternatives. While high-fidelity simulation is critical to the analysis, it is often expensive and time consuming to work with. In addition, the required pace for analysis needs to be accelerated as technology and threats rapidly evolve. In response to these challenges, the research in this thesis focuses on accelerating two central parts of simulation-based analysis in capability design. The first part focuses on improving methods for searching for emergent behavior, which is critical for building up understanding with simulation. The second part focuses on including stochastic responses from simulation in parametric models used during tabletop design exercises, which are critical for comparing alternatives. To accelerate simulation-based analysis of emergent behavior, a specific definition of emergent behavior is synthesized from the literature that prompts optimization approaches to be used for searching more quickly than with brute-force Monte Carlo Simulation (MCS). This definition also allows formulation of the new ENFLAME (Exploration of Nonlinear and stochastic Future behavior under Lack of knowledge using simulation-based Analysis to Manage Emergent behavior) framework for structuring activities working to manage emergent behavior with simulation-based analysis. Specifically in this work, the new LANTERN (Low-cost Adaptive exploratioN to Track down Extreme, Rare events using Numerical optimization) methodology for searching for emergent behavior as rare, localized and stochastic extreme events is developed that accelerates the process using novel Bayesian Optimization (BO) techniques that adaptively query the simulation to find rare events. In experiments with test problems based on the behavior expected with an Agent-Based Modeling (ABM) simulation approach, the new BO techniques show significant improvement over MCS. For accelerating analysis of stochastic behavior during tabletop design exercises, the ECDF-ROM surrogate modeling approach that uses Reduced-Order Modeling (ROM) techniques combined with a new field representation is developed. The surrogate modeling approach is shown to work effectively with distributions like those expected with military simulation, allowing parametric, interactive queries of distributions. A final demonstration of the techniques was completed using two scenarios developed in simulation with the Advanced Framework for Simulation, Integration, and Modeling (AFSIM). First, a Suppression of Enemy Air Defenses (SEAD) scenario was used to demonstrate the effectiveness of the new techniques at searching for rare, localized extreme events. Second, a four vs. four air combat scenario was used to demonstrate the effectiveness of the new technique for searching for rare, stochastic extreme events, and to demonstrate the new distribution surrogate modeling approach. The results together show that the LANTERN methodology accelerates the search for emergent behavior effectively for iterative simulation-based analysis of military scenarios and the ECDF-ROM approach enables parametric models of stochastic outcomes.
Rotorcraft takeoff analysis and classification to detect outlier operations that could present a safety risk

(Georgia Institute of Technology. School of Aerospace Engineering, 2023-06) da Silva, Ricardo F. ; Achour, Gabriel N. ; Payan, Alexia P. ; Johnson, Charles ; Mavris, Dimitri N.

Various reports from entities such as the Federal Aviation Administration (FAA) and the National Transportation Safety Board (NTSB) have shown a recent increase in the number of incidents involving helicopters. The versatility of rotorcraft operations makes the establish ment of safety metrics challenging. Yet, flight data monitoring (FDM) programs enable the implementation of data-based models and analyses that can contribute to improving the safety of helicopter operations. Traditionally FDM programs have featured exceedance-based data analyses by defining safety thresholds. However, recent advances in data science, and more particularly in deep learning techniques, have paved the way for a more reliable definition of safety thresholds via the use of outlier detection algorithms. This paper focuses on the implementation of an anomaly detection model for the takeoff phase which represents a large portion of incidents in rotorcraft operations. After generating training data and augmenting the dataset, the takeoff segment is extracted from each flight data record. Then, the type of takeoff performed is identified through a classification algorithm, and finally, a recurrent neural network composed of long short term memory cells is implemented to detect anomalies or outliers in the input takeoff data.
Development of a Parametric Drag Polar Approximation for Conceptual Design

(Georgia Institute of Technology. School of Aerospace Engineering, 2023-06) Sampaio Felix, Barbara ; Perron, Christian ; Ahuja, Jai ; Mavris, Dimitri N.

The present work proposes an efficient parametric approximation of mission drag polars by combining multi-fidelity surrogate models with parametric reduced order modeling techniques. Traditionally, semi-empirical aerodynamic analyses are used to provide drag polars needed for mission analysis during the conceptual design of aircraft. The database needed for these methods is unavailable for unconventional vehicles, and for this reason, many studies rely on higher-fidelity models typical of preliminary design to perform design space exploration for novel vehicle geometries. Due to the high computational cost and evaluation time of these higher-fidelity models, researchers constrain the design space exploration of vehicles by either relying on single discipline optimization or obtaining mission drag polars for a few vehicle geometries within their design loop. The present work demonstrates the application of Hierarchical Kriging surrogate models to obtain mission drag polars for fixed vehicle geometries. Then, the proper orthogonal decomposition reduced order model with Kriging interpolation is used to approximate the coherent structure of mission drag polars. The proposed method is demonstrated on a supersonic commercial aircraft. Experiments showed that both the multi-fidelity surrogate model and the reduced order model are able to emulate vehicle drag polars well for fixed and varying vehicle geometries, respectively.
Aviation-BERT: A Preliminary Aviation-Specific Natural Language Model

(Georgia Institute of Technology, 2023-06) Chandra, Chetan ; Jing, Xiao ; Bendarkar, Mayank ; Sawant, Kshitij ; Elias, Lidya R. ; Kirby, Michelle ; Mavris, Dimitri N.

Data-driven methods form the frontier of reactive aviation safety analysis. While analysis of quantitative data from flight operations is common, text narratives of accidents and incidents have not been sufficiently mined. Among the many use cases of aviation text-data mining, automatically extracting safety concepts is probably the most important. Bidirectional EncoderRepresentations from Transformers (BERT) is a transformer-based large language model that is openly available and has been adapted to numerous domain-specific tasks. The present work provides a comprehensive methodology to develop domain-specific BERT model starting from the base model. A preliminary aviation domain-specific BERT model is developed in this work. This Aviation-BERT model is pre-trained from the BERT-Base model using accident and incident text narratives from the National Transportation Safety Board (NTSB) and AviationSafety Reporting System (ASRS) using mixed-domain pre-training. Aviation-BERT is shown to outperform BERT when it comes to text-mining tasks on aviation text datasets. It is also expected to be of tremendous value in numerous downstream tasks in the analysis of aviation text corpora.
BERT for Aviation Text Classification

(Georgia Institute of Technology, 2023-06) Jing, Xiao ; Chennakesavan, Akul ; Chandra, Chetan ; Bendarkar, Mayank ; Kirby, Michelle ; Mavris, Dimitri N.

The advent of transformer-based models pre-trained on large-scale text corpora has revolutionized Natural Language Processing (NLP) in recent years. Models such as BERT (Bidirectional Encoder Representations from Transformers) offer powerful tools for understanding contextual information and have achieved impressive results in numerous language understanding tasks. However, their application in the aviation domain remains relatively unexplored. This study discusses the challenges of applying multi-label classification problems on aviation text data. A custom aviation domain specific BERT model (Aviation-BERT) is compared against BERT-base-uncased for anomaly event classification in the Aviation Safety Reporting System (ASRS) data. Aviation-BERT is shown to have superior performance based on multiple metrics. By focusing on the potential of NLP in advancing complex aviation safety report analysis, the present work offers a comprehensive evaluation of BERT on aviation domain datasets and discusses its strengths and weaknesses. This research highlights the significance of domain-specific NLP models in improving the accuracy and efficiency of safety report classification and analysis in the aviation industry.
Formalizing the Decomposition Process Between Elements in the RFLP Framework Using Axiomatic Design Theory

(Georgia Institute of Technology. School of Aerospace Engineering, 2023-06) Omoarebun, Ehiremen Nathaniel ; Cimtalay, Selçuk ; Mavris, Dimitri N.

As interactions and communication between elements in engineered systems continue to increase, the need to manage complexity, mitigate risks and improve human understanding of a system behavior becomes important. Over the years, systems engineering has emerged as a way to address the design of complex systems. However, traditional systems engineering comes with its own limitations, and this has led to the emergence of Model-Based Systems Engineering (MBSE) which aims to provide a better solution of design of complex systems. A lot of MBSE approaches are still based on heuristics and sometimes there is no clear structure on the process of design, especially during product and process decomposition. This paper aims to address that gap through the introduction of a formal structure by integrating concepts from Axiomatic Design Theory with the Requirement, Functional, Logical and Physical (RFLP) framework used in MBSE. The axioms from axiomatic design guide the decision-making process and the zigzagging aspect of the theory aids in the development of a structure during design. This process also aids in the identification and mitigation of potential coupling in design. A coupled spring-mass-damper system will serve as a demonstration to verify the proposed approach
Design Space Reduction using Multi-Fidelity Model-Based Active Subspaces

(Georgia Institute of Technology, 2023-06) Mufti, Bilal ; Perron, Christian ; Gautier, Raphaël ; Mavris, Dimitri N.

The parameterization of aerodynamic design shapes often results in high-dimensional design spaces, creating challenges when constructing surrogate models for aerodynamic coefficients. Active subspaces offer an effective way to reduce the dimensionality of such spaces, but existing approaches often require a substantial number of gradient evaluations, making them computationally expensive. We propose a multi-fidelity, model-based approach to finding an active subspace that relies solely on direct function evaluations. By using both high- and low-fidelity samples, we develop a model-based approximation of the projection matrix of the active subspace. We evaluate the proposed method by assessing its active subspace recovery characteristics and resulting model prediction accuracy for airfoil and wing drag prediction problems. Our results show that the proposed method successfully recovers the active subspace with an acceptable model prediction error. Furthermore, a cost vs. accuracy comparison with the multi-fidelity gradient-based active subspace method demonstrates that our approach offers comparable predictive performance with lower computational costs. Our findings provide strong evidence supporting the usage of the proposed method to reduce the dimensionality of design spaces when gradient samples are unavailable or expensive to obtain.
Methodology for Aircraft Architecture Selection & Design Optimization

(Georgia Institute of Technology, 2023-04-30) Harish, Anusha

Growing concerns about the environment have led to aviation agencies around the world such as IATA, ICAO, NASA, and ACARE to set targets to curb noise, emissions and fuel consumption in the coming years. In order to achieve these goals, several new aircraft technologies and concepts in the areas of unconventional airframe configurations (such as blended wing body and truss-braced wing), advanced propulsion systems (example, open rotor engines and electrified propulsion), alternative energy sources (such as hydrogen, battery, etc.) as well as propulsion-airframe integration concepts (distributed propulsion, boundary layer ingestion, etc.) have been proposed. There are over 100,000 possible combinations of these technologies. However, this vast architecture space has not yet been fully explored. Therefore, there is a need for a lower-order analysis methodology capable of rapidly analyzing different combinations. This research aims to propose a methodology for rapid generation and assessment of architectures in order to identify promising ones that are capable of meeting future environmental goals. There are 3 key aspects to this problem - generation of alternatives, evaluation of the design space for the architectures, and finally the optimization of the aircraft designs. The first research area focuses on the generation of architecture alternatives using Constraint Programming for every aircraft configuration with known propulsive-airframe integration concept, given the compatibility between different components. Since there is currently no methodology that automatically generates architecture alternatives, this proposed methodology is validated by comparing its results against known or studied architectures in the literature. The second research area is aimed at developing a ”pre-conceptual” design methodology that can quickly evaluate and optimize architecture alternatives with fewer design details and consistent set of assumptions and requirements. Parameters such that the energy and the power split between different components, and the path for power flow from the energy source to the thrust producing device at both sizing points as well as throughout the mission segments are proposed and used in the determination of key performance indicators such as global chain efficiency, energy specific air range and thrust specific power consumption. The objective of the final research question is the optimization of the aircraft design for each generated architecture. A multi-objective optimization algorithm is implemented to optimize each design with aircraft weight and energy consumption as the two objectives, while meeting all aircraft requirements such as range, payload, cruise altitude and speed, mission power requirements, etc. Thus, a complete, generalized, universal architecture enumeration and pre-conceptual design and optimization methodology is proposed. The capability of this methodology is demonstrated in the final use case where architectures with different alternatives in terms of energy sources – jet fuel, batteries (high specific power, high specific energy) and hydrogen; and advanced propulsion system architectures with distributed propulsion – electrified propulsion and hydrogen propulsion hybrids, are generated, evaluated and optimized for a 2050 Entry-into-Service. Furthermore, the impact of technologies on the aircraft performance is investigated through a technology sensitivity study.
Implementing the Digital Thread - A Proof-of-Concept

(Georgia Institute of Technology, 2023-01-25) Oroz, Juan ; Roohi, Zayn ; Abelezele, Sabastian ; Fronk, Gabriel ; Al Fawares, Ruby ; Pinon-Fischer, Olivia ; Gharbi, Aroua ; Marvis, Dimitri ; Petersen, Melissa ; Karl, Alexander ; Matlik, John ; Schwering, Bryan

Current engineering processes are heavily document-centric, which can add time and cost to projects, limit data accessibility, and make it difficult to actively manage models and data consistency throughout the lifecycle of a product. Additionally, traditional data siloes limit data accessibility across departments. Similar issues exist with tools: departments use software with different standards and formats, making it time-consuming and difficult to accurately propagate changes and requirements throughout. Aerospace projects and vehicles are also often a level of magnitude more complex than products developed in other industries, requiring the coupling of multiple disciplines, which intensifies these problems. Digital Engineering and Model-Based Systems Engineering (MBSE) provide the context, methodologies and tools to address some of the aforementioned challenges. In particular, this paper presents the development and implementation of a Digital Thread proof-of-concept for a minimum viable product. In doing so this research demonstrates solutions to the challenges of data acquisition and management, model and data connectivity, tool, and platform integration, eventually leading to the realization of an authoritative source of truth across the product’s lifecycle. Additionally, this research highlights some of the key benefits brought about by the Digital Thread, which include increased collaboration and communication, managed consistency across models and data, as well as the ability to conduct model verification, validation, and calibration - an important tenet of MBSE.
Development of a Parametric Structural Analysis Environment to Support the Design, Manufacturing, and Production of a Composite UAV Wing

(AIAA, 2023-01) dos Santos, Marcos ; Cox, Adam ; Fischer, Olivia J. Pinion ; Mavris, Dimitri N.

As digital and physical systems become more complex, collect voluminous quantities of data, and move towards greater integration over time, the concept of digital engineering has become of great interest in the engineering world. The integration of digital methods with traditional engineering approaches in product lifecycle management has posed challenges on how techniques such as digital twins can be best used during the design and manufacturing phases of the product lifecycle. To address this need, this research supports the integration of design, manufacturing, and production by assessing the structural integrity of various designs of a parametric UAV wing built with a composite material. A systematic and efficient environment is developed to modify the wing design parameters, develop and analyze the finite element model, obtain structural data, and identify feasible design regions for decision making. The sharing of models, data, and analyses with the manufacturing and production segments of the lifecycle permits integration of the various disciplines in early design phases to allow greater design freedom and avoid great costs during the design of the product. The results indicate that (1) the need for a trade-off analysis between key disciplinary considerations in UAV wing design decision making can be addressed and that (2) the developed capability enables decision makers to choose the configurations to be studied in later design stages after the structural integrity and weight considerations are assessed for multiple wing designs.