Fridays, 12:30 to 1:30 in Gugg 442, the AE School gives its best students a chance to shine
The Brown Bag Lecture Series is a tradition that gives select AE undergrads and grad students a unique opportunity to share their research before an audience that includes both their peers and their academic mentors and advisors. For some, it's the first time they've ever presented their work to anyone; for others, it's a practice run for a doctoral proposal. What all Brown Bags have in common is the rigor of their research and finesse of their presentations. Below, we will feature the abstracts of the Brown Bag presentations from the current semester. If you would like to be considered for a future Brown Bag presentation, speak with your advisor or contact faculty coordinator, Michelle Hall.
Pizza provided. Curiosity required.
"Parametric Aerothermodynamic Hypersonic Design Optimization"
In the coming decades, hypersonic flight for space travel, commercial transport, and defense will require the development of optimized vehicles for highly specific roles. The design of hypersonic vehicles is a highly coupled system, having to evaluate feasible vehicles which are both able to meet mission requirements, while being thermally protected from hypersonic heating. To further develop this field, quick and accurate hypersonic optimization tools must be developed that implement mission constrains, geometric considerations, and survivability considerations. This research considers these requirements and developed a parametric model of hypersonic vehicles, which is used to optimize reentry vehicle trajectories, coupled with sizing the thermal protection system of these vehicles. Optimized reentry vehicle designs can then be considered for upgrades to existing launch platforms, or as payloads for next generation launch systems.
"Real Time Data Table Generation"
An efficient method for real time data table generation is described that is based on a specialized recursive least squares algorithm. The method can be used as part of a control system that adaptively computes a plant model. The algorithm was tested on several parafoil and payload aircraft to form a table of turn rate versus asymmetric brake input. Studies are reported that compare the resulting table structure as a function of the number of data points, size of the table, and forgetting factor.
Shubham Basavaraj Karpe
"Soot Modelling using Higher Order MOMIC Approach"
Energy conversion and propulsion devices operating in a non-premixed burning mode often lead to an incomplete combustion of the fuel due to lack of complete mixing of the fuel with the oxidizer. Additionally, such devices operate at higher temperatures in the range of 1500-2000 K. Both these factors, being favorable for soot evolution leads to the formation of soot precursors which subsequently grow, nucleate and agglomerate to eventually develop into soot particles of larger size. Soot, being one of the major pollutant impacting environ-mental welfare, has motivated significant research in the field for alternate fuels. Therefore, cost-effective numerical simulations accompanied with experimental validations are key to have accurate estimations of soot evolution from combustion processes. However, lack of understanding of chemical kinetics governing the formation of soot precursors, uncertainty in the coupling between fluid dynamics, heat transfer, chemical kinetics and turbulence leads to uncertainty in the quantitative prediction of soot formation.
The current presentation outlines use of method of moments with interpolative closure (MOMIC) coupled with LES to examine sooting behavior in a canonical reacting problems. In MOMIC, the transport equations for the integer moments of the soot size distribution function are solved while those with fractional indices are interpolated, and local statistical quantities like soot number density and volume fraction are estimated. This approach is computationally affordable compared to other prevalent methods such as Lagrangian particle tracking method and the sectional method. Typically, LES filtered equations contain many terms that require closure modeling. In this approach, a soot subgrid model based on the linear-eddy mixing (LEM) model will be used for the unresolved small-scale interactions between soot, turbulence and chemistry. The end result of this study is to develop computational framework using MOMIC approach coupled with LEM subgrid closure model and use it to study soot formation and evolution in canonical non-premixed turbulent flames.
"Design and Manufacture of a Short Takeoff and Landing Cargo Aircraft for Urban Package Delivery"
This talk will describe the design and manufacturing of a sub-scale research aircraft to investigate the effectiveness of externally blown flaps for achieving extreme short takeoff and landing (STOL) performance for urban cargo delivery applications. The design leverages distributed electric propulsion with six propellers oriented along the wing leading edge for wing blowing. The aircraft is a three-surface configuration incorporating a canard, wing, and horizontal tail to achieve low induced drag at all flight conditions and ample pitch control authority in low speed operation with wing blowing.
The talk will describe the preliminary and detailed design of the aircraft, the prototyping and manufacturing process, and work to integrate an air data system for a planned flight test campaign. Ongoing work on design optimization of a full-scale blown wing cargo aircraft based on the lessons learned in the sub-scale effort will also be discussed.
"Design, Analysis, and Test of a Modular GN&C Architecture for High-Powered Launch Vehicles"
This presentation outlines the late-stage development, integration, and test of a modular GN&C architecture for high-powered launch vehicles. Given the results of an in-house 6-DOF simulator, preliminary testing of the canard-based active control system began with the characterization of the inertial sensors. The frequency stability of the IMU was determined via Allan Deviation plots which map the bias instability and random walk characteristics.
The state estimation algorithm employs the IMU measurements to formulate a navigation solution, and the PID controller uses this solution to compute the necessary deflections of the canards to maintain the target orientation of the vehicle. Results from the first launch highlighted the need of a higher fidelity navigation system, and thus the designs of Kalman Filters for position and attitude estimation were built and simulated via an in-house, Simulink 6-DOF model. The end result of this study is a robust GN&C and flight software architecture for any launch system that employs canard actuation for active control.
"Relative Positioning and Tracking of Tethered Small Spacecraft Using Optical Sensors"
This study analyzes the attitude dynamics and tracking control system for an object attached via tether to a primary spacecraft using optical sensors. Requirements of the tracking include body pointing of the main spacecraft to the tethered target, in addition to optimized sun pointing for maximum power generation from the primary spacecraft’s solar panels. This study shows that a modified PD controller with 3 reaction wheels can track the tethered object given only relative position information, i.e. the relative target position and the sun pointing vector in the body frame, generated from an optical camera and sun sensor.
"Understanding High-Throughput Satellites: Market Disruptions, Technology, and Value Analysis"
High-Throughput Satellites (HTS) are a distinctive class of communication satellites that provide significantly more throughput per allocated bandwidth than traditional wide-beam communication satellites. They are the proverbial wave of creative disruption in the space industry and are poised to disrupt the communication market in significant ways. The first objective of this work is to introduce these satellites, discuss their markets, and present their underlying technology base. The second objective is to develop a decision-analytic framework for assessing the value of HTS and to provide meaningful results of the value of such systems under realistic design, operational, and market conditions. We develop the cost and revenue models of HTS.
To build the revenue model, we develop a hybrid data-driven and scenario-based load factor model that combines historical data based on financial records from current HTS operators with extrapolations based on best-, nominal-, and worst-case scenarios. We then integrate the cost and revenue models within a stochastic simulation environment and perform Monte-Carlo analysis of the net present value (NPV) of HTS. One important result is that a medium-size HTS significantly outperforms a roughly equivalent traditional wide-beam satellite, even under the worst-case loading scenario. Another important result here identified and quantified is the trade-off between the average revenue per user (ARPU) and average loading of the satellite and how it is mediated by the downlink speed provided to consumers. This result can be used in different ways, for example, by helping define the boundaries of what is competitively achievable in terms of ARPU and downlink speed offerings. The implications of these results for satellite operators and investor are that they delineate the pathways to financial success or failure and the boundary conditions beyond which an HTS will be value-negative.
"The Benefits of Continued Advances in the Propulsive Capability of the Electric GEO Communications Satellite"
The implementation of electric propulsion systems on Earth orbiting satellites has dramatically improved the fuel efficiency, and in turn the capability and resilience, of such satellites when compared to their chemical propulsion counterparts. Electric propulsion systems flown to date have achieved power levels up to 5 kW per string for GEO communications satellites enabling electric orbit raising, orbit positioning, and station keeping functions. This presentation will investigate the GEO mission capabilities and user benefits of further advances in the operating power and performance of electric systems. It is shown through mission analysis that a modest improvement in performance over the current flight systems results in a substantial improvement in mission and launch vehicle flexibility. The benefits of such an improvement are quantified with a comparison between reference missions that utilize existing commercial electric systems, and those that utilize an electric system with an incremental advancement in performance. The substantial mission improvements discussed in this presentation justify the need for continued advances in electric propulsive capability.
"Design, Build, Fly: Developing a System of Systems Approach for the SAE Advanced Class Competition"
The SAE Aero Design Competition rules for the Advanced Class aircraft simulate a colonization mission. Points are awarded for dropping two types of ballistic payload and deploying autonomous gliders into a drop zone. In order to increase score, an optimal ratio of payloads must be used while maximizing payload fraction. The aircraft was designed utilizing a system of systems approach to tackle the many challenges. This presentation will cover the design, manufacturing, testing, and results of the Georgia Tech Advanced Class aircraft at the SAE Aero Design Competition.
"Autonomous Landing of an Unmanned Aerial Vehicle using a Deep Neural Network Vision Learning Algorithm"
As we continue to push to an increasingly autonomous world, the successful landing of autonomous unmanned aerial systems (UAS) is crucial for many practical applications. Two different vision-based algorithms, the density-based spatial clustering of applications with noise (DBSCAN) method and a deep neural network (DNN) method, can successfully identify and track a designated landing target. With an average of seven times faster processing speed and a five times greater useful altitude range, it is clear that with a robust training data set the DNN method outperforms the DBSCAN method in nearly all possible situations. The research work also explores different ways to generate useful training data for the DNN in the simulation environment Gazebo. Creating a multi-axis PID controlled gimbal to gather onboard images from many angles while simultaneously varying the simulation environment’s parameters, allows for the collection of robust training data to prepare the DNN for any potential situation to include high altitude, cluttered, and GPS-absent environments.
"Optimizing a Coupled CFD Solution for the Orion Crew Capsule Heat Shield"
The goal of this project is to study the convective and radiative heat flux predictions using computational fluid dynamics (CFD) on a smooth body representation of the NASA Orion Multi-Purpose Crew Vehicle (MPCV). The Orion MPCV is designed for long duration missions in space and will play a crucial role in transporting astronauts in the upcoming Artemis missions. With Orion’s aim for cost-effective reusability comes a need to study its Thermal Protection System (TPS), which allows for safe entry into the atmospheres of the Earth, Moon, and Mars.
In order to study the effects of the TPS on test flights and better understand the aerothermal environment, CFD is a necessary tool used to create simulations of entry vehicles and calculate the heat load and heat flux. However, solutions are often performed using an uncoupled approach between the flow and the radiation solver. This leads to an overestimation of the overall heating effects because they do not consider radiative cooling. Furthermore, the effects of radiation on the aft-body is an area of research with significant uncertainty.
Through the use of codes developed at NASA Ames, DPLR and NEQAIR, a coupled solution can be attained for the Orion capsule which will more accurately solve for the effects of radiation in both the fore and aft-body regions. This yields more accurate design margins for the TPS, which produces a better estimate of the thickness required. By developing an accurate coupled solution of the TPS on the Orion MPCV, planners will have greater flexibility in carrying out future missions to the Moon and beyond."
Experimental Evaluation and Re-design of a Small Gas Turbine Engine for Improved Thrust to Weight Ratio
As small-scale UAVs become more prevalent, there is a great demand for improvement in the propulsion system. The purpose of this project was to improve the thrust to weight ratio of a JetCat P100 RX engine, a system often used in model aircraft.
Initial ideas included adding an afterburner, redesigning the compressor, modifying the nozzle, and manufacturing lighter components. With the construction of an accompanying NPSS model, we opted to add deswirl vanes in the nozzle and reduce the weight of the compressor shroud. This resulted in a substantial thrust increase and minor weight reduction. Both components were additively manufactured.
"Machine Learning and its Applications to Control"
Classical control schemes have been demonstrated to work well for a variety of systems including flying vehicles such as quadrotors. The traditional approach to control is effective when applied to systems whose dynamics are nearly linear and in low uncertainty environments. However, when dealing with more complex nonlinear dynamics, or when the systems are pushed near their stability limits, classical control methods perform poorly.
The application of machine learning algorithms to control has been proven to be effective for accomplishing difficult tasks in partially unknown environments while allowing the system to adapt to dynamically changing situations. In this presentation, we will cover some of the algorithms developed at the Autonomous and Decision Systems (ACDS) Laboratory and provide examples of systems where these algorithms have been successfully implemented.
"Vehicle Performance Evaluation for High Fidelity Agent-Based Battle Theater Modeling"
Becoming of interest to the military, Attritable Unmanned Aerial Vehicles are reusable military platforms that resiliently adapt to and fulfill mission requirements, but are cost effective enough to be considered expendable. In an air-to-air scenario, facing dynamic & highly-mobile threats, vehicle performance of the vast design space becomes critical to understanding the attritability and capabilities of these aircraft. This presentation will cover the methods used to evaluate the mission performance of viable AUAV designs to improve the fidelity of agent-based scenarios in a battle theater.
"Validation of a Radiofrequency Generator for a Heritage Ion Thruster"
When designing an ion thruster, there are two main aspects to address: efficiently ionizing the propellant and accelerating the ionized propellant to produce thrust. In this work, we seek to experimentally validate the performance of a recently developed radio frequency generator in ionizing xenon propellant in comparison to current designs. The performance will be analyzed in terms of efficiency parameters defined in a theoretical model of the average beam ion energy cost of an ion thruster. The radio frequency generator performance is characterized under high-vacuum conditions via plasma diagnostic measurements. Additionally, this work seeks to investigate the validity of certain assumptions made during the experimental procedure.
"Electrical Power System for RECONnaissance of Space Objects (RECONSO) Cube Satellite"
RECONnaissance of Space Objects (RECONSO) is a multidisciplinary cube satellite project worked on as part of the Space Systems Design Lab (SSDL) and the Air Force Research Lab (AFRL). The project combines students from several different engineering programs such as mechanical, aerospace, electrical, and computer engineering. The purpose of the satellite is to demonstrate a low cost and simple method to improve situational awareness of objects in low Earth orbit. The satellite will fill a current gap in space debris detection for objects between 1 to 10 centimeters. The mission is designed last over a 6 month period and it will be launched as a secondary payload. The satellite operates by using a low light camera to detect and track objects that are moving relative to the star field in the background. These tracks would then be sent back to the ground and compared with existing catalogs of space objects. It will primarily focus on observation of regions with high densities of space debris such as around the poles.
"Study of Oxy-Combustion at Elevated Pressures in a Premixed System Using Reactor Network Modeling"
Chemical reactor network (CRN) modeling allows investigation of t he interaction of combustion with simplified flow characteristics so that parametric design studies can be performed in a cost-effective manner. In this study, CRN modeling is used to evaluate the performance of an axially-staged premixed combustor under practically relevant high-pressure operating conditions. First , the computational approach is established by comparing the results against the available experimental and numerical data. Afterward , a detailed parametric investigation is conducted to characterize the role of auto ignition, 0 2 as an oxidizer instead of air , CO 2 as the crossflow, and react ants st aging ratio at different operating pressures to establish near-optimal operating conditions for an axially-staged premix ed combustion device. The parametric studies are perform ed by using a skeletal mechanism comprising of 19 species and 93 reaction s, which is shown to predict results matching the more detailed USC [? ] mechanism. The near-optimal design condition, which is characterized in terms of a reduced level of pollutant (CO) emission is found to correspond to oxy-combustion at a high pressure (100 - 200 atm) with a hot CO2 as cross-flow (50% or higher by mass) at the inlet of the first stage of the device.
"Structural Iterative Design for the RECONSO CubeSat Mission"
As satellites have become more cost-effective and are higher in demand, the role of CubeSats has become increasingly important. RECONSO is a student-led CubeSat mission that will autonomously detect and track space debris for 6 months in Low Earth Orbit. Throughout the duration of the mission timeline, the structures subsystem team has been working to iteratively update the satellite design in order to meet all the specifications for other subsystems. The structures team has supported this design process by performing a number of tasks: using SolidWorks to model the satellite, creating engineering drawings to effectively communicate design changes, contributing to the manufacturing and fabrication of various parts, and providing documentation to the Air Force Research Laboratory (AFRL). This presentation will cover the past and current state of the CubeSat, and the individual tasks that were completed to support the project.
"Flame Visualization of Hall Thruster Propellant Distributors"
The anode of a Hall Thruster is a two-purpose component that serves to both distribute propellant into a Hall thruster discharge chamber and complete the electrical plasma circuit. This propellant distributor must uniformly distribute neutral propellant around the circumference of the thruster discharge channel to provide an optimal environment for propellant ionization and acceleration. In this work, we apply a previously developed premixed-flame visualization method to verify the uniformity of newly manufactured anodes. Additionally, we investigate the presence of flow uniformity degradation of previously inspected anodes over the lifetime of Hall thruster operation. The visualization method used correlates the image intensity of a premixed flame at each azimuthal location to the relative flow density across the propellant distributor.
"Convergent Aeronautics Solutions (CAS) for Lithium-Oxygen Batteries in NASA Electric Aircraft (LiON)"
The paramount obstacle to enable NASA’s vision of Green Aviation is the extraordinary energy storage requirements for electric aircraft. Lithium-Air batteries have the largest theoretical energy storage capacity of any battery technology and if realized will transform the global transportation system. A significant problem for current Lithium-Air batteries is large scale decomposition of the battery electrolyte during operation leading to battery failure after a handful of charge/discharge cycles. Thus, development of rechargeable, large scale, ultra-high energy, and safe Lithium-Air batteries require highly stable electrolytes that are impervious to decomposition under operating conditions. The ultimate goal for this research is to discover the “design rules” for ultra-stable electrolytes for Lithium-Air batteries. These high energy batteries have the potential to meet the energy storage challenges of current and future NASA aeronautics and space missions in addition to many terrestrial transportation applications.
"Utilization of PLA in a Closed-Loop Space Based Agriculture System for Human Exploration"
The large mass required for human caloric intake can be reduced with on-site cultivation of agricultural products. This would reduce the launch mass and volume requirement to store the food supplies. Due to the restricted water availability on the prime destinations for human exploration, aeroponic technology has the capability of using 95% less water than traditional farming techniques and 40% less water than hydroponics. While limited in the variety of crops that can be grown, high lactic acid produce, such as soybeans, have high promise for protein requirements as well as conversion to PLA. The utilization of produce that produces PLA allow for a reduction of resupply mass for large structural components and replacement parts. This is accomplished by additive manufacturing on-site after conversion to PLA to expand the agriculture structure to support higher populations of astronauts. Therefore, such a system would only require an initial “seed” to then self-expand beyond its initial infrastructure. While this system would not completely supplement the human dietary needs, it is a decent start for further development of human exploration and provides a frame work for other agricultural technologies such as aquaponics, hydroponics, and soil-based farming to be developed during the mission time.
"Effect of Enhanced Flight Vision Systems on Pilot Performance"
Due to a need to compensate for missing visual cues in IFR flight, Enhanced Flight Vision Systems (EFVS) comprised of a variety of sensors are gaining popularity. In this presentation, we will demonstrate the effects of the portrayal of runway markings and sensor range limitations of those systems on pilot performance in terms of relevant approach and landing metrics as well as pilot workload. This study was conducted through human trials on a general aviation flight simulator in the Cognitive Engineering Center and the results have implications for the types of visual cues pilot’s need as well as the adverse effects of not providing those cues. Ultimately, this research serves as a first glance at how EFVS impacts pilots’ information processing that will require more in-depth analysis now that some focal points have been identified.
"Attitude Determination and Control System (ADCS) Development & Verification for RECONSO CubeSat Mission"
The RECONnaissance of Space Objects (RECONSO) is a low-cost CubeSat mission being developed to help catalog unknown objects in Low Earth Orbit (LEO). With the lowered cost and increase in missions to LEO, space debris becomes a hazard that can jeopardize satellite missions and restrict access to space. To support the RECONSO mission, a 3-axis inertial Attitude Determination and Control System (ADCS) is being designed, integrated, and tested in the Space Systems Design Lab (SSDL). This presentation summarizes the current status of the RECONSO ADCS, verification activities, and results. First, sun sensors were built in-house alongside algorithms to convert raw sensor output. An Extended Kalman Filter was created for orbit determination. Control algorithms are being developed to meet the payload’s pointing requirements. Finally, a subsystem verification test is being produced to ensure that subsystem requirements are met.
"Development of an Electrical Power System for the TARGIT CubeSat"
As more ambitious scientific experiments are placed onto CubeSat missions, there is an important balance to strike between scientific payload and system necessities required for the CubeSat to operate. One of the most important of these necessities is the electrical power system, the heart of the satellite. This presentation outlines the design and decision-making process behind the electrical architecture of the TARGIT mission through trade studies, simulations, and testing. Beyond the mission timeline of TARGIT, a focus was made on developing an effective method of solar panel fabrication for future Georgia Tech CubeSat missions.
“Designing Systems of Systems: The Coupling between Sensing Capabilities & Agent Behaviors”
Attritable Unmanned Aerial Vehicles have been gathering interest as reusable military platforms that fulfill mission requirements yet are low enough in cost to be expendable. With a large array of possible sensor suites available, the on-board sensors carried by some or all of these platforms are a large driving factor in the cost and design of these vehicles. This presentation will cover methodologies being used to explore this large design space and discover the correlation between sensor choice, behavior of the vehicle swarm, mission effectiveness, and overall cost.
"Characterization Of Transient Blowout Dynamics Of A Swirl Stabilized Flame Using Simultaneous OH And CH2O PLIF"
Gas turbine engines have operational limits that are heavily governed by the operational limits of the combustor. One of the key concerns while operating a combustor is the possibility of rich or lean blowoff. Focused research is done to study the lean blow off phenomenon, particularly to aid in its prediction so as to enhance the operational envelopes of gas turbine engines. This presentation will deal with methods to better characterize lean blowoff in an unwrapped annular swirling flow combustor.
"3-D Stress-Strain Histories for Composite Beams in Nonlinear Transient Structural Analysis"
This work demonstrates the capability of in-house tools developed in Prof. Hodges' group to obtain the 3-D stress-strain histories for composite beams undergoing dynamic loading in a nonlinear transient structural analysis. In the present work, we used isotropic and composite cantilever beams to demonstrate the results, which have been compared with results obtained from commercially available FEA tools such as ANSYS. It is observed that the use of in-house tools, VABS, and GEBT for the transient structural analysis is significantly advantageous over 3-D FEM. The time required for such an analysis greatly reduces from the order of multiple hours to a few minutes just for an isotropic beam. This will no doubt provide a cutting edge tool for preliminary structural designs as the preliminary design relies on approximations and simplistic analytical models which are neither reliable nor advantageous over tools based on high fidelity theory.
"Signal Processing and Software Development for the RANGE CubeSat Mission"
As Georgia Tech research groups continue to develop space systems in the years ahead, a robust and sophisticated ground support system will be an invaluable tool in tracking and maintaining simultaneous communications with several orbiting bodies. This research concerns creating software to track and communicate with RANGE, Georgia Tech’s first CubeSat mission, and diagnosing hardware and software failures as they occur to build up a capable satellite ground station infrastructure. Signal processing tools like the Python-based GNU Radio were used to develop scripts to decode radio signals to binary packets, and knowledge of the data packet formatting is used to remove data frames and output actual sensor measurements. Beyond realizing the science objectives of the RANGE mission itself, a focus is also placed on developing the software in a robust way to allow future automation of manual processes, and noting the difficulties with RANGE to inform the development of future flight software and communication system designs. This presentation will introduce GNU Radio software to demonstrate the general method of signal processing with RANGE, and an introduction to the ground station operations in Van Leer.