Barbara Sampaio Felix
(Advisor: Prof. Dimitri N. Mavris)

will propose a doctoral thesis entitled,

A Methodology for Design Space Exploration of Novel supersonic Aircraft Using High-Fidelity Aerodynamic Analysis

On

Wednesday, February 15th at 2:30 p.m. EST

Weber Space Science and Technology Building (SST II)

Zoom meeting: https://gatech.zoom.us/j/99036097632?pwd=MW52aktkWG1MNGszYW0xZFBlNWRwZz09

Meeting ID: 990 3609 7632

Passcode: 866605

Abstract

Aerospace industries and research institutions have shown an increased interest in the design of novel commercial Supersonic Transport (SST) as the market share for this vehicle class has increased. Due to performance limitations shown by the past SSTs, the design of future supersonic vehicles is a clean sheet design and not an improvement upon past configurations. Following the traditional aircraft design process, the first step in the early vehicle development phase of a new aircraft is Design Space Exploration (DSE) and aircraft sizing of vehicle geometries using a predetermined design mission. Within the discipline of aerodynamics, designers ideally want to evaluate the vehicle characteristics for the purpose of performance, handling qualities and loads in every orientation. It is desirable to predict the distribution of pressure and shear around a vehicle body, and then obtain their integrals and derivatives. Nonetheless, only the lift and drag coefficients of the vehicles are required for the initial evaluation of mission analysis. During this early design phase, it is common to use lower fidelity models to quickly estimate vehicle aerodynamic characteristics given a limited definition of the aircraft geometry. These codes rely on historical data to guarantee the accuracy of the results. And without an abundance of data representing the vehicle geometry in question, there is a large risk associated with making design decisions based on these data-based models. To avoid this problem designers have sought novel ways to use higher fidelity models for aircraft conceptual design. Nonetheless, a detailed vehicle geometry is required to use these higher fidelity models, which adds complexity to the DSE step. The additional geometry design parameters increase the number of samples required to cover the design space. Furthermore, the higher the fidelity of the model, the larger the computational burden and designer effort associated with it. For these reasons, the literature currently lacks a methodology to obtain a parametric representation of the aerodynamic tables required for SST design at the proper level of fidelity and within a reasonable development time.

The objective of this thesis is to enable the DSE using mission analysis of SSTs with higher fidelity aerodynamic analysis models by leveraging the recent developments of two key enablers: data-fusion techniques and Reduced Order Models (ROM). The proposed methodology is a two-step process that first uses data fusion techniques to generate a multi-fidelity drag polar for a given vehicle geometry, and then builds a parametric representation of this table using ROM. The first step of this research will define the flight conditions that must be present in the mission drag polar for the given design mission. Furthermore, it will determine the appropriate fidelity of the aerodynamic models used to generate the mission drag polar. The second step of the methodology uses ROM techniques to approximate the mission drag polar with changing vehicle geometry. The geometry parameters that most impact the SST mission drag polar characteristics and the minimum number of drag polar samples used to train the ROM will be established. A set of experiments will be used to answer opened questions on the methodology steps and show evidence of whether the proposed methodology is suited to solve the problem described. Once all experiments are performed, a detailed procedure on how to generate a parametric mission drag polar approximation for SSTs will be obtained. This result will be used in a final test to perform DSE and aircraft sizing for a novel SST. The last experiment in this thesis will evaluate the effectiveness of the proposed methodology to identify which areas of the SST design space present favorable fuel burn characteristics. For this reason, all the steps of the proposed methodology will quantify the impact of using approximation models on the estimate of vehicle fuel burn. And the aircraft performance of a vehicle obtained with the proposed methodology will be compared to a benchmark vehicle published in the literature. This dissertation contributes to the research of novel supersonic commercial aircraft and the broader work on aircraft DSE and sizing using high-fidelity aerodynamic models. The steps developed in this document can also be applied to other unconventional vehicles, which do not have abundant empirical data available to calibrate lower fidelity models.

Committee

• Prof. Dimitri Mavris – School of Aerospace Engineering (advisor)
• Prof. Daniel P. Schrage – School of Aerospace Engineering
• Prof. Lakshmi Sankar – School of Aerospace Engineering
• Dr. Sriram Rallabhandi – NASA Langley
• Dr. Jimmy Tai – Research Engineer, Aerospace Systems Design Laboratory