Organizational Unit:
Daniel Guggenheim School of Aerospace Engineering

Permanent Link
https://hdl.handle.net/1853/70742
Research Organization Registry ID
Description
Previous Names
Parent Organization
Parent Organization
Organizational Unit
Includes Organization(s)
ArchiveSpace Name Record
https://finding-aids.library.gatech.edu/agents/corporate_entities/106

Filters
1
2021 1
1
1
1
Reset filters

Settings
Author: Caglar, Havva Irem × End date: 2022 × Start date: 2020 × search.filters.applied.f.isOrgUnitOfPublicationTitle: College of Engineering × search.filters.applied.f.resourcesubtype: Thesis ×

Publication Search Results

Now showing 1 - 1 of 1
Thumbnail Image
Item

A METHODOLOGY FOR CONDUCTING DESIGN TRADES FOR A SMALL SATELLITE LAUNCH VEHICLE WITH HYBRID ROCKET PROPULSION

2021-07-28 , Caglar, Havva Irem

The commercial space industry has recently seen a paradigm shift related to the launch of a small satellite into Low Earth Orbit. In the past, a small satellite was launched as a secondary payload with a medium or heavy launch vehicle where the primary payload placed a constraint on the orbit and schedule. Today, a dedicated launch of a small launch vehicle is the main operational concept to launch a small payload. Many Smallsat Launch Vehicles (SLV) have been under development by the commercial space industry to improve these launch services in recent years. Despite these efforts, the specific prices per launch are still high, and reducing these prices further remains a challenge. One promising technology candidate to reduce costs for SLV is hybrid rocket propulsion which has matured recently with some cost and safety advantages. Although hybrid rocket propulsion faces a number of challenges, including a low regression rate and combustion instabilities, academia and commercial companies have invested significant resources in developing this technology. With this motivation, this thesis has focused on the conceptual design of SLV with hybrid rocket propulsion. Moreover, a cost reduction strategy currently used by the commercial space industry was observed to be the development of a unique engine and using multiple of them in a launch vehicle. Following this trend, the vehicle concept investigated in this thesis was an expendable ground-launched vehicle with some architectural variables such as the number of stages and the number of hybrid motors in each stage. The design trade-off studies of such a small multistage launch vehicle with multiple hybrid motors in each stage require very long times especially when traditional point design approaches are used. As the number of design variables increase, the design space exploration becomes even more challenging. To provide a solution to this problem, a methodology for rapid conceptual design of such a vehicle was presented in this thesis. A physics-based conceptual design approach was followed in this study since SLV are relatively new concepts without much historical performance data. To conduct a multidisciplinary analysis, a physics-based, integrated modeling and simulation environment was constructed with four core disciplines: trajectory analysis, aerodynamics, propulsion, and weight. Aerodynamics and propulsion analysis were conducted using a first-principles approach, which was based on fundamental theories. A 3 Degree of Freedom (DOF) industrial, transparent, physics-based trajectory analysis software was used in this study based on availability. However, any other trajectory analysis software that a system designer is familiar with can be used in its place. In other words, the methodology developed in this thesis would remain unchanged if another trajectory analysis software were used. The weight discipline was represented at a high level by using Propellant Mass Fraction (PMF) design variable. A multidisciplinary modeling and simulation environment for launch vehicles may be computationally expensive depending on the fidelity levels of each discipline. Moreover, trajectory optimization is included in a launch vehicle design process conventionally which may be also computationally expensive depending on the optimization method. This expense poses a difficulty in performing a trade-off study for hundreds of vehicle design alternatives within the constraints of the schedule in the conceptual design phase. Because of this, trajectory optimization was removed from the design process to speed up the process by selecting a constant controller design. The methodology developed in this thesis consisted of two sequential steps. In the first step, a surrogate modeling approach was followed to replace the Modeling and Simulation (M&S) environment. A DOE method and a surrogate modeling method suitable to this problem were searched in this part. To cover the design space, a hybrid DOE consisting of a Fast Flexible Filling DOE and a three-level Full Factorial DOE was chosen. Artificial Neural Networks method was selected to fit approximation models because of the type of design variables (both continuous and discrete variables) and nonlinearity of the problem. The first experiment was conducted to test this hypothesis. As a result, it was demonstrated that this approach can provide accurate surrogate models for any desired response. In the second step, the specific mechanical energy-based design trade-off method was developed using some statistical methods. This method estimates the lower bound of the vehicles’ actual specific mechanical energy where the vehicles can be rapidly designed by using surrogate models. This lower bound was predicted with the help of the prediction interval of the specific mechanical energy’s model fit error. To fit the surrogate models, the necessary data were gathered by running the DOE in the integrated M&S environment while imposing some terminal conditions on the altitude of the vehicles analyzed in this environment. Specifically, the surrogate models of specific mechanical energy and flight path angle were used to design the vehicles rapidly. The second experiment was conducted to test this hypothesis. As a result, the actual specific mechanical energies computed via trajectory optimization were found to be consistent with the predictions. Overall, it was demonstrated that the proposed method enables a system designer to rapidly design some feasible vehicles, which can then proceed to the next design phase for further comparison, analysis, and design.

View