The AE Brown Bag Lecture Series is a Daniel Guggenheim School tradition: select AE undergrads and grad students share their research before an audience that includes their peers and academic advisors. If you would like to present a Brown Bag, speak with your advisor or contact faculty coordinator, Michelle Hall to schedule.
COVID-19 Programming Adjustment
Most Fridays during the Spring 2021 Semester, we will host 2 Brown Bag presentations, beginning at 12:30 p.m. EST. They will be viewable from your laptop via BlueJeans. If you are not able to attend, you may view a video recording of the presentation by clicking on the presentation title, below.
"Applying Machine Learning to the Angles-Only Initial Orbit Determination Problem"
As the number of satellites, and amount of debris, in Earth’s orbit continue to dramatically increase further developments in Space Situational Awareness are critical to responsible and safe use of near-Earth space. This presentation will focus on a data-driven approach to the initial orbit determination problem. In particular, development of a machine learning pipeline for real-time initial orbit determination from optical telescope data will be discussed. A convolutional neural network was trained to accept a series of inertially mapped orbits as features and outputted Keplerian orbital elements as labels. The created model was evaluated using both synthetic and real data collected at the Georgia Tech Space Object Research Telescope (GT-SORT).
"Development of a Soft Robotics Tensegrity Rover for Planetary Exploration"
Attained from the term tensional integrity, tensegrity describes a system comprised of isolated components under compression within a tensile network. These tensile forces help stabilize the structure, even when the compression elements experience buckling. Because they are generally lightweight and able to withstand high velocity impacts, tensegrity systems present an interesting possibility of adopting the concept for planetary landing and exploration applications. As a unique and exciting approach towards designing more efficient and cost-effective planetary rovers, the tensegrity rover could be implemented for future missions to Mars, asteroids, and more. This presentation features highlights of the engineering efforts involved in the design and development of a system which enables the tensegrity lattice to move robotically.
"A Study of Flow Dynamics on A Ballistic Bluff Body Combustor"
The objective of this research is to develop computer models of a reacting flow over a ballistic bluff body at low Reynold’s Numbers. Ballistic bluff bodies are present in a wide spectrum of applications, from industrial boilers to turbojet and ramjet afterburners, and are known for their flame stabilization characteristics. These simulations will be used to inform experimental results conducted in the Ben T. Zinn Combustion Lab and potentially provide more insight on flow fields. Several simulations were run in ANSYS Fluent, each with increasing complexity to achieve this goal. Firstly, Reynolds Averaged Navier Stokes (RANS) laminar and turbulent k-E viscous models were solved to determine the validity of the geometry and resultant solution. With these test-cases validated, unsteady RANS (URANS), k-w SST models were run for a sweep of Reynolds Numbers from 80 to 250. The results from this sweep showed that the Strouhal number, which is associated with the Von Kármán shedding off the bluff body, corresponded closely with that of the well-studied cylinder profile. As the Reynolds number increases, this vortex street grows in magnitude and frequency which can have serious implications for the flame stability. Future study would involve a Large Eddy Simulation (LES) in reacting flow. This additional investigation will be planned to gain insight on the breakdown of local vortices near the flame and irregular oscillations in the flow near lean blowoff conditions. By understanding when and how blowoff occurs with this geometry, the operating ranges of many combustion systems could be expanded and improved.
"Human-Integrated Tool for Proactive and Reactive Security in Cyber-Physical Systems"
As Unmanned Aerial Vehicles (UAVs) and Autonomous Systems (AS) become more prominent, it becomes increasingly important for a security network as well as effective collaboration between human operators and AS. To combat this issue, this presentation provides a comprehensive defense software for cyber-physical systems, comprising both proactive and reactive mechanisms to handle cyber-attacks. Specifically, the demonstrated tool allows a human operator to remain aware of the system’s health and operation, while an autonomous subsystem applies a switching rule based on the principles of Moving Target Defense (MTD). Furthermore, the man-machine interaction implements a trust metric, that allows either the autonomous mechanism or the human agent to have more control over the system based on attack detection and mitigation history. Finally, to demonstrate the tool’s security and collaboration capability, it is simulated using an ADMIRE aircraft.
As many companies begin to test prototype eVTOL aircraft, finding the optimal trimmed transition corridor between rotor-borne and wing-borne flight is critical for concepts with tilting rotors and wings. This presentation details the work done as part of a conceptual vehicle design of a eVTOL aircraft at Georgia Institute of Technology. A three degree of freedom model with linear aerodynamic stability derivatives and a surrogate propulsion model was developed for the given aircraft. Interactional effects between rotors and lifting surfaces were neglected during the first development iteration of the tool. The optimizer SNOPT (Sparse Nonlinear OPTimizer) was used to drive the minimization of a specified objective function. Results provide essential insight into the optimal control and constraints of the transition for eVTOL aircraft.
With the total number of satellites orbiting earth scheduled to increase by a factor of five over the next few years, the need for a cheaper and more environmentally conscious way of getting to space is more pertinent now than ever before. Given that the typical payload fraction is between 1 and 10%, any reduction in the amount of fuel needed would be significant both financially and environmentally. SpaceShot provides the conceptual development and simulated analysis of a hybrid launch system, from a primary maglev track to a secondary on-board propulsion system, for smaller unmanned missions, mainly the deployment of small satellites.
Nicolas San Miguel
Safe planning and control of multi-drone missions is an important problem for Urban Air Mobility, among other systems, where missions require an Unmanned Aerial Vehicle (UAV) to execute various tasks while satisfying spatial, temporal, and reactive requirements. This presentation will discuss the use of Signal Temporal Logic (STL) to describe the temporal requirements of a multi-UAV mission concisely and unambiguously and will discuss methods for the planning and control of a dynamically feasible vehicle trajectory that satisfies the specified requirements. The controller tracks a UAV along its trajectory with a bounded tracking error which is within a predetermined robustness margin. In addition, this talk discusses the characterization of the tracking error using sensitivity and robustness metrics.
In recent years, the focus on the development of hypersonic vehicles has increased significantly. Though the hypersonic flight regime presents many unique challenges, it also presents a broad range of benefits ranging from reconnaissance to commercial flight. Some of the challenges concerning research in this area include a limited understanding of the highly interdependent physics as well as classified tools pertaining to hypersonic vehicles. In order to address these issues, this Grand Challenge project seeks to create a multidisciplinary design analysis and optimization (MDAO) environment for a hypersonic vehicle using low fidelity and unclassified data and tools.
Model-Based Systems Engineering (MBSE) allows for effective representation of complex systems and has become increasingly utilized in the engineering design process. Using SysML (System Modeling Language) in the MagicDraw Tool, MBSE techniques were applied to a mission to characterize the lunar exosphere. This presentation will cover the application specifically from a requirements engineering approach, and emphasis was placed on the modeling of key and driving requirements, as well as the traceability of specific mission constraints originating from the request for proposal. The integration of external environments such as MATLAB and SolidWorks into MagicDraw for verification purposes is also briefly discussed.
GT-1 is a mission designed to train students in creating and operating a 1U CubeSat bus, in a fast development timeline. The avionics system is integral to mission success, providing the flight computer, attitude determination sensors, radio communications, solar panels, power supply, system monitoring sensors, and more. This presentation will give an overview of the GT-1 mission, avionics design decisions that worked well, and ones that didn’t, along with other important takeaways that may be useful on other small satellite missions.
"Effect of Heater Location on Horizontal Rijke Tube Hysteresis"
Identifying and mitigating combustion instabilities is one of the biggest areas of research in combustion and propulsion. In doing so, experimental rigs are observed in controlled burning, with temperature, pressure, and other variables tracked and recorded. Often, this raw data undergoes analysis attempts that require plenty of time and programming to do. A better method is to first observe the pressure and temperature data to find hysteresis zones, to narrow down the search for the instabilities. In order to exemplify this metric, a simpler analysis on a horizontal Rijke tube is conducted with preset conditions to find where hysteresis zones may lie.
"Development of Precision Deployment for the TARGIT Mission’s Tether System"
The Tethering And Ranging mission of the Georgia Institute of Technology (TARGIT mission) is one of the latest CubeSats being developed and built by the Space Systems Design Laboratory at Georgia Tech. The tether subsystem of TARGIT is responsible for deploying tether attached to a target inflatable that will be used for validation and testing of a miniaturized LiDAR instrument. To fulfill all requirements for the subsystem the relationship between deployment speed and hardware specifications was characterized, and a model for proper deployment developed. Following thorough testing, this model was incorporated into the TARGIT mission software for precision deployment of the inflatable target.
"ARL Autonomous Drone Docking and Charging System"
The United States Army Research Laboratory (ARL) commissioned a drone-charger interface that can perform fully autonomously. The project is headed by two labs at Georgia Tech. Dr. Jonathan Rogers’ lab has designed an autonomous drone capable of landing on a moving target, while Dr. Anirban Mazumdar’s lab designed and prototyped the charging dock. The landing pad dimensions were optimized in MATLAB using Monte Carlo simulations. The system requirements call for an autonomous charging dock that can charge through the feet of a drone regardless of landing orientation. The system had to be rated for 25 amps to charge 5000 mAh capacity LiPo batteries at 5C. With the optimal design point finalized, the dock frame was equipped with circuitry capable of withstanding the required loads. An Arduino Mega serves as the active controller onboard the drone and dock. Each Arduino is equipped with an XBee radio communication device to synchronize the stages in the charging cycle. In this presentation, I will discuss the charging process, design considerations, fail-safes built into the system, and applications of this system in recon drone missions.
"Reduction of Pilot Workload in Rotorcraft Shipboard Landings – A Quantitative versus Qualitative Analysis"
Currently, rotorcraft shipboard landings are associated with high levels of pilot workload. Various visual flight lead cues for real-time guidance were developed and tested in a pilot-in-the-loop flight simulation. For an approach-turn-land maneuver and an approach-land maneuver, the associated pilot workload is analyzed. A quantitative metric based on time-varying power frequency, and qualitative metrics, NASA Task Load Index (TLX) scores and the Deck Interface Pilot Effort Scale (DIPES), were obtained during the simulation. This presentation will outline the current understanding of the most accurate quantitative versus qualitative representation of this pilot workload data.
"Electrical Topology for eVTOL Concept"
The primary focus of this presentation is the development of the electrical topology for a multi-rotor eVTOL aircraft. Motor and battery type and number were already chosen based on powertrain needs. Given the powertrain, the electrical topology is designed to enable safety and low weight. The first important consideration for safety is isolation of motors from one another such that if any component of the powertrain (battery, motor, or inverter) fails, only the opposite motor is potentially affected, preventing failure propagation. Another important safety consideration is ensuring that, in the event of a failure in a circuit, the aircraft continues to maintain full flight control. This is achieved through servos capable of accepting redundant power drawing from two isolated circuits. In addition, servos with redundant motors allow for maintaining control if a servo motor fails. In order to reduce the weight of the circuitry, the power to the ailerons and flaps is carried by the same wires as the power to the motors until near the actuators, where the wires are branched and a DC/DC converter drops the voltage from the motor voltage to the actuator voltage. For actuators that are distant from any motors, such as actuators for tail control surfaces, the DC/DC converter drops the voltage near the batteries, minimizing the wire gauge required to carry power to the actuator.
"SAE Aero Design: Developing a Highly Competitive System of Systems"
For the past 6 years, Georgia Tech has taken first place at the annual SAE Aero Design Advanced Class competition. The competition objectives are inspired by a Mars colonization mission and tasks teams with prototyping and flying an aircraft capable of accurately dropping water bottles, Nerf Aero Howlers, and autonomous gliders into a 50’ target. The aircraft is subjected to a 11’ maximum wingspan and a 750W power limit. Each glider must carry as many “colonists” (represented by table tennis balls) as possible, glide to and land autonomously within the target zone, and weigh less than 9 oz. Score is determined by payload quantity as well as an optimum ratio between the three types of payload delivered.
The iterative design process includes but is not limited to aircraft sizing, aerodynamics analysis, structural analysis, propulsion system selection, and systems testing. The analysis, modeling, manufacturing, and testing methods refined over multiple years have contributed to overall success. Constant flight testing and a history of learning from failure leads not only to a reliable aircraft system but also to a body of adept students that can compete with confidence. Lastly, a constant cycle of peer student-to-student mentoring ensures continued improvement of the team’s ability to design, build, and fly aircraft rapidly.
"2D Automatic Table Generation"
Having multiple parameters and being able to get a solution for an unknown in real time is crucial in fields where this information can be used to make split second decisions. The automatic generation of a 2D table allows for adaptive updates to take place throughout the entire procedure/mission. These computations happen in a continuous manner where the table is updated rather than recomputed. This presentation highlights the iterative methods used to generate the table.
"Development of a Cold-Gas Thruster Component Testbed"
The volumetric constraints of small-satellites, specifically CubeSats, limit the availability of propulsion systems for small-satellite missions. The Lightsey Research Group in the GT Space Systems Design Lab develops custom cold gas thrusters for small-satellite missions, making use of additive manufacturing to integrate the main tank, plenum tank, nozzles, and tubing into one monolithic structure that optimally uses the available volume. However, the compact design significantly limits the ability to test individual components, particularly the solenoid valves and their electronic controller. A scaled-up model of the cold-gas system would allow for simple testing of components along with measurement of pressure and temperature at intermediate points in the flow path. This presentation will discuss the design, development, and upcoming testing of an expanded cold-gas thruster component testbed.
Current Mars entry missions use a rigid thermal protection system, which is limited in diameter by the payload fairing of the launch vehicle. To account for the larger payload masses and decelerator surface areas required for future surface missions, NASA has conducted studies into Hypersonic Inflatable Aerodynamic Decelerators (HIADs), which allow large decelerators to be packaged into small volumes. However, HIADs require a flexible thermal protection system (FTPS), which is currently at a low TRL and requires high-fidelity modeling prior to full-scale system testing. This presentation includes the Martian atmospheric entry trajectory modeling for HIAD use cases, stagnation point heat flux calculation for a given trajectory, and the preliminary modeling of FTPS layups to determine the necessary material thicknesses for a given entry trajectory.
Interest in small satellites, specifically CubeSats, has surged in recent years as they offer opportunities to conduct scientific investigations and technology demonstrations in space in such a way that is cost-effective, timely and relatively easy to accomplish. GT-1 is a student-led 1-U CubeSat mission being developed in the Space Systems Design Laboratory with a goal of developing a fully functioning and robust spacecraft bus to be implemented on future GT series CubeSat missions designed to operate in Low Earth Orbit. A critical feature of the spacecraft bus is the UHF antenna and solar panel deployment mechanisms. The small form factor of the 1-U CubeSat (10cm×10cm×10cm) imposes several constraints on such deployment mechanisms and therefore require unique solutions to meet the mission requirements. This presentation aims to explain the development and decision-making process involved in designing, manufacturing, testing, and integrating the deployment mechanisms including the problems encountered and the lessons learned throughout the past year
Isaac Del Valle
Tensegrity, a term derived from tensional integrity, refers to a specific class of structures composed of rigid bars and cables. When these bars and cables are connected to each other, they form a lattice-type structure which is capable of undergoing severe deformation due to the buckling of its bars. Utilizing the ability to deform or compress, a tensegrity lattice demonstrates advantageous characteristics as a rover for micro or low gravity environments, including being robust to failure, impact tolerant, and capable of energy efficient modes of locomotion. This presentation highlights the manufacturing processes used to build tensegrity structures, as well as the methods that have been employed to provide locomotion to a tensegrity lattice for future planetary exploration.
One of the recent areas of focus in aerospace has been the use of Model-Based Systems Engineering approaches to aid in the system engineering process. This talk will give an overview of the application, benefits, and limitations of MBSE to modeling aspects of an unmanned mars mission system using SysML in MagicDraw. This includes modeling aspects such as requirements, system structure and hierarchy, and parametric analysis. Additionally, this talk will also briefly discuss the aerodynamic and trajectory analysis code developed for planetary entry as well as the current limitations of MATLAB integration within MagicDraw.
"Conceptual Design and Data Visualizations for Georgia Tech Interuniversity Mission Operations Control Center (GTMOCC) for Cubesat Missions"
As Georgia Tech expands its space systems capabilities with current and planned future Cubesat Missions, it requires a strong foundation for mission operations to be able to send commands and downlink telemetry from these systems. GT-1 and TARGIT are scheduled to be the first customers of the MOCC, however this system will be designed for versatility to accommodate multiple future missions including GT-2+, and formation flying mission like SWARM-EX and VISORS. This presentation aims to provide a proposed layout of how data received from the satellites will be processed to generate mission critical visualizations, and describe the different MOCC operating modes. Integrating AGI Systems Tool Kit along with MATLAB will provide the functionality to propagate estimated orbits for the satellites and calculate access data regarding when estimated next passes occur in order to aid the operations team in changing operating modes.
Utilizing hydrocarbon fuels for rocket engines is beneficial in terms of cost, reliability, and environmental friendliness. To enable more informed and cost-effective designs of rocket engines, it is key to understand the combustion processes involved in the combustion chamber; one approach to do so is through numerical simulations. Numerical simulations facilitate deeper exploration of the physics involved in comparison to experiments. In this study, the numerical simulation of a 7-injector, GCH4/GOX, experimental rocket engine (developed at the Technical University of Munich) is performed with an in-house LES CFD solver. The numerical methodology adopted is based on the well-established, second-order accurate (in both space and time) finite-volume solver for the unsteady Favre-filtered multi-species compressible Navier-Stokes equations. A hybrid scheme, which switches between a second-order-accurate central scheme and a third-order-accurate MUSCL (Monotone Upstream-centered Schemes for Conservation Laws) scheme is employed. The reacting flow and flame features/characteristics are studied from the simulation results.
"Finite Element Model Characterization of a Regeneratively Cooled Nozzle"
Regeneratively cooled nozzles often circulate cryogenic compressed fuel through a nozzle composed of hundreds of tubes. This process cools and helps maintain the strength of the metal components during rocket engine operation. Engine designers use a simplified engine system finite element model to determine loads for component design. A two-dimensional shell mesh can be used for the nozzle in this load analysis model. Beam theory can be leveraged to compute approximate orthotropic material properties such that the shell elements will exhibit the same behavior as the nozzle tubes. A “truth model” can be made using cyclic symmetry. The mode shapes and frequencies of the shell model can be tuned to this truth model using Attune by adjusting the shell orthotropic material properties.
The AIAA undergraduate engine design competition committee has requested proposals for a new engine design for their Mach 2.1 business jet entering into service in the year 2030. The YJ-2030 is Georgia Tech’s response to the AIAA Request For Proposal (RFP). The engine is an afterburning mixed flow turbofan that allows the aircraft to cross the Atlantic in under 5 hours, while still offering extended range at subsonic cruise. This presentation will cover the requirement definition, cycle design, component design and the material and weight analysis of the YJ-2030 engine.
Adrian Vincente La Lande
“3-Dimensional B-Field Profile of a Hall-Effect Thruster”
Like many forms of electric propulsion, Hall Effect thrusters rely on the interaction of electro-static forces. One of the most important physical features is the magnetic field that is distributed both axially and radially inside the discharge channel and allows the thruster to create its characteristic Hall current. This is a critical function of the thruster, as the Hall current sustained by this magnetic field is necessary for creating the plasma ultimately used to produce thrust. Understanding the magnetic field topology of Hall effect thrusters is critical for improving their performance. Despite this, we do not have many ways of spatially characterizing said magnetic fields in our lab. To enhance the diagnostic capabilities of the High-Power Electric Propulsion Lab (HPEPL), a B-Field Mapper (BFM) tool was developed to allow users the ability to accurately and 3-dimensionally map the magnetic field flux of a thruster. This presentation covers the development, application, and performance of the BFM, as well as the magnetic-field topography it measured from a T-220 Hall-effect thruster.
"Developing Preliminary Sizing Optimization and Safety Analysis Tool for Wing Substructure"
Early design decisions made during the conceptual and preliminary design phases of a development program are largely influential to the final weight, design, and cost of the aircraft. This presentation discusses the development and integration of a tool designed to optimize the weights of primary substructure groups within a user-defined wingbox. The optimization process is constrained to ensure all relevant structural margins of safety are satisfied under various shear and bending loads experienced during flight. The resulting optimal weights are summed to return to the user an evaluation of the ideal weight of the analyzed wingbox. This allows for fast and accurate sizing considerations early in the design process.
"Optimal Control for In-Host Virus Models"
Mathematical virus models bear a close resemblance to dynamical systems common in aerospace engineering. This similarity enables the use of optimal control methods to determine improved treatment regimes. The talk will give an overview of ordinary differential equation and continuous time markov chain models (a type of jump process) for viral in-host infections. Control emerges by interrupting infection of new cells with medication. The cost function encodes minimizing amount of medication, and thus side effects, while still eliminating the viral population. Differential dynamic programming will be discussed for the ordinary case and model predictive path integral for the stochastic case.
"Modelling and Analysis of Satellite Megaconstellations"
Much of the anxiety surrounding the deployment of constellations of hundreds or thousands of satellites stems from a position of concern as to the potential interference impact such constellations will have on communications infrastructure. By making use of the capabilities found in STK, especially those related to communications, it was determined that certain types of ground receivers, under the right conditions, are more exposed to interference from the downlink satellites of these megaconstellations. The project also developed a workflow procedure to efficiently and accurately model megaconstellations by using generating two line element files for the individual assets of the constellation.
Since the proposed constellations have not yet been deployed as of this project, certain assumptions had to be made to generate a realistic notional constellation, including altitude, RAAN, and inclination of the orbital shells. A custom-made python script was used to iterate through values in the TLE to generate evenly distributed networks of satellites in each shell based on input altitude, planes per shell, and satellites per plane. A notional constellation of 12,000 satellites distributed among three different shells was created. Deck Access was then used to determine which of these fictional spacecraft would fall within the field-of-view of the ground-site within the analysis period.
"Initial Prototyping and Testing of a Modular Delivery Drone"
Delivery companies are looking to drones as a method for delivering packages quickly and reliably to consumer doorsteps. However, the development of a universal delivery vehicle leads to a tradeoff between payload capacity and flight efficiency. One solution is to build a modular drone such that smaller loads would be carried by a single drone while larger loads would be carried by, for example, four drones together. One such prototype was developed as a series of quadrotors organized in an X-pattern. After a single quadrotor was built and flown, a lightweight aluminum frame was constructed to interlock four drones and a single flight controller.
Flight testing was accomplished by tethering all drones to the ground, enabling a controllable flight radius ranging from three to thirty feet. The drone was able to be piloted directly with full directional control, but with noticeable flight instability. In an attempt to solve this, GPS systems were added to the drone to enable autopilot functionality. However, the drone was unable to maintain a constant position and altitude. Most recently, a single drone was flown with a suspended load using both direct and remote control with great success. As such, it was determined that flight controller redesign in the form of PID gain manipulation would allow for stable flight of the modular drone system