* Note that appearance on this list does not guarantee that the
abstract has been or will be accepted. All submitted abstracts
will be reviewed for suitability and technical content.
Acceptance will be confirmed via email with the submitting author.
Acoustics & Applied Aerodynamics
Abstract ID: 50DCASS-021
Aerodynamic Valve Integration and Performance Analysis in Valveless Pulsejets
Remya Nair
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
Previous research conducted at the University of Cincinnati extensively examined the dynamics and performance of Curtis-Dyna valved pulsejets across twelve geometric configurations, challenging traditional assumptions about their operation. These studies revealed that Curtis-Dyna pulsejets operated predominantly as Helmholtz resonators rather than the commonly assumed quarter-wave tubes. Experimental and computational investigations showed that parameters such as tailpipe length, combustor volume, and nozzle flare significantly influenced resonance modes, operating frequencies, thrust, and pressure dynamics. High-speed visualization techniques, coupled with pressure and ionization data, provided new insights into the interplay between combustion and fluid dynamics. It was observed that rapid reed valve closure generated toroidal vortices, with secondary shedding vortices sustaining the unsteady operation. The addition of nozzle flares enhanced thrust and stability, with optimal performance achieved using medium-length tailpipes and larger combustors. Different fuels were tested, with ethanol producing higher pressure peaks but less stable operation compared to gasoline. Critical findings included twin-peak pressure behaviors, shock wave generation during combustion, and flow-flame interactions, which dictated operational stability and led to mixed resonance modes arising from complex flow-field dynamics.
Research into valveless pulsejets has also made significant advancements, employing experimental, computational, and theoretical approaches to optimize their design and performance. Studies on small-scale engines, including the smallest ever reported at 8 cm in length, highlighted how inlet geometry and fuel type affected thrust and efficiency. Engines with rearward-facing inlets demonstrated significantly higher thrust but were less efficient. Numerical simulations using CFD (CFX) and combustion modeling offered insights into gas dynamics, acoustics, and chemical kinetics, validated by experimental data. Acoustic modeling treated valveless pulsejets as combinations of Helmholtz resonators and wave tubes, accurately predicting operating frequencies and optimal geometries. Additional research revealed that peak thrust occurs when inlet and exhaust frequencies align, emphasizing the critical role of geometry in ensuring adequate air intake, fuel-air mixing, and pressure wave generation.
Building on this foundation, our research seeks to integrate aerodynamic valves into the existing Curtis-Dyna pulsejet design to enhance performance and efficiency. The Long Combustor-Medium Tailpipe-Flare Off configuration has been identified as the optimal baseline for this integration. We plan to investigate the effects of various aerodynamic valve geometries on pressure dynamics, thrust, and stability. Acoustic analyses will be conducted to understand how the aerodynamic valve interacts with the combustion and flow-field dynamics. Our approach also includes studying backflow suppression mechanisms and optimizing valve designs to minimize pressure drops and enhance overall performance. These efforts aim to bridge gaps in the current understanding of valved and valveless pulsejets, paving the way for their practical implementation as cost-effective, efficient propulsion systems.
Abstract ID: 50DCASS-042
Use of Sweeping Jet Actuator for Supersonic Jet Noise Reduction
Kaurab Gautam
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
Sweeping Jet Actuators (SJAs) are fluidic oscillators that generate sweeping jets at specific frequencies without relying on any moving mechanical components. These actuators have been widely employed in flow control applications due to their simplicity and robustness. The sweeping frequency of SJAs is inherently dependent on their geometric design and operating conditions, making them highly versatile for tailored applications. In this study, we investigate the influence of SJA geometry and operating parameters on their sweeping frequency and evaluate their potential for jet noise reduction in supersonic flows. SJAs were integrated into multiple nozzle configurations, including rectangular and GE F404 nozzles, to study their impact on acoustic and flow field characteristics. Comprehensive acoustic measurements revealed that the implementation of SJAs resulted in a 3
Abstract ID: 50DCASS-049
Exploration of Slot Injection for Noise Reduction on 2D-CD Style Nozzles
James Cramer
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
Acoustic emissions from military-style aircraft have become a topic of upmost importance, driven by complaints of excess noise around military installations and VA claims from military personnel, specifically around ship-borne applications [1]. Over the past year, slot style fluidic injections on 2D-CD style nozzles have been studied at the Gas dynamics and propulsion Lab and has shown great promise. Early research indicated a 3db OASPL reduction beyond the contribution of screech, with reduction in Broad-Band shock noise and large-scale mixing noise with injection mass flows under 6% of core flow. The study is currently ongoing and is looking at the effect of injection angles, injection pressure ratio, slot thickness and number of injectors on the shock structures and shear layers.
[1] Allan Aubert, and Richard McKinley, "Measurements of Jet Noise Aboard US Navy Aircraft Carriers," AIAA Centennial of Naval Aviation Forum @ Years of Achievement and Progress, American Institute of Aeronautics and Astronautics, 2011,
Abstract ID: 50DCASS-070
The Impact of Upstream Flow Variations on Acoustic Behavior in overexpanded nozzle
Arshad Mohammed
University of Cincinnati
Gautam Kaurab
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
Conventional nozzle design typically emphasizes parameters such as area ratio, design Mach number, and nozzle height or equivalent diameter, with limited attention given to the inlet flow conditions. This study investigates the performance of twin rectangular nozzles with distinct flow characteristics: one incorporating a diffused, smooth flow path, and the other featuring an abrupt area contraction, commonly referred to as plug flow. The analysis focuses on the overexpanded regime, specifically for a nozzle pressure ratio of 3, highlighting the differences in behavior between these nozzle configurations.
Aircraft and UAS Design & Applications
Abstract ID: 50DCASS-002
Multi-Fidelity Aerostructural Optimization of ESAV including Flutter Constraints
Markus Rumpfkeil
University of Dayton
Phil Beran
Air Force Research Laboratory
The traditional design process relies heavily upon low-fidelity models for time and cost savings. However, the reduced accuracy of low-fidelity tools can often lead to the discovery of inadequacies late in the design process. These deficiencies can result either in costly redesigns or the acceptance of worse-than-advertised performances. Multi-fidelity methods attempt to blend the reduced cost of low-fidelity models with the increased accuracy and reliability of high-fidelity ones. In this presentation, a gradient-based multi-fidelity constrained optimization approach is applied to an aeroelastic drag minimization of an efficient supersonic air vehicle (ESAV). The highest and lowest fidelity levels considered are Euler and panel solutions, respectively, all combined with a modal structural solver. Imposed constraints are for lift, and pitching moments as well as flutter suppression. The results show that a drag reduction is possible while satisfying all constraints and that a developed multi-fidelity optimizer performs better than a purely high-fidelity one.
Abstract ID: 50DCASS-003
Development of a High-Speed Vehicle Conceptual Design Process using ADAPT Framework
Samuel Atchison
AFIT Contractor
Matthew Madayag
Air Force Institute of Technology
Jose Camberos
Air Force Institute of Technology
The significant multidisciplinary challenges of high-speed vehicle design and development make the design process for these vehicles more complex and integrated than that of traditional aircraft. Along with this integration, the higher speeds invite more physics that the selected analysis tools must capture for accurate analysis and evaluation. To meet this need, design frameworks allow the integration of analysis tools at selected (and variable) fidelity levels. One such design tool built specifically for aircraft design is the Aircraft Design, Analysis, Performance, and Tradespace (ADAPT) framework. This study describes the implementation of a design process for high-speed vehicles through the initial integration of the Engineering Sketch Pad (ESP), Configuration-Based Aerodynamics (CBAero), and Automatic Structural Layout Tool (AutoSaLT) as plugins within the ADAPT framework. We discuss the integration process of these tools in detail to show how we developed the plugins and linked them in ADAPT's workflow. We then show geometric, material, and structural layout trade studies on a selected geometry to showcase the capabilities of ADAPT, the potential of developing a design process within this framework, and to provide some insights from the trade study results. Distribution Statement A: Approved for Public Release; Distribution Unlimited. Public Release # 88ABW-2024-0904
Abstract ID: 50DCASS-056
Aerospace Vehicle Conceptual Design Trade Studies Using ADAPT
Bryce Helmer
Air Force Institute of Technology
Samuel C. Atchison
Air Force Institute of Technology
Jose A. Camberos
Air Force Institute of Technology
Driven by the transition to near-peer conflict, interest in aerospace
vehicles has continued in recent years, creating a demand for new designs.
The process of designing aerospace vehicles comes with challenges,
fundamentally different from traditional design techniques and methodology
due to the interconnectedness of all the aircraft’s components. The geometry,
aerodynamics, propulsion system, structure, etc. are all affected by one
another and thus must all be considered during the entirety of the design
process. One component cannot be designed in isolation from the other
components and expect the design to integrate effectively. This leads to the
need for programs that can incorporate the various aircraft components and
effectively perform studies on design and flight parameter changes. The
Aircraft, Design, Analysis, Performance, and Tradespace (ADAPT) program
can be used to execute trade studies on aerospace vehicles. This project
explores the conceptual design tradespace by leveraging the novel fixed-wing
aircraft design environment ADAPT. The project will analyze the sensitivity
of vehicle flight characteristics to changes in various design and flight
parameters. ADAPT, in conjunction with disciplinary tools such as ESP
(geometry), CBAero (aerodynamics), SQu1D (propulsion), and AutoSaLT
(structures), will be used to drive the workflow. The ADAPT architecture
will be tested and exercised by running various trade studies and current
vehicle configurations will be assessed for flight performance using the tested
workflow. ADAPT will then be leveraged to conduct configuration
optimization studies across the design space.
Distribution Statement A: Approved for Public Release; Distribution Unlimited. Public Release #88ABW-2025-0047
Abstract ID: 50DCASS-072
Flight Test Validation of Tandem Propeller Performance with Vertical Offset
Michael Foster
University of Dayton
Jessica DeMoor
University of Dayton
Sidaard Gunasekaran
University of Dayton
Jielong Cai
University of Dayton
The power efficiency of a multirotor can be improved in edgewise flight by positioning the aft rotors above the plane of the front rotors. To validate the findings of a previous experimental wind tunnel campaign at the University of Dayton Low Speed Wind Tunnel (UD-LSWT), a custom-built multirotor was developed to conduct flight tests examining staggered rotor geometry. The multirotor accommodated multiple vertical offset configurations of the aft rotors and utilized GPS to sustain altitude and velocity in edgewise flight, thereby ensuring repeatable flight paths. Test runs were repeated several times for each vertical offset configuration at multiple flight speeds. Across test cases, the center of gravity of the multirotor remained centered across its longitudinal and lateral axes, and its mass was held constant, to isolate the effects of vertical offset on performance. A custom-built anemometer and wind direction sensor was developed to measure and record ambient weather conditions throughout testing. All flight test data utilized in analysis exhibited calm ambient conditions, as recorded by the anemometer and wind-direction sensor.
Flight test data of the current study validates the findings of the previous wind tunnel study by demonstrating a significant improvement in multirotor power consumption efficiency as a result of vertical offset of the rear rotors. Specifically, at averaged advance ratios between 0.15 and 0.45, a vertical offset of 20% of the propeller diameter led to a reduction in power consumption of over 15% as compared to the baseline configuration without offset. Further increases in the vertical offset of the rear rotors yielded only minimal further improvement in efficiency. The findings of the current study further suggest the feasibility of implementing vertical rotor offset to improve multirotor efficiency.
Abstract ID: 50DCASS-079
On coupling approaches for mixed-fidelity aeropropulsive analysis and optimization
Neal Novotny
Air Force Research Laboratory
Jan Kiviaho
University of Dayton Research Institute
Awaiting public release.
Abstract ID: 50DCASS-080
Gradient-based aeroelastic optimization of a generic flying wing model
Neal Novotny
Air Force Research Laboratory
David Sandler
University of Dayton Research Institute
Nathan Wukie
Air Force Research Laboratory
Physics-based design processes that account for discipline coupling are critical to the design of highly integrated and novel air vehicle configurations. In this work, we implement a framework for gradient-based, high-fidelity, aeroelastic design optimization based on Engineering Sketch Pad, FUNtoFEM, FUN3D, TACS, and OpenMDAO. A test-bed generic flying wing model was constructed with a parameterized aerodynamic outer mold line and internal structure in order to facilitate capability demonstration. In depth discussion of the model configuration, design variables and geometric design variables is presented. Finally, the framework is demonstrated through two example cases. The first example case is a study on the effect of the KS-weight parameter and number of design variables on optimization convergence. Results demonstrate lowering the KS-weight resulted in an increasingly conservative vehicle design but did not predictably reduce the computational work required for optimization but did improve the tendency for the optimization to converge. Additionally, increasing the number of design variables increased the tendency of the optimization to converge. Finally, a demonstration of gradient-based, static, aeroelastic design optimization formulated to minimize a blended drag and mass functional subject to lift, stress and vehicle volume constraints is presented. The demonstration results exhibit reductions in drag and structural mass while satisfying aerodynamic and structural constraints for a feasible design.
Combustion & Fuels
Abstract ID: 50DCASS-059
Altitude System Design for Two-stroke Engines Between 50-200 cc
Jacob L. Ahles
Air Force Institute of Technology
Marc Polanka
Air Force Institute of Technology
Joseph K. Ausserer
Air Force Institute of Technology
Thomas C. Balaj
Air Force Institute of Technology
Jacob A. Baranski
Innovative Scientific Solutions Inc.
An altitude system for testing small 50-200 cc piston engines was designed and built by the Air Force Institute of Technology (AFIT) and the Air Force Research Lab (AFRL). When piston engines operate at altitude the air density decreases which results in decreased power output and changes to engine efficiency. Testing engines at ground level with simulated altitude conditions is a cost effective and safe way to collect engine data. This altitude system used a turbocharger cart, plenum, and supercharger cart to condition the air. The turbocharger cart primarily works to lower the air temperature by expanding compressed gas through its turbine. The supercharger cart primarily works to lower the air pressure. The plenum connects the turbocharger cart and supercharger cart with the engine and allows a bypass to cool the engine exhaust and allows the system to run at a steady state separate from the engines intermittent and transient conditions. Since piston engines are susceptible to intake and exhaust tuning, care was taken in the design of the intake and exhaust sections of the altitude system. To better understand the impact of the intake and exhaust, high speed pressure transducers were installed for comparison between ground level operation data and altitude operation data.
Abstract ID: 50DCASS-024
Analysis of Rotating Detonation Wave Interaction with Reactive Core Flow
Tyler Pritschau
University of Cincinnati
Rachel Wiggins
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
Interaction between between a pre-mixed center flow, and a rotating detonation wave sustained in a hollow Rotating Detonation Combustor (RDC) is captured with Chemiluminescence
imaging. In this study, a RDC is supplied Hydrogen and air while a core flow is supplied with
Ethylene and air. The dissimilar fuels are chosen deliberately so that the Chemiluminescence
properties of the two reactants streams differ. CH* Chemiluminescence is dominated by the
core reaction because only trivial amounts of carbon are present in the RDC reactants. This
allows OH* Chemiluminescence to be used to characterize system-wide heat release profiles
while CH* Chemiluminescence depicts primarily heat release from core reactants. Detailed analysis of phase-averaged imaging is conducted across a wide range of flow conditions to characterize the changes in this interaction as relative fueling and oxidizer flux vary.
Abstract ID: 50DCASS-027
Near Lean Blowout Behavior in Lean Direct Injection Nozzle
Shyam Murali
University of Cincinnati
Yuvi Nanda, Tharun Reddy, Grace Fischer, Kranthi Yellugari
University of Cincinnati
Dr. Ephraim Gutmark
University of Cincinnati
To better design and operate the propulsion systems as per new emission norms require very low NOx emissions pushing propulsion systems to operate in lean fuel regime. Operating in lean conditions reduces thermal NOx as well as reduce fuel consumption. As fuel-air mixture approaches lean conditions, flames stability is drastically reduced and experiences local extinction and re-ignition and thus it is crucial to study and understand this phenomenon. This research paper investigates near lean blowout instabilities in TARS combustor burning propane fuel operating in non-premixed mode. The TARS combustor was operated at atmospheric back pressure and the inlet air is preheated to 600K. Synchronized sound pressure, OH chemiluminescence imaging and Z-type schlieren imaging was acquired at 40kHz and 50kHz for imaging respectively. Nonlinear signal analysis like symbolic recurrence and Hurst exponent were used to understand the complex nonlinear behavior of the system and SPOD to understand the flame-flow dynamics from the OH chemiluminescence and schlieren images. In the future Jet-A will be used to compare near lean blowout instabilities using liquid and gaseous fuels.
Abstract ID: 50DCASS-034
Effects of Fuel-Throttling on the Operation of a Flow-Through Rotating Detonation Engine
Anthony Centofanti
University of Cincinnati
Peter Glaubitz
University of Cincinnati
Tyler Pritschau
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
Rotating Detonation Engines (RDEs) offer a potential step increase pressure gain compared to traditional combustors which undergo a loss in stagnation pressure. Recent studies have shown promising trends in the development of the technology; however, many technical challenges are still present. One of these challenges is understanding the effect of throttling the RDE. This poses a challenge due to the greater heat release characteristic of detonations, which require the test article to be cooled to be effectively studied on the time scale where supply valves may respond to throttling commands. The annular geometry of most RDEs compounds this problem due to the thermal boundary of the outerbody and centerbody. Hollow RDEs, which omit the centerbody, have been studied previously and are a prime candidate for water cooling due to the simplicity of only the outerbody requiring cooling. This idea can be taken further by adding a separate core flow in place of the center body, which results in the flow-through configuration. Results of fuel-throttling the flow-through RDE show a clear extension of operability when throttled to a rich condition once a stable detonation wave is established as opposed to solely igniting at the same rich condition. Additionally, if the resulting mode exhibits some thermoacoustic behavior it may still differ from the ignition conditions.
Abstract ID: 50DCASS-037
BOS Analysis of TARS Combustor
Grace Fischer
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
Shyam Murali
University of Cincinnati
Yuvi Nanda
University of Cincinnati
Tharun Reddy
University of Cincinnati
Background Oriented Schlieren (BOS) is used to capture images of flow experiments conducted in a in a Triple Annular Research Swirler (TARS). BOS is advantageous as it is a non-invasive way to capture flow fluctuations and is simple to set up as it has few components. The TARS combustor was operated at atmospheric pressure and ambient air temperature. TARS was operated in a non-premixed mode and is a swirl-stabilized combustor with a lean-direct injection fuel nozzle. It consists of three air passages that each have their own swirler that can be changed independent of the others. This allows differing configurations of TARS by changing the swirlers either to different swirling angles or rotating directions to form differing swirling flow fields. Using the TARS configuration C4500C45 (where C indicates counter rotating) two cases were run with an equivalence ratio of 0.57 and 0.85 respectively both being stable combustion. Images were gathered using the Background Oriented Schlieren imaging technique, allowing images to be taken of the density fluctuations to compare the BOS imaging capabilities at two different equivalence ratios. Background oriented schlieren captures this by having a camera view a noisy background through the flow and see the density changes in the flow causing the light to refract and displace the background pattern. Future plans include further analyzing these images to gain a visualization of the temperature gradients of the flow.
Abstract ID: 50DCASS-039
Unwrapped Detonation Wave Structure from Thermal Imaging
Peter Glaubitz
University of Cincinnati
Tyler Pritschau
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
A mid-wavelength infrared radiation (MWIR) camera recorded side-view imaging of an optically accessible flow-through rotating detonation combustor (FT-RDC) at approximately 25,000 fps. The region recorded is significantly limited due to hardware constraints of the camera. Through the use of additional diagnostic equipment including additional highspeed camera recording aft imaging, and highspeed pressure transducers, the highspeed thermal images can be associated with a certain detonation wave location. An “unwrapped” reconstruction of the detonation wave can then be generated by stitching the thermal images together to create a phase-averaged image. The resulting image will reveal key structures in the FT-RDC flow field including the detonation wave, oblique shock, and refill zone. Further work to calibrate the thermal camera will reveal temperature profiles of the FT-RDC flow field.
Abstract ID: 50DCASS-040
Nonlinear behavior of thermoacoustic instabilities in swirl-stabilized combustor
Yuvi Nanda
University of Cincinnati
Shyam Muralidharan
University of Cincinnati
Tharun Reddy
University of Cincinnati
Ephraim Gutmark
University of Cincinnati
This study incorporates the Triple Annular Research Swirler (TARS) which is a lean direct injection nozzle, installed in an atmospheric combustor. The nozzle is composed of two axial swirlers and one radial outer swirler arranged concentrically. Fuel is injected through two independently controlled fuel circuits; the Pilot which supplies fuel to the passage between the axial swirlers, and the Main, which supplies fuel to the outer passage. Thermoacoustic instabilities arising from the coupling between the unsteady heat release and acoustic field have been identified in this combustor by altering the fuel supply to the Pilot and Main fuel circuits. The study aims to examine the nonlinear characteristics of these combustion oscillations by implementing phase space reconstruction methods such as phase portraits and recurrence plots to scalar pressure measurements. The corresponding flame dynamics are investigated by employing high-speed simultaneous OH* chemiluminescence and schlieren imaging which allow the capture of the flame heat release along with the flow field dynamics of the system.
Abstract ID: 50DCASS-044
Simultaneous Z-type Schlieren and OH Chemiluminescence for the Analysis of Combustion Dynamics
Tharun Srinivas Karnam Reddy
University of Cincinnati
Ephraim J. Gutmark
University of Cincinnati
Thermoacoustic instability is a phenomenon that occurs in jet and rocket engines due to the interaction between heat release and acoustic waves, potentially leading to efficiency losses and engine failure. Schlieren is an optical imaging technique that captures density variations by detecting changes in the refractive index as light passes through regions of differing density.
In this study, thermoacoustic instability in a Triple Annular Research Swirler nozzle is examined using the Schlieren visualization technique. Details of the Z-type schlieren setup and comparisons of different light cutoffs at the focal point using a knife-edge are presented. The occurrence of thermoacoustic instability depends on various factors, primarily operating conditions such as temperature, equivalence ratio, fuel type, nozzle configuration, and combustor design.
The underlying dynamics of thermoacoustic instability with rich mixture (fuel-to-air ratio is higher than the stoichiometric ratio) conditions are analyzed using simultaneous Z-type schlieren and OH* chemiluminescence techniques. Fast Fourier Transform (FFT) and Spatial Proper Orthogonal Decomposition (SPOD) methods are employed to study the behavior of thermoacoustic instabilities across various operating conditions.
Abstract ID: 50DCASS-062
Thin Filament Pyrometry to Measure Temperature in a Rotating Detonation Engine
Theodore B. Guetig
Air Force Institute of Technology
Marc Polanka
Air Force Institute of Technology
Larry P. Goss
Innovative Scientific Solutions Inc.
Brian C. Sell
Innovative Scientific Solutions Inc.
Rotating Detonation Engine (RDE) technology is becoming more practical and is closer to implementation. The detonation combustion of an RDE proves beneficial by creating a pressure rise across the combustor rather than a pressure drop seen in traditional aircraft combustors. The increased pressure allows the turbine to extract more work. Coupled with this high pressure is a high temperature in the detonation engine that is difficult to measure. Thin-Filament Pyrometry (TFP) was performed on a 6-in RDE to measure these high temperatures. While TFP has previously provided decent results, the quality of TFP depends on the frequency response and the diameter of the filaments. These characteristics are inversely related and must be balanced. Temperature profiles were obtained over a variety of massflows, equivalence ratios, and at different axial locations within the RDE providing insight into the mixing process and where the heat release occurs. Successful determination of these temperature profiles within the combustor with TFP provides information that could lead to improved thermal management techniques for these engines including film cooling. This would subsequently enable longer operation and bring RDEs closer to implementation.
Computational Fluid Dynamics
Abstract ID: 50DCASS-067
Development of a Polar Invariant Map for Turbulence Modeling
James Wnek
Wright State University
Christopher Schrock
Air Force Research Laboratory
Eric Wolf
Ohio Aerospace Institute
Mitch Wolff
Wright State University
Invariant maps are a useful tool for turbulence modeling, allowing different turbulent states to be visualized in an interpretable manner and providing a mathematical framework to analyze or enforce realizability. The rapid growth of machine learning-enhanced turbulence modeling research has renewed interest in them, but current invariant maps are limited in their versatility due to the need for potentially costly coordinate transformations or eigendecomposition at each point in the flow field. Here we introduce a new polar turbulence invariant map based on an angle describing the ratio of the principal stresses and a scalar describing the anisotropy magnitude relative to a maximum value. The polar invariant map reframes realizability in terms of a limiting anisotropy magnitude. This leads to a direct link between realizability and general eddy viscosity models via a scaling factor without the need for coordinate transformations or explicit eigendecomposition. The relationships to the Lumley triangle and barycentric map are illustrated through examples of fully developed channel flow and square duct flow. Applications to machine learning-enhanced turbulence modeling are discussed. The polar invariant map provides a foundation for new approaches to enforcing realizability constraints in Reynolds-averaged turbulence modeling. Distribution Statement A: Approved for Public Release; Distribution is Unlimited. PA# AFRL-2024-5686.
Abstract ID: 50DCASS-101
Computational Fluid Dynamics for Distributed Lift Configurations
Ainsley Rousseau
The Ohio State University
Scott Sherer
Air Force Research Laboratory
Michael Mongin
Air Force Research Laboratory
Aaron Altman
Air Force Research Laboratory
A computational fluid dynamics study was performed using the OVERFLOW CFD solver with a Reynolds-Averaged Navier-Stokes (RANS) turbulence model on a multi-wing distributed lift configuration consisting of four lifting surfaces on each side of the aircraft fuselage. A traditional baseline monowing configuration was also simulated for comparison purposes. Two values of Reynolds Numbers were simulated for both configurations to investigate the sensitivity of the force components to this parameter across a range of angles of attack. Individual lift and drag coefficients as well as their ratio (L/D) on the entire configuration were compared to experimental data and showed reasonably good agreement given the variation between computational and experimental set-ups. Force components on the individual wings were also examined in the multi-wing case. Results showed that the first wing of the multi-wing configuration behaves similarly to the monowing, with the remaining wings providing additional but reduced lift compared to the first wing. The lift-curve slopes for the individual wings was found to decrease from front to back at low angles of attack of less than approximately 10 degrees, but would return to approximately the lift-curve slope of the leading wing at greater than ten degrees. Distribution Statement A: Approved for Public Release; Distribution is Unlimited. PA\# AFRL-2024-3408
Abstract ID: 50DCASS-082
Implicit Large Eddy Simulations of F-BoLT Boundary Layer Transition
Tim Leger
Ohio Aerospace Institute
A set of implicit Large Eddy Simulations were performed to complement an experimental test campaign, in which the effects of leading edge curvature on boundary layer transition were investigated. Six scenarios were considered, two different nose curvatures, each at three Reynolds numbers. To initiate transition, disturbances were introduced outside the bow shock. Stanton number contours from the simulations were in generally good agreement with the experimental measurements. The simulations revealed a maximum difference of 15% in drag and acreage heating, along with a nearly 85% difference in turbulent surface area between the two nose curvatures at the same Reynolds number. These
quantities were also estimated from the laminar solution using Reθ/Me, and from simulations using the Coder Spalart-Almaras transition model, but did not show as much of a difference between nose curvatures. Auto spectral analyses were performed on surface pressure along a line outboard of the streamwise vortex. The results show that while some energy is being dissipated by the grid near the end of the vehicle, the transition region is well captured for all six scenarios. Distribution Statement A: Approved for Public Release; Distribution is Unlimited. AFRL-2024-4564; Cleared 19 AUG 2024
Abstract ID: 50DCASS-097
CFD Analysis of Fuel Mixing in a Flow-Through Rotating Detonation Combustor
Bret Lane
University of Cincinnati
Ephraim J. Gutmark
University of Cincinnati
This work studies the pre-ignition flow field of a flow-through rotating detonation combustor (FTRDC) with gaseous ethylene injected into the Core flow as well as the RDC using CFD. With dual-flows, the mixing that occurs within the individual flows as well as in the shear layer impacts local fuel distribution, a critical parameter for successful ignition to stable detonation of the FTRDC. Changes in local equivalence ratio, displayed as mixture fraction, with varying parameters are studied to determine trends in fuel distribution at different axial locations of the combustor. Interesting results are obtained with varying air mass flow rates, equivalence ratios, and air temperatures for both Core and RDC flows. Geometric changes in Core area restriction with a nozzle and Core injection location are also studied. Results from the computational work are compared to experimental data to corroborate on conclusions about the limitations to obtain stable detonations in different regions of ignitability testing as well as the key phenomenon observed in both.
Data Analysis & Uncertainty Quantification
Abstract ID: 50DCASS-004
Uncertainty Propagation in Transient Heat Transfer from an Extended Surface
Edwin Forster
Air Force Research Laboratory
Rama Gorla
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-103
Multifidelity Surrogate Modeling of a Generic Hypersonic Vehicle
Evan Burke
Air Force Research Laboratory
Markus Rumpfkeil
University of Dayton
This work presents a case study of multi-fidelity surrogate models for predicting the
aerodynamic responses of a generic hypersonic vehicle (GHV) over a five-dimensional input domain. Multi-fidelity surrogates enable the use of computationally expensive high-fidelity data in conjunction with lower-fidelity data for assessment of a system. Surrogate models are created using polynomial chaos expansion. Three distinct aerodynamic analysis codes of increasing fidelity are considered: CBAERO, Cart3D, and FUN3D, and a grid convergence study is presented for each. Mono- and multi-fidelity surrogate models are created to predict lift, drag, and pitching moment responses of the GHV. Monofidelity surrogate models are shown to be capable of capturing the aerodynamic response with a relatively sparse number of samples. Bi- and tri-fidelity surrogates provide better accuracy than monofidelity in comparison to a high-fidelity FUN3D database, especially when high-fidelity samples are limited.
Digital Engineering
Abstract ID: 50DCASS-066
Variable-Fidelity Analysis of Aerothermal Effects on a Generic Aerospace Vehicle
Andrew Setzer
Air Force Institute of Technology
Jose A. Camberos
Air Force Institute of Technology
Modern day aircraft development relies heavily on the use of digital twins for various reasons,
this includes initial testing, finalizing designs, and proving air worthiness. In many cases, design
iterations are populated into a shared file which is then disseminated to engineering groups for
additional testing and retesting, this is done to ensure that new design changes maintain, at a
minimum, previous performance characteristics and safety. This design framework is highly
labor intensive, subject to user error, and inefficient for large-scale projects; some researchers
have therefore suggested the use of Model Based Systems Engineering (MBSE) software to
expedite and automate design processes. In order to achieve this, the MBSE software would
require numerous points of data from a host of aerodynamic solvers at different fidelities to
accurately represent the characteristics of a given design. Recently, the CFD Research Center
(CFDRC) acquired funding from the Ohio Federal Research Network (OFRN) to build a MBSE
aircraft development system, which is in need of multi-fidelity aerothermal data. Therefore, in
support of CFDRC, this project will generate and compare multi-fidelity aerothermal data of a
Generic Hypersonic Vehicle (GHV) in a ballistic and phugoid trajectory. Configuration-Based
Aerodynamics (CBAero) will act as the low-fidelity solver used to generate data, while Kestrel
will act as the research’s high-fidelity solver. The data will then be compared to the existing
Aeroheating and Thermal Analysis Code v2 (ATAC2) data which is already implemented in
the MBSE design model. Additional analysis will occur post-data-validation to determine
the differences between each solver’s workflow in the MBSE design model, highlighting the
strengths and weaknesses of alternate fidelity solvers when processing aircraft design in an
automated manner. Distribution Statement A: Approved for Public Release; Distribution Unlimited. Public Release # 88ABW-2025-0050
Abstract ID: 50DCASS-051
Automating Geometry Generation, Meshing, and Aerodynamic Modeling
Christopher Humphrey
University of Cincinnati
Jose Camberos
Air Force Institute of Technology
With an ongoing interest in aircraft design, researchers need to fully invest and utilize digital simulation tools. Assessing the accuracy of these software tools and techniques against accepted results requires careful consideration. NASA's low-fidelity tool Configuration Based Aerodynamics (CBAero) provides accurate and efficient results at supersonic speeds. Benefits include the increased ability to explore a (design) trade space and reduce the time needed to test new configurations. Executing the process of design exploration with simulations requires the automation of a geometry to simulation procedure. The research performed included three parts: First, create and link the geometry and the meshing. Second, input and simulate the aerodynamics with CBAero. Third, analyze the data generated from CBAero. The first part used Engineering Sketch Pad (ESP) to generate the geometry of a generic aircraft. A Python script with AFLR4 is utilized to generate a two-dimensional surface mesh, creating a link between ESP and the mesh. The Python script controls the automatic unstructured meshing and has several options to refine the mesh. The second piece uses CBAero to simulate a variety of flight conditions. The third section added the ability to change the output data formats and the ability to analyze the results. The final result generates a streamlined process from geometry generation and meshing to simulation data and analysis in a simple and easy to use package.
Abstract ID: 50DCASS-054
Introduction to AIAA's Digital Engineering Integration Committee
Rick Graves
Air Force Research Laboratory
AIAA’s Digital Engineering Integration Committee (DEIC) was chartered in 2020 to
accelerate digital transformation across the American Institute of Aeronautics and Astronautics, the premier aerospace professional organization in the United States.
This collaborative group is a blend of industry, academic, and Government practitioners
intent on digital transformation promotion across the aerospace domain, with application
to commercial and Department of Defense programs. The depth of digital engineering
activities organized by DEIC has grown since 2020 to its current state at AIAA SciTech
2025 of 18 technical working sessions and 4 after-hours workshops on emerging digital transformation topics. Integrated products teams within DEIC cover diverse topics such as
digital thread and digital twin best practices, and design and implementation of digital ecosystems. This committee also address emerging capabilities such as artificial
intelligence, machine learning, and rigorous validation and verification for aerospace
systems. In this presentation, the authors will describe the goals and objectives of
AIAA’s Digital Engineering Integration Committee, and discuss near term and longer term
DEIC objectives, with a particular focus on STEM outreach for digital engineering for
Region 3.
Abstract ID: 50DCASS-055
Digital Materiel Management for a Bio-Inspired Rotating Empennage Aircraft
Rick Graves
Air Force Research Laboratory
Digital Materiel Management (DMM) is an approach to accelerate the acquisition
materiel lifecycle to satisfy Air Force and Air Force Materiel Command strategic
objectives. A digital materiel management approach enables an accelerated and
synchronized delivery of integrated capabilities to a system’s primary stakeholders.
Using the DMM approach, models are developed and managed in lieu of traditional
documents, structured data replaces fragmented information, and digital collaboration
between stakeholders is emphasized to minimize the impact of traditional decision-making
silos. To realize this approach, it is critical that early phases of the engineering
lifecycle, especially those focused on basic research and science and technology, are
aligned with digital materiel management goals and objectives.
The U. S. Air Force Research Laboratory (AFRL) is currently researching the viability of
tactical aircraft that leverage bio-inspired control subsystems. The Bio-Inspired Rotating
Empennage (BIRE) concept modifies the traditional F-16 by removing the vertical tail, and
the aft section of the aircraft is redesigned to accommodate a rotating empennage with control surfaces resized to extend range and sustain performance. This concept, funded
by the Air Force Office of Scientific Research (AFOSR), has been researched extensively by
AFRL and academic partners over the last several years. Digital engineering implementation
for this concept will explore best practices for developing and communicating aerospace
system design changes across the F-16 digital engineering ecosystem. A digital materiel
management approach explored for the BIRE concept enables AFRL to rapidly share
authoritative data and models (digital artifacts) developed during science and technology
activities with the existing F-16 digital engineering ecosystem.
Abstract ID: 50DCASS-064
Implementation of a Digital Repository for Vehicle Analysis, Design, and Decision Making
Jack Bruer
Air Force Institute of Technology
Samuel C. Atchison
Air Force Institute of Technology
Jose A. Camberos
Air Force Institute of Technology
Aerospace vehicles continue to be driven by Great Power Competition and remain an area
of interest for the Department of Defense (DoD). To meet this challenge, the conceptual design and analysis of high speed vehicles must be rapidly conducted. One approach to accelerate these efforts relies on legacy historical data, including geometries and prior analysis results. To
support this effort, the authors propose developing, implementing, and curating a data repository of aerospace vehicles. This repository would provide an accessible and searchable catalogue of existing and proposed vehicle designs, including available research documents and data. The
effort is intended to support future projects for advanced research and technology development of high speed vehicles. By consolidating historical resources into a single database,
future projects can leverage off the shelf information for analysis, design, and decision making. This presentation will describe the implementation and organization of the envisioned data repository. Distribution Statement A: Approved for Public Release; Distribution Unlimited. Public Release # 88ABW-2025-0044
Abstract ID: 50DCASS-089
Leveraging MBSE for Integration and Stakeholder Engagement in Autonomous Aerial Refueling Flight Tests
Kevin Keth
Air Force Institute of Technology
David R. Jacques
Air Force Institute of Technology
The Department of Defense has pushed to implement Digital Engineering (DE) and Model-Based Systems Engineering (MBSE) into acquisition processes, publishing various supporting documents such as the Systems Engineering Guidebook, DoDI 5000.97 Digital Engineering, and DoD Reference Architecture Description. The adoption of DE aims to establish a digital authoritative source of truth accessible to all stakeholders responsible for system architecture. However, within multidisciplinary teams, individuals often come from diverse professional backgrounds unrelated to DE and systems engineering, making it challenging to fully leverage the benefits of the digital model. This paper describes an MBSE model that adopts a reference architecture used for small, unmanned aircraft systems (SUAS) and a validation framework to enable model-based design reviews (MBDR). The model is applied to an academic team researching Autonomous Aerial Refueling using SUAS as part of a broader effort to support digital transformation. Furthermore, the research team is observed and analyzed to identify additional technical views that can be developed within the MBSE model, enabling better integration of multidisciplinary knowledge and fostering synergy within the digital framework.
Abstract ID: 50DCASS-102
Uncertainty Propagation for a Hypersonic Vehicle through a Model Validation Hierarchy Framework
Alexander Wade
CFD Research Corporation
Andrew Kaminsky
CFD Research Corporation
Alexander Schreiner
CFD Research Corporation
Awaiting public release.
Experimental Methods
Abstract ID: 50DCASS-035
Characterizing the Wavelength and Angular Dependence of Radiation Properties in TPS Materials
Yejajul Hakim
University of Kentucky
Ahmed H. Yassin
University of Kentucky
Ayan Banerjee
University of Kentucky
Savio J. Poovathingal
University of Kentucky
Michael W. Renfro
University of Kentucky
Radiative heat transfer is a key factor in the thermal response of Thermal Protection System (TPS) materials during planetary re-entry missions with hypersonic vehicles. Although radiative heat flux accounts for only 25-35% of the total heat load during Earth's atmospheric entry, its impact becomes more significant in planetary entry missions to other celestial bodies. This study aims to measure the angle- and wavelength-dependent transmission, scattering, and backscattering properties of TPS materials with varying thicknesses, which are then used to calculate the absorption and scattering coefficients, as well as the refractive index of the material. In this experiment, the properties are measured for LI-2200 samples with varying thicknesses, ranging from 0.4 mm to 1.3 mm, which are enclosed between two transparent glass slabs, with a fiber-coupled stabilized light source directed at them. Two types of light sources are used to cover different wavelength ranges, from UV to near IR. To generate and collect collimated beams from these light sources, a two-lens tube system is installed at both the source and collection ends. The collection end is linked to a high-resolution spectrometer, covering a range from 200 nm to 1100 nm, and is positioned opposite the sample from the light source during transmission and scattering measurements. Both the sample and the receiver are mounted on motorized stages that allow for rotation, enabling adjustment of the incidence and collection angles. Twenty-five different angular combinations were measured for transmission and scattering, while forty-five combinations were recorded for backscattering. These measurements were normalized by comparing transmission through air for transmission and scattering through the sample, and reflection from two different types of mirrors for backscattering from the sample in both the ultraviolet and visible spectra. The absorption coefficient and refractive index is then computed from these measurements using Beer-Lambert's law, Fresnel's equations, and the Kramers
Flight Dynamics & Controls
Abstract ID: 50DCASS-007
6DOF Reentry Reachability Optimization of a Vehicle Resembling SpaceX's Starship
Emma Webb
Air Force Institute of Technology
Robert A. Bettinger
Air Force Institute of Technology
Six degree-of-freedom (6DOF) optimization problems are difficult to solve with collocation methods due to the varying time scale of the translational and rotational states. This research utilizes GPOPS-II transcription software to solve the 6DOF reentry reachability problem for a reentry vehicle resembling SpaceX's Starship (CRV-1). The maximum downrange and maximum cross-range problem is solved in 6DOF with a running cost function designed to decrease the frequency of the rotational states, which decreases computation time and aligns with realistic vehicle structural stability requirements. It is shown that a linear interpolation between collocation points in the optimal solution may not satisfy the problem dynamics; in order to numerically reconstruct the optimal solution, a Newtonian root-finding problem is combined with a quaternion tracking controller to converge on a surface deflection solution that resembles the optimal solution. Additionally, the differences between the point-mass and 6DOF optimization problems are highlighted to express the advantages provided by both solutions.
Abstract ID: 50DCASS-009
Adaptive Online Optimization-Based Control for Soft-Landing Maneuvers on Asteroids: Performance and Safety Assurance Despite Gravity Field Uncertainty
Felipe Arenas Uribe
University of Kentucky
T. Michael Seigler
University of Kentucky
Jesse B. Hoagg
University of Kentucky
The exploration of small celestial bodies, such as asteroids and comets, has gained significant attention due to the scientific insights these missions offer. While recent missions like NASA's OSIRIS-REx and JAXA's HAYABUSA have showcased the potential of small-body exploration, the challenge of powered descent and soft-landing on asteroids remains unresolved. This capability is critical for enabling advanced missions that require surface interaction. Soft-landing maneuvers are complicated by the uncertain and irregular gravity fields of asteroids, resulting from their irregular shapes, nonuniform density distributions, and high synodic periods. Traditional guidance and control approaches, which rely on open-loop execution of pre-computed trajectories, lack the robustness and real-time adaptability necessary for these environments. While closed-loop tracking of these trajectories neglect state and actuator constraints, such as glideslope conditions for collision avoidance and actuator limits. This work introduces an adaptive, online optimization-based control framework to address the challenges of asteroid soft-landing maneuvers. The proposed architecture integrates three key components to ensure high performance and safety despite uncertainties in gravity models and asteroid physical parameters. First, an extended high-gain observer is developed to estimate external disturbances in real-time, enabling precise compensation for unmodeled dynamics. Second, a feedback-linearizing control law is designed to track a reference trajectory generated online, while compensating for disturbances using the observer's estimates, which generates a desired control input. Third, a safety filter ensures forward invariance within a safe set defined by a composition of control barrier functions given by state and input constraints, using a closed-form safe control input which minimizes the intervention from the desired control input. Simulation results demonstrate the proposed framework's ability to achieve reliable soft-landings on irregular asteroids, even under scenarios with significant gravity field and mass uncertainties. By combining adaptive feedback control with an optimization-based safety filter, this approach addresses limitations in traditional guidance and control architectures, paving the way for fully autonomous asteroid landing missions.
Abstract ID: 50DCASS-028
Aerospace Vehicle Trajectory Analysis Using Stability & Controllability Bracketed Analysis
Rhonin Edwards
Air Force Institute of Technology
Will Lorenzo
Air Force Institute of Technology
Jose Camberos
Air Force Institute of Technology
Aerospace vehicle conceptual design requires trajectory analyses to
determine flight characteristics. Cutting-edge air-breathing vehicles
designed for experimental flight testing, and understanding system
aerodynamic performance is crucial for advancing aerospace technology.
This research project will conduct a trajectory analysis of a given
Computational Fluid Dynamics (CFD) database utilizing the TS-COBRA
(Tool for Stability & Controllability Bracketed Analysis) simulation
environment. Through the use of TS-COBRA one can generate skymaps that
visually represent the system’s performance at specific mach numbers and
altitudes. This representation allows further consolidation of data with
greater insight into vehicle performance and limitations. This work will
conduct a comparative study of the vehicle’s performance by leveraging the
extensive CFD Database with data generated by lower fidelity panel methods.
This comparative study will verify whether the panel methods have sufficient
accuracy for modeling aerospace vehicle flight performance.
Abstract ID: 50DCASS-029
Aircraft Design, Analysis, Performance, and Trade Space (ADAPT) Plug-In for Stability & Controllability
Luis Javier Silva-gomez
Air Force Institute of Technology
Will Lorenzo
Air Force Institute of Technology
Samuel Atchison
Air Force Institute of Technology
Jose Camberos
Air Force Institute of Technology
The Aircraft, Design, Analysis, Performance, and Trade Space (ADAPT) and Tool for Stability and Controllability to Bracketed Requirements Analysis (TS-Cobra) programs allow for analyzing aircraft performance, along with stability and control characteristics in the conceptual design phase of the hypersonic regions. This project will focus on developing and implementing the ADAPT plug-in for TS-Cobra. The workflow begins by generating a hypersonic vehicle in Engineering Sketch-Pad (ESP) and then incorporating it into the ADAPT plug-in. This capability will allow us to assess the stability and control characteristics through the ADAPT workflow. A comparative analysis of these results will identify discrepancies and provide guidance for revision to correct the predictive capability of the ADAPT plug-in. Distribution Statement A: Approved for Public Release; Distribution Unlimited. Public Release #88ABW-2025-0051
Abstract ID: 50DCASS-036
A Bio-Inspired Fuzzy-Driven Control Approach for Real-Time Autonomous Spacecraft Inspection
Daegyun Choi
University of Cincinnati
Donghoon Kim
University of Cincinnati
In-space servicing missions aimed at extending spacecraft mission life have received significant attention, with autonomous spacecraft inspection playing a critical supporting role. While traditional optimal control and learning-based approaches have been explored, they often face limitations in real-time implementation and interpretability. To address these challenges, this study proposes a control strategy for deputy trajectory planning that integrates a fuzzy inference system with a bio-inspired optimization technique. The approach seeks to minimize cumulative delta-v and maximize inspection efficiency while adhering to mission constraints, such as maintaining admissible relative distances, accommodating optical sensor field of view, and restricting thrust force and torque capability. The fuzzy-based controller is trained across diverse scenarios to ensure robustness and reliability. Numerical simulations validate the proposed method, demonstrating that the deputy achieves near-optimal trajectories while satisfying mission constraints, offering a practical and interpretable solution for autonomous spacecraft inspection.
Abstract ID: 50DCASS-063
X-15 High Fidelity Stability and Control Study
Hannes Rogers
Air Force Institute of Technology
Jose Camberos
Air Force Institute of Technology
The interest in aerospace vehicle development continues to be a main focus of the Department of the Air Force. Leveraging progress from legacy programs will be vital in supporting future aerospace system development, which will rely heavily on the use of digital design tools. Modeling and simulation requires sufficiently accurate tools to deliver results with confidence. This study will compare high-fidelity computational aerodynamic analyses against existing low fidelity models created using historic X-15 flight data. The computational fluid dynamic software (CFD), Kestrel, will be used to perform the simulated analyses across a Mach range of 3-6. The low-fidelity calculations are available from the following sources; CBAero (hypersonic aerodynamics), TSCobra (flight dynamics), and VORLAX (subsonic aerodynamics). The Kestrel data will be used to generate "Skymaps" that will help display the pitch, roll and yaw moments vs. Mach number to help define the flight envelope of the X-15 and will be compared to those generated from the low-fidelity models. (Case Number: 88ABW-2025-0049)
Fluid Dynamics
Abstract ID: 50DCASS-001
Numerical Simulation of Circulation Control For a Wing Section Using a Coanda Jet
Donald Rizzetta
Air Force Research Laboratory
Daniel Garmann
Air Force Research Laboratory
Large-eddy simulations were carried out to describe the subsonic flow over a wing section, using Coanda-jet circulation control to augment lift. The section geometry consists of a modified supercritical airfoil, with a jet that is blown over a one quarter circular cylindrical trailing-edge Coanda surface. Because the configuration does not include a slotted trailing-edge flap, the mechanical complexity and weight may be reduced. The computations correspond to an experimental investigation, that provides data for comparison. High-fidelity solutions were obtained to the unsteady three-dimensional Navier-Stokes equations, at a chord-based Reynolds of 475,000 and freestream Mach number of 0.1. Several angles of attack were considered, and solutions were obtained both with and without circulation control. In the control cases, three different jet-mass flow rates were simulated at each angle of attack. A grid-resolution study was carried out to assure numerical accuracy. Comparisons are made between the respective cases and with the experimental data, and features of the flowfields are characterized. It was found that Coanda-jet control was able to augment lift in excess of factors of three with reasonable amounts of jet-mass flow.
Abstract ID: 50DCASS-053
Characterizing the Dynamics of a Spanwise Homogenous Shock Train using Large Eddy Simulation
Jack Sullivan
The Ohio State University
Datta V. Gaitonde
The Ohio State University
Turbulence phenomena and unsteady physics occurring in a spanwise homogeneous shock train are investigated using wall resolved large eddy simulations. Shocks waves within the train impart strong spatial pressure gradients on the underlying wall boundary layer, resulting in each individual shock wave/turbulent boundary layer interaction (STBLI) in the system modifying the dynamics, structure, and transport mechanisms of the local turbulent flow. These variations are quantified using a host of diagnostics techniques, including Reynolds stress amplification analyses, stress anisotropy maps, turbulent kinetic energy budgets, and descriptions of scale growth and evolution. Statistical descriptions of the turbulent flow indicate that classical inner layer mechanisms, which dominate the upstream undisturbed boundary layer, are interrupted by strong shock induced separations. Downstream of the observed time mean separation region, the wall bounded flow adopts a mixed boundary layer/shear layer type state, where prominent outer layer processes emerge. At sufficiently downstream distances, where local Mach numbers have decreased towards unity and shock waves become significantly weaker, evidence for the re-emergence of canonical inner layer behaviors is noted.
In addition to altering the local turbulent flow, the STBLIs in the shock train display unsteadiness across a wide spectrum of spatiotemporal scales. Characterization of these dynamics is undertaken through analysis of wall pressure fluctuations and shock oscillations, with particular emphasis devoted to unveiling the frequency dependent behavior displayed by the system. Initial results indicate the entire shock train possesses a low frequency streamwise vibration, with the leading separation shock initiating the coherent motion and downstream shocks responding after a finite time delay. In addition to this prominent rigid body reaction, energetic mid band frequencies are observed in the wall pressure signals for all locations downstream of separation, with the 2-point correlations identifying these as footprints of large outer layer eddies and coherent convective structures within the turbulent boundary layer. Correlation between the perturbation of downstream shock waves and these larger boundary layer features is observed, creating a minor mode of shock oscillation in response to repeated interaction with outer layer structures.
Abstract ID: 50DCASS-075
Experimental Investigation of PRANDTL-D3C Near-Wake
Julian Pabon
University of Dayton
Sidaard Gunasekaran
University of Dayton
Grace Shreyer
University of Dayton
Michael Mongin
Air Force Research Laboratory
Aaron Altman
Air Force Research Laboratory
The disappearance of the tip vortex in the near wake of a wing challenges conventional
aerodynamic theories and presents new opportunities for drag reduction. Recent
computational studies suggested that the PRANDTL-D3C wing
Abstract ID: 50DCASS-091
Experimental Setup for a Closed-Loop Controlled Active Flow Control Jet
Julian Pabon
University of Dayton
Sidaard Gunasekaran
University of Dayton
Michael Mongin
Air Force Research Laboratory
Active flow control (AFC) devices can be utilized for stall mitigation, enhancing aerodynamic performance, and improving maneuverability by altering the localized flow around different regions of a vehicle. A blown jet was installed on a finite wing along the upper surface of a flat plate in the University of Dayton Water Tunnel (UDWT). The mass flow rate and the corresponding momentum coefficient were controlled using a closed loop PID controller to achieve a desired mass flow rate through the actuator. The resultant jet profile was characterized using Particle Image Velocimetry (PIV).
Abstract ID: 50DCASS-092
Characterization of Agricultural Nozzle Spray Geometry in Propeller Wake using Spray Patternator
Brock Greenwood
University of Dayton
Abdul Khan
University of Dayton
Sidaard Gunasekaran
University of Dayton
What happens when the precision of agricultural spraying collides with the turbulence of a
propeller wake? This study answers that question by investigating the impact of propeller wake
on the spray geometry of standard ASABE agricultural nozzles. A 3D spray patternator was
designed, built, and integrated into the University of Dayton Spray Test Section to capture
spray patterns both with and without the influence of a propeller. The nozzles were positioned
directly beneath the hub of the propeller, and spray patterns were captured from the nozzle
tip to the bottom of the spray at 1.4-inch intervals. These images were used to reconstruct and
quantify the spray geometry, including spray width, depth, and droplet distribution. The
results reveal that finer nozzles, such as the 11001 and 11003, experienced up to a 100%
increase in spray width and a 50% reduction in depth, with significant deformation of the flat
fan geometry. Coarser nozzles, like the 8008, showed minimal changes. Quantification of
droplet dynamics indicated that finer nozzles, with higher drag coefficients, are more
vulnerable to propeller-induced turbulence.
Abstract ID: 50DCASS-093
Wake Characteristics of Non-Elliptical Lift Distribution
Charles Cain
University of Dayton
Sidaard Gunasekaran
University of Dayton
The correlation between the wake characteristics and the lift distribution of wings with non-elliptical lift distributions was examined using analytical and numerical methods, which were then compared to a panel-method simulation (FlightStream). Initial results show that, for a non-elliptical lift distribution in an inviscid flow, the wingtip vortex location can be moved inboard and flattened, such that it may interact with a viscous shear layer in ways that a typical wingtip vortex would not. The downstream location where the vortex appears changes with the taper ratio. This is part of a continued study to understand the underlying physical mechanism behind a nearly two-dimensional wake that has been observed in particular non-elliptical lift distributions, such as Prandtl’s bell-shaped lift distribution. The goal of this research is to characterize the causes of the nearly two-dimensional wake and its corresponding reduction in wingtip vortex strength. This is preliminary work that will support a theoretical understanding of the flow in preparation for wind-tunnel testing.
Additionally, in order to better understand the linkage between lift distribution and wake characteristics, an analytical study was conducted to examine the mathematical correlations between lift distribution and resultant upwash. These results show that a lift distribution can be determined from a given downwash velocity profile, and vice versa, suggesting a potential analytical basis for wake vortex control.
Abstract ID: 50DCASS-094
Effects of a Pulsed Jet in Crossflow on Flow Characteristics
Laura Senkar
University of Cincinnati
Daniel Cuppoletti
University of Cincinnati
The development and understanding of flow control techniques is essential to improving the performance of aerodynamic and propulsion systems, and can be utilized in a variety of applications. The interaction of a jet in cross flow is a well-studied phenomenon, and it is known that a steady jet in cross flow will generate pressure differentials and vortical structures, and that flow characteristics vary with jet geometry. This experimental study investigates the effects of a pulsed jet in cross flow, an active flow control technique, on those flow characteristics as well as momentum flux. A series of configurations were tested, varying jet skew angle and pulse frequency. Experiments were conducted in a low-speed wind tunnel, and a range of injection frequencies and jet configurations were tested. Air was injected into the crossflow, flush with the injection wall, using a solenoid-controlled fluidic injector to pulse the jet at a 50% duty cycle for each of the various pulse frequencies tested. Multiple jet geometries were tested, having various injection skew angles. Pressure measurements were taken using Kulite pressure transducers set in an airfoil-shaped rake. The flow field was mapped at multiple jet to diameter ratio (x/D) locations, and these mapped planes are used to analyze jet penetration and growth/decay of the area of influence. These pressure measurements provide an understanding of the flow structures generated, and results show trends in the flow control method effectiveness. Future work includes the use of IR imaging techniques to visualize the flow for the same and additional jet configurations. This method of measurement is nonintrusive, unlike the current rake, which will allow for a better understanding and mapping of the flow field while employing a simpler and more efficient experimental method.
Heat Transfer & Thermal Management
Abstract ID: 50DCASS-005
Investigation of the Operating Mechanism of the Ranque-Hilsch Vortex Tube
Matthew N. Fuqua
Air Force Life Cycle Management Center
James Rutledge
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-008
Adiabatic and Overall Effectiveness Superposition Theory for Upstream Phantom Cooling on a Film Cooled Leading Edge
Nathaniel Stout
Air Force Institute of Technology
James Rutledge
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-026
A Nitinol-Actuated Passive Dynamic Radiator for Spacecraft Electronic Component Cooling
Anthony Lococo
University of Dayton
Dr. Rydge Mulford
University of Dayton
Abigail Boyer
University of Dayton
Ashley Anderson
University of Dayton
Andrew Gabriel
University of Dayton
This project aims to develop and optimize a dynamic radiator fin to cool any electronic device in a space environment. Historically, these devices have been subject to temperature limits, as they cannot become too hot or too cold within their duty cycle. This duty cycle corresponds to the orbit a given spacecraft takes, as the device needs to either stagnate or perform functions along certain arcs during its orbit. Traditionally, managing this has been done through creating a conduction pathway to space directly from the electronics, or through constructing static radiators which protrude into space. Both of these issues are problematic, as they cannot adapt to the variable heat loads which the electronics induce. The solution is a dynamic radiator, which is able to retreat inside the spacecraft and protrude outward according to the demands of the system. When inside, it will collect the heat of the system, and when outside, it will release it into space. Additionally, having the radiator inside the system will increase the steady state cold temperature of the electronic component. In summary, a dynamic radiator prevents the device from getting too cold when inactive, and too hot when active. This technology can be coupled with the usage of a nitinol phase change material (PCM). This enhancement of thermal management is obtained through the use of nitinol
Abstract ID: 50DCASS-043
Fuel Pump Power and Thermal Conceptual Design Investigation for a High Speed Vehicle
Jacob Jadischke
AFIT Contractor
Jose Camberos
Air Force Institute of Technology
Ramana Grandhi
Air Force Institute of Technology
Danielle Hollon
Air Force Research Laboratory
Mitch Wolff
Wright State University
The design of a high-speed vehicle presents new challenges when compared to a lower speeds.
At high speeds, there is no turbine or other rotating component within the propulsion system
to generate electrical power, and or drive the fuel pump, necessitating an alternative means
to turn the device to sustain thrust. As it is desirable to consider the power generation system
earlier in the conceptual design process of high-speed vehicles, a means by which the power
requirements of a generic geometry can be acquired quickly must be obtained, which for high
speeds includes the power requirements of the fuel pump. In prior work, a conceptual design
level 6DOF simulation was created to model the power and thermal requirements of high-speed
vehicle subsystems. This work expands the previous work by implementing a fuel pump
driven by an electric motor. The pump; modeled as a variable displacement pump, and motor
are modeled and controlled in the developed SIMULINK model. A power and thermal analysis
was performed for the entire subsystem and the profiles for each were extracted from the
SIMULNK model. Upon simulation, both the power and thermal profiles for the fuel pump and
motor subsystem were available for analysis at the conceptual design level.
Distribution Statement A: Approved For Public Release. Distribution is Unlimited. PR-4596 AFRL-2024-6516
Abstract ID: 50DCASS-065
Predictive Aeroheating for Conceptual Vehicle Design and Analysis
Amy Cinnamon
Air Force Institute of Technology
Jose Camberos
Air Force Institute of Technology
Conceptual design of hypersonic vehicles is completed early in the design cycle. The
prediction of aero-thermodynamic heating is a significant factor in design decisions at this stage.
Configuration-Based Aerodynamics (CBAero) is an engineering-level software tool used to
predict the conceptual aero-thermodynamic environments of aerospace configurations. CBAero
tends to overpredict heating at hypersonic flight conditions by up to 20 percent or more, tending
to result in vehicles which are over-designed. The accuracy of CBAero’s heating predictions may
be improved if the heat transfer correlations used are modified to incorporate an air chemistry
model based on experimental data supplemented by first-principles air chemistry rates. The
results from the low-fidelity CBAero may be compared against a high-fidelity tool such as Kestrel
to demonstrate a reduced margin of error. (Public Release no. 88ABW-2025-0043)
Abstract ID: 50DCASS-077
CFD Analysis of a TPMS-Based Conformal Heat Exchanger
Nathan Lewan
Wright State University
Mitch Wolff
Wright State University
Advancements in aircraft systems and supporting equipment have resulted in high thermal loads within new generations of military aircraft without allocating the space requirements that would traditionally be needed for additional cooling. Therefore, the necessity arises of creating more efficient thermal management systems to increase the cooling capabilities of the aircraft. Heat exchangers play a critical role in these systems and operate by transferring energy from a higher potential medium to a lower one. Conventional heat exchanger design has been governed by manufacturability, but with recent improvements in additive manufacturing, conformable triply periodic minimal surface heat exchangers (CTPMS-HX) are now feasible. This work presents a computational fluid dynamics (CFD) study over a CTPMS-HX created in nTopology software. The bulk of this process included importing geometry to ANSYS, setting up boundary conditions for varying tests, meshing, and finally the solution. This research is designed to serve as a basis for which future CFD studies can be compared to as well as improve upon exiting methods for this type of analysis. Distribution Statement A: Approved for Public Release; Distribution is Unlimited. PA# AFRL-2025-0067
Abstract ID: 50DCASS-085
Conformal Triply Periodic Minimal Surface Heat Exchanger Design and Evaluation
Tanner Barber
Air Force Institute of Technology
John Brewer
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-098
Implementation of Genetic Algorithms to Model the Modes of Heat Transfer in Fibrous Insulation Materials
James Senig
University of Kentucky
John F. Maddox
University of Kentucky
Hypersonic reentry vehicles experience extreme thermal loads due to aerodynamic heating. Thermal protection systems (TPS) are used to protect reentry vehicles from these extreme thermal environments during their descent. Fibrous, porous materials such as Phenolic-Impregnated Carbon Ablator (PICA) are commonly used in constructing TPS for these reentry vehicles. FiberForm is a common substrate for PICA and is composed of a complex network of carbon fibers. There are three primary modes of heat transfer that take place within fibrous, porous materials such as FiberForm: solid conduction, gaseous conduction, and radiation. Solid conduction occurs due to the contact between the individual fibers comprising the fiber matrix and is a function of the bulk material conductivity, fiber-volume fraction, and temperature. Gaseous conduction happens due to the presence of gases within the pores of the fiber matrix. It is a function of temperature, pressure, participating gases, and fiber geometry. Finally, radiation is emitted from the individual fibers and is a function of temperature and material optical properties. The three modes of heat transfer are interconnected and the contributions from each mode cannot be easily separated. In recent work, the effects of gaseous conduction have been mitigated by measuring the thermal resistance under vacuum using the comparative cut-bar method, leaving solid conduction and radiation as the primary contributors to heat transfer. Due to the nonlinear relationship of radiation with temperature, traditional linear methods for determining the heat transfer parameters associated with each mode are not feasible. Nonlinear methods, such as genetic algorithms, can be used to separate the contributions from solid conduction and radiation. This process requires collecting thermal conductivity data over a wide range of temperatures and pressures. Once the heat transfer parameters associated with each mode are determined, each mode of heat transfer can be independently modeled as a function of temperature and pressure. In previous work, genetic algorithms have been used to separate solid conduction and radiation. This work will look at applying genetic algorithms to solve for all three modes of heat transfer in FiberForm, reducing the need to test at medium to high-vacuum conditions.
Abstract ID: 50DCASS-105
Experimental Effect of In-Depth Radiative Heating with Coupled Argon Plasma Flow on Fiborous LI2200
Colby Gore
University of Kentucky
John F. Maddox
University of Kentucky
Savio Poovathingal
University of Kentucky
Awaiting public release.
High School
Abstract ID: 50DCASS-047
Evaluation and Comparison of Large Language Models
Isaiah Goble
Stebbins High School
Kara Combs
Air Force Research Laboratory
Given the rise in popularity of large language models (LLMs), there is a misconception they have developed human-level thinking capability. Analogical reasoning is an example of a higher-level cognitive ability (frequently leveraged by humans), allowing new information to be inferred through analogies without the need for additional training data. Twenty-one LLMs were tested on the Rattermann story analogy dataset, which includes sixteen problem sets consisting of five stories, denoted with letters A through E, with varying degrees of similarity to a sixth
Abstract ID: 50DCASS-069
On the Frequency Assignment of Packet Radio Networks
Grant Zhou
Oakwood High School
A Packet Radio Network (PRN) connects a web of stations and transmits data through packet-switching techniques. It has found applications in satellite radio, mobile communications, and other related areas. In a PRN, two stations can communicate with each other if a channel exists between them. PDNs are prone to secondary interference, which refers to two channels connected by a third channel. Frequencies are assigned to channels to transmit data in packets. Multiple channels may share the same frequency if secondary interference is prevented. One research problem is to find an upper bound for the frequencies required to avoid all secondary interferences.
I studied Packet Radio Networks with the following properties: in any sub-network of the given PRN, each station is connected to, on average, at most 3.5 channels. I proved that any such PRN has a frequency assignment with at most 45 frequencies that prevents secondary interference. My result improved a previous one by Lai and Luo, where they proved that if the average value is less than 2.8, then at most 2?+1 frequencies are needed. Here ? is the largest number of channels a station connects to in the PRN.
Abstract ID: 50DCASS-083
Constructing an Affordable Homemade Wind Tunnel
Audrey Flaute
Anna High School
Many find themselves curious about the field of aerospace engineering. In order to study this subject, it is essential to have a basic understanding of the physics behind flight. The visual and physical representation of it aids in this understanding. Most wind tunnels are large and pricey, and not everyone can access them. Even homemade wind tunnels can be quite costly as well. This project aimed to create an affordable, homemade subsonic wind tunnel that can measure the forces of lift and drag. The wind tunnel in this project was successfully built using scraps of cardboard, wood, and other easily accessible materials and borrowed tools. Force sensors were then attached to the tunnel, which successfully read forces of lift and drag of the airfoil. However, the readings were likely not extremely accurate. This was acceptable for the sake of affordability. The cost of the tunnel amounted to only fifteen dollars. This project can be recreated for a similar price if the materials are chosen carefully. Otherwise, it would still likely cost under one hundred dollars. This project may allow those who want to build their own wind tunnel a simpler guide to doing so while saving money.
Imaging & Diagnostics
Abstract ID: 50DCASS-071
Evolution of H-TALIF Signal After Nanosecond Discharges in Methane-Air Mixtures
Katherine Opacich
National Research Council
Timothy M. Ombrello, Campbell D. Carter
Air Force Research Laboratory
Joseph K. Lefkowitz
Other - please contact webmaster
Matthew K. Hay, Waruna D. Kulatilaka
Texas A&M University
Femtosecond two-photon-absorption laser-induced fluorescence (TALIF) was used to measure the spatial and temporal evolution of atomic hydrogen (H) after single and multiple nanosecond discharges. Experiments were conducted at ambient pressure and temperature in mixtures of 4% and 2% methane by volume in air. Results show that the spatial and temporal distribution of the H-TALIF signal is primarily influenced by the hydrodynamics related to the discharge-induced flow field. However, questions remain regarding the decrease in the H signal observed in the inter-electrode region at earlier times, typically 1~2 µs, following the ~10-ns discharge. In comparing the 4% to the 2% methane-air mixture results, it was found that reducing the fuel concentration lowered the H-TALIF signal at each delay time but did not impact the spatial distribution, an expected result since methane acted as the source of H in all these cases. Finally, the H-TALIF signal was recorded after the second, fifth, and tenth discharge pulse in a burst of discharges operating at a 200 kHz repetition rate. A decrement in signal in the inter-electrode region was observed at early times after each pulse in the sequence. The spatial extent of the H-TALIF signal grew with each pulse in the train, predominantly due to the hydrodynamics related to consecutive nanosecond discharges and discharge-induced jetting motion. Combined with computational models, these results can provide valuable insights to the plasma-assisted combustion community regarding the effectiveness of nanosecond discharges for ignition and flame holding.
Machine Learning
Abstract ID: 50DCASS-011
Bio-Inspired Motion Strategies for Robotic Arm-Equipped Satellites Using Computer Vision and Animal Pose Estimation
Alhim Adonai Vera Gonzalez
University of Cincinnati
Daegyun Choi
University of Cincinnati
Donghoon Kim
University of Cincinnati
Robotic arm-equipped satellites are essential in modern space operations, addressing critical challenges, such as orbital servicing, debris removal, satellite assembly, and in-space repairs. These systems are designed to operate in complex and constrained environments, including microgravity, where traditional operational strategies often face limitations in stability, adaptability, and maneuverability. As space missions become increasingly ambitious, there is a growing need for autonomous robotic arm-mounted satellites that can perform delicate and precise tasks without human intervention. Therefore, reliable and efficient operational strategies for such systems in zero-gravity conditions are essential. This study investigates bio-inspired control strategies for robotic arm-equipped satellites by analyzing the movement patterns of animals, such as lizards and squirrels. The ability of these animals to execute righting reflexes that enable posture stabilization using their bodies, limbs, and tails provides valuable insights for robotic systems operating in constrained or zero-gravity environments due to morphological similarities between these animals and robotic systems.
As a preliminary step, this work primarily focuses on analyzing animal behaviors by examining various models and exploring pose estimation approaches. By analyzing the morphological similarities between animals and robotic arm-mounted satellites using videos and literature, models that similarly describe robotic system motion are identified. To analyze motion patterns, animals’ posture information is extracted from multiple videos using pose estimation tools like DeepLabCut and OpenPose combined with animal datasets. After detecting each body segment, such as the upper and lower bodies, limbs, and tail, the relative angles between each body segment are calculated, forming time-series trajectories that replicate the animals’ motion. Through this analysis, the primary body segments, key motion patterns, and the body segments’ movement ranges are identified and represented in this work. As future work, the findings from the animals’ motion patterns will be leveraged to formulate operational strategies for robotic arm-equipped satellites.
Abstract ID: 50DCASS-014
Multi-Fidelity Machine Learning Modeling for Aerodynamic Response Prediction of Aerospace Vehicles
Ethan Jackman
Air Force Institute of Technology
Jose Camberos
Air Force Institute of Technology
Harok Bae
Wright State University
The design of Aerospace Vehicles requires knowledge of the aerodynamic response across the flight domain. This paper presents a new multi-fidelity machine learning modeling approach to balance the accuracy of Computational Fluid Dynamics with the abundance of low-fidelity models using multi-fidelity surrogate modeling to predict the aerodynamic field responses. The proposed architecture combines an autoencoder for dimensionality reduction and an emulator embedded neural network for leveraging multi-fidelity training data in latent spaces. In the final paper, a blunted cone with fins example will be discussed for the practicality of the proposed approach along with multiple analytical examples.
Distribution A: Approved for Public Release PA Number 2024-1054
Abstract ID: 50DCASS-038
Dreaming Falcon: Advanced Drone Control using Physics-Informed Model-Based Reinforcement Learning
Eashan Vytla
The Ohio State University
Bhavanishankar Kalavakolanu
The Ohio State University
Andrew Perrault
The Ohio State University
Matthew McCrink
The Ohio State University
Traditional control methods such as PID controllers have been widely used for quad-copter control. However, these approaches often struggle to maintain stability and adapt to variable environmental conditions or vehicle dynamics. Additionally, they may not fully exploit the capabilities of modern drone hardware. Therefore, there is a pressing need for innovative control algorithms capable of overcoming these limitations. Deep Reinforcement Learning (DRL) has the potential to enhance the stability and maneuverability of drones, particularly in dynamic environments or adverse weather conditions. However, DRL requires hundreds of hours of training data from simulation or flight tests. To increase the sample efficiency of DRL, the algorithm proposed in this work, named Dreaming Falcon, presents a solution using physics-informed model-based Reinforcement Learning. Our algorithm consists of two major components: a world model to dynamically learn vehicle and environmental conditions using an autoregressive differential network, and a behavior model to find the optimal trajectory based on a defined reward function using Reinforcement Learning. A data buffer is implemented to train the world model in real time using batch processing. This approach can simultaneously learn flight dynamics while also optimizing control actions. Preliminary results demonstrate that the algorithm can learn an accurate world model from synthetic data generated by a simple 3-DOF simulation. Our final report will explore the stability and feasibility of Dreaming Falcon in adverse environmental conditions compared to a baseline PID controller. Our work will allow EVTOLs to learn and adapt to changes in the vehicle and adverse environments.
Abstract ID: 50DCASS-041
Online Identification of Free-Floating Dual-Arm Space Robot Dynamics With Interpretable Physics Informed Machine Learning
Aaron Borger
University of Cincinnati
Donghoon Kim
University of Cincinnati
Recently, there has been a growing interest in active debris removal as well as servicing, assembling, and manufacturing spacecraft in orbit. The desire primarily intends to avoid the Kessler syndrome and extend the life of valuable space assets, reducing the overall cost. Free-flying and free-floating robotic systems with multiple manipulators have been proposed as likely ideal solutions as they provide a generic platform that could perform a large variety of procedures. However, free-flying and free-floating systems have highly non-linear dynamics, meaning classical manipulator control algorithms cannot be used due to the reaction force and torques applied to the base as manipulator joints move. Understanding the dynamics of a system typically helps when developing a control algorithm, so previous methods derive the system dynamics for free-floating dual-arm space robots from the law of conservation of momentum and the Lagrangian formalism. However, several effects have not been accounted for, which may reduce the accuracy of the dynamics and prevent the dynamic formulation from transferring to free-flying space robots. These effects include external forces such as the propulsion mechanism firing, which would change the momentum, mass, and center of mass due to fuel usage or fuel sloshing, flexible structures such as solar panels or long robotic links, thermal expansion, and contraction, which can alter the center of mass or inertial properties, motor dynamics changing due to the thermal environment, and interactions with a client satellite which has been captured and docked and may have uncertain inertial properties and contact dynamics at the docking interface. An interpretable physics-informed machine learning approach is proposed, which allows learning the non-linear system dynamics online while leveraging an initial dynamics formulation that is likely close to the true dynamics and can, therefore, reduce the computation and training data required. (Approved for Public Release; distribution is unlimited. Public Affairs release approval # AFRL-2025-0557)
Abstract ID: 50DCASS-050
Accelerated and Confidence-based Machine Learning with Dimension Reduction via Global Sensitivity Analysis for Conceptual Aerospace Vehicle Design
Harok Bae
Wright State University
Atticus Beachy
Air Force Institute of Technology
Jose A. Camberos
Air Force Institute of Technology
Ramana V. Grandhi
Air Force Institute of Technology
Jonathan Boston
Air Force Research Laboratory
A new machine learning modeling approach is proposed to quantify the prediction confidence of a trained model and reduce the dimensionality of complex modeling problems by identifying important design variables. The first- and second-order global sensitivities of variables are leveraged to enhance the efficiency of an ensemble modeling framework using emulator embedded neural network for high-dimensional design exploration. There are two primary ways of enhancing modeling efficiency: integrating sub-models of important variables as emulators to accelerate model training, and reducing the dimensionality of design space by aggregating unimportant variables into a random source variable. The emulator embedded ensemble model provides prediction bounds that reflect the uncertainty arising from the aggregated variables, fostering informed decision makings. In a preliminary study, the proposed approach demonstrated promising performance, particularly in scenarios where a small subset of variables significantly influences the system behavior.
Abstract ID: 50DCASS-058
Isotopic Analysis of Lithium Hydroxide Monohydrate Using Laser-Induced Breakdown Self-Reversal Isotopic Spectrometry (LIBRIS) and Machine Learning
Madison Moran
Air Force Institute of Technology
Ashwin Rao
Air Force Research Laboratory
Anil Patnaik
Air Force Institute of Technology
Monitoring isotopic abundance levels of 6-Li and 7-Li is essential to nuclear energy infrastructure and the enforcement of nuclear non-proliferation policies. At present, mass-spectrometry and other costly, complex, and time consuming in-lab processes are typically used to determine isotopic abundance in Li compounds. Optical emission spectroscopy techniques coupled with advanced machine learning algorithms provide a pathway to the development of rapid, portable, in-situ Li isotope detection devices. In this study, high-resolution Laser-Induced Breakdown Self-Reversal Isotopic Spectrometry (LIBRIS) is implemented to record the 15.8 pm Li I 670.8 nm isotopic shift in LiOH*H2O samples of varying 6,7-Li abundance. A simple univariate linear regression demonstrates an experimentally acquired isotopic shift of 13.813 plus/minus 1.21 pm with a relative error of 12.57 % in samples varying from 3 to 95 6-Li atom percent. Supervised machine learning regressions are trained on the self-reversal dip wavelength location for each sample in order to quantify their 6-Li abundance. A stacked ensemble model using two supervised regression base learners yields the superlative characterization of isotopic content with RMSE of 5.66 at% and a detection limit of 18.8 at%. Using LIBRIS under atmospheric pressure instead of LIBS under vacuum conditions simplified the experimental parameters to mimic the environmental conditions for in-situ Li analysis. Combining LIBRIS with advanced machine learning models yields better accuracy and sensitivity than traditional chemometric analysis.
Abstract ID: 50DCASS-061
A Geometrically Flexible Machine Learning Approach to Predict Stress Intensity Factors for Computationally Efficient Fatigue Analysis
Christopher Turner
University of Cincinnati
Karen DeMille
Air Force Research Laboratory
The availability of efficient and accurate predictive models for stress intensity factors (SIFs) is crucial for fatigue analysis of aerospace structures. Traditional analytical and numerical analysis methods are often limited by their geometric assumptions and high computational costs, respectively. To enable earlier incorporation of fatigue analysis in design processes, there is a need for more flexible and efficient approaches. Addressing the need for more flexible and efficient approaches, this work aims to investigate the feasibility of using machine learning models to provide a geometrically flexible approach for rapidly predicting SIFs in complex geometries under arbitrary loading conditions. In this investigation, various machine learning models, including multilayer perceptron (MLP) and convolutional neural network (CNN) models, are explored as efficient and accurate tools for SIF prediction. The MLP models are trained to provide a baseline for SIF prediction accuracy using traditional geometric parameterization. The MLP models are given inputs related to geometric and load parameters used in verified, closed-form SIF solutions. However, the requirement for geometric parameters limits the MLP models to predicting SIFs for well-defined, parameterized geometries. To remove this limitation, the CNN models are used to evaluate the effectiveness of predicting SIFs without geometric parameterization. The CNN models are given virtual images of cracked geometries as input, thus removing the requirement of parameterized geometries associated with the MLP models. The geometrically flexible CNN can produce SIF predictions in a fraction of the time required for numerical simulations while removing geometric parameters required for analytical solutions. The MLP and CNN models are evaluated on their prediction of SIFs obtained from both closed-form SIF solutions and finite element analysis (FEA). Assessments of the MLP and CNN based on prediction accuracy, geometric flexibility, and efficiency are underway. The models being explored in this work have the potential to accelerate the evaluation of SIFs, enabling more computationally efficient fatigue analysis to assist in early stages of aerospace structural design.
Distribution Statement A: Approved for Public Release; Distribution is Unlimited. PA# AFRL-2025-0365
Abstract ID: 50DCASS-068
Testing the Cognitive Predictive Performance Equation Model on a Duolingo Dataset
Trevor Cross
Air Force Institute of Technology
Aihua Wood
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-074
Using Unsupervised Machine Learning to Detect Flow Separation on Low-Pressure Turbine Blades
Aaron Suter
Air Force Research Laboratory
Christopher Marks
Air Force Research Laboratory
Jared Kerestes
Air Force Research Laboratory
Mitch Wolff
Wright State University
Flow separation in the low-pressure turbine section of gas turbine engines reduces
efficiency significantly. Vortex generator jets (VGJs) have been shown to prevent separation on
turbine blade surfaces, and their efficiency can be improved by activating them only when
separation occurs. Unsupervised machine learning makes this possible by analyzing pressure
data along the blade surface and categorizing the flow as separated or unseparated, enabling
conditional jet activation. In this study, low-pressure turbine blades equipped with pressure taps
and VGJs will be tested in a low-speed linear cascade wind tunnel. A machine learning model
will be trained to categorize the flow with surface pressure data for Reynolds numbers between
25k and 100k. Total pressure loss measurements across the midspan flow will quantify VGJ
effectiveness. The binary nature of the boundary layer flow over the blades suggest that
clustering methods will accurately identify separation. This proof-of-concept study will
demonstrate the potential for machine learning to enhance active flow control, which can then be
applied to more complex flow conditions. Distribution Statement A: Approved for Public Release; Distribution is Unlimited. PA# AFRL-2025-0327.
Abstract ID: 50DCASS-095
Enhancing RANS Turbulence Predictions for 3D Automotive Flows Using the Ensemble Kalman Filter and Tensor Basis Neural Network
Tarun Kakarlamudi
The Ohio State University
Dr. Lian Duan
The Ohio State University
Reynolds-Averaged Navier-Stokes (RANS) solvers remain the industry standard for Computational Fluid Dynamics in automotive external aerodynamics due to their computational efficiency. However, their accuracy in predicting turbulent flows, particularly in wake regions and flow separation, is limited by inherent model-form uncertainties in Reynolds stress closure. This research investigates an innovative approach to improving RANS-based turbulence modeling through the Ensemble Kalman Filter (EnKF) coupled with a Tensor Basis Neural Network (TBNN), with a specific focus on its application to three-dimensional turbulent flow over the Ahmed Body geometry.
The EnKF framework is employed to iteratively update turbulence parameters based on sparse observational data, refining Reynolds stress predictions without requiring full-field stress data. Sparse observations were extracted from an Improved Delayed Detached Eddy Simulation (IDDES) of the Ahmed Body, with velocity and pressure data used as reference measurements for the EnKF updates. The velocity profiles were strategically sampled in the wake and slant regions, with a few profiles along the centerline and additional profiles normal to it, ensuring robust constraints on the flow structure. Additionally, pressure data was extracted from the slant, base, and nose regions to capture key aerodynamic features influencing drag and wake behavior.
The integration of EnKF with a TBNN enables the model to capture anisotropic turbulence effects more effectively, resulting in improved flow predictions. Initial results indicate a 75% improvement in RANS predictions on a coarse computational mesh, demonstrating enhanced agreement with high-fidelity IDDES results in wake velocity distribution and pressure recovery. By offering a non-intrusive and solver-compatible correction mechanism, the EnKF-TBNN framework presents a viable pathway for improving turbulence modeling in the automotive industry without additional computational costs comparable to Large Eddy Simulations (LES) or Direct Numerical Simulations (DNS). Future work will focus on validating this approach against experimental data and extending its application to a wider range of automotive flow configurations.
Abstract ID: 50DCASS-090
Satellite Simulation and Association Utilizing Machine Learning for Congested Space
Mathew Core
Air Force Institute of Technology
Lt Col Rachel M. Derbis
Air Force Institute of Technology
Maj Timothy I. Machin
Air Force Institute of Technology
David W. Meyer
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-100
Hyperparameter Optimization for Training Multi-fidelity Robust Physics Informed Neural Networks
Mark Bergman
Wright State University
Harok Bae
Wright State University
This paper presents a practical framework for optimizing deep neural network architectures for aerospace vehicle design using multi-fidelity datasets. Addressing challenges of limited data, modeling uncertainty, and system nonlinearity, the proposed approach integrates Physics-Informed Neural Networks (PINNs) to enforce physical constraints and enhance predictive stability. Bayesian Optimization HyperBand (BOHB) is employed to refine hyperparameter selection, balancing training cost with model accuracy while leveraging multi-fidelity data. The methodology is demonstrated on high-dimensional, highly nonlinear analytical datasets to evaluate performance and robustness. By developing this proposed framework, we aim to develop performance metrics for BOHB, streamline architecture selection processes, and provide a robust NN architecture for a given situation considering modeling complexity, uncertainty in data, and limited training resources.
Materials & Structures
Abstract ID: 50DCASS-012
Mechanical Response of Triply Periodic Minimal Surface Gyroid Structures Under Combined Loading
Capt Jay B. Patel
Air Force Institute of Technology
Maj John S. Brewer
Air Force Institute of Technology
Elizabeth K. Bartlett
Air Force Research Laboratory
Additive manufacturing has opened up a vast array of structural designs not previously achievable via traditional machining. One such example is triply periodic minimal surface (TPMS) structures. These are lattice-based designs with potential aerospace applications such as heat exchangers and optimized structural components. TPMS structures can be composed of different cell types, but the focus of this investigation is the gyroid unit cell. Regarding mechanical properties, TPMS structures exhibit high torsional resistance when compared to vertically-inclined structures [1]. These properties are important as this work seeks to characterize TPMS structures in combined tension-torsion loading. Previous research on combined loading of TPMS structures utilized compressive loading and these structures are routinely generated on rectangular cell maps [2]. In contrast, this research will aim to explore combined tensile and torsional loads applied to additively manufactured Inconel 718 specimens employing a gyroid unit cell with two variations of cylindrical cell maps, as well as a rectangular and spherical cell map. These four variations in lattice structure are tested in an axial-torsion test frame using equal parts of axial and angular displacement control until failure. The data from the tests is compared to finite element analysis models to understand when yielding begins. Finally, the fracture surfaces are investigated with optical and scanning electron microscopes to characterize the failure modes.
Abstract ID: 50DCASS-018
Design of a Force Transmission System Within a Real-Time Apogee Control System on a High-Powered Model Rocket.
Seth Mitchell
Cedarville University
Joseph Copeland, Daniel Hogsed Keneth Lee III, Jackson Maley, Grant Parker, Elisa Schmitt, Joseph Miller, and Thomas Ward
Cedarville University
As part of the Cedarville University Mechanical Engineering Senior Design Capstone Course, a high-powered model rocket was designed for the NASA Artemis Student Launch Challenge. For the first time, airbrakes are integrated as a secondary experimental payload to dynamically alter the altitude of the rocket during flight using real-time data. The airbrakes consist of flaps actuated using a crank slider mechanism with a near-constant angular velocity ratio. The mechanical force transmission system was analyzed rigorously using computational (Computational Fluid Dynamics and Finite Element Analysis), analytical, and experimental techniques to determine both internal and external forces on the system. A lockup friction study was conducted for a Polyethylene Terephthalate Glycol collar on a carbon fiber rod to determine the coefficient of friction of this combination of materials. The controls system uses a probabilistic state space model, and the controller runs a Raspberry Pi Pico to control apogee. Full-scale tests are planned to validate the integrated system performance before final rocket delivery.
Abstract ID: 50DCASS-020
Finite Element Modeling of a Taylor Test to Determine Material Properties
Armando Deleon
Wright State University
Anthony N. Palazotto
Wright State University
The Taylor test involves the dynamic deformation of a cylindrical sample that hits at a preset velocity a rigid target. The impact parameters and the degree of sample deformation, determined by geometric measurements, are a measure of the dynamic properties of the examined material. As part of this project, a finite model was developed in ABAQUS. The Johnson-Cook plasticity formulation was utilized to model the material behavior of the specimen. The Johnson-Cook model is a particular type of Mises plasticity model with analytical forms of the hardening law and rate dependence which is suitable for high-strain-rate deformation of many materials, including most metals, it is typically used in adiabatic transient dynamic simulations. Once a convergence study was performed, the model was compared to testing results to include length and diameter of the deformation region. The results were then analyzed to test Taylor
Abstract ID: 50DCASS-023
Experimental Characterization of Bearing Fatigue of Fiber Metal Laminate Without Film Adhesive
Matthew Fadden
Air Force Institute of Technology
John Brewer
Air Force Institute of Technology
Michael Gran
Air Force Research Laboratory
Awaiting public release.
Abstract ID: 50DCASS-031
Evaluating Wave Propagation using Johnson-Cook Equation
Katie Bruggeman
Wright State University
Anthony Palazotto
Air Force Institute of Technology
Dan Young
Wright State University
In recent work, JC parameters of wrought Inconel718 (IN-718) specimen are analyzed after they have been subject to Taylor Impact testing at various pressures via gas-gun. The Johnson-Cook (JC) equation characterizes the response of a metal which is subject to a high strain rate and a viscoplastic regime.
The JC flow equation consists of three parts which describe the relationship between strain rate and temperature. The first term within the equation characterizes the strain hardening effect of the material. The second term characterizes the strain rate strengthening effect through the use of a natural logarithmic function. The third term describes the material while subject to varying temperatures, which are further defined subsequently.
The strain rate ratio is a relation of variable strain rates and a reference strain rate. The variable temperature of the material being considered is displayed as T, Tm is the melting temperature, and Tref is a reference temperature which is normally considered as room temperature. All of the temperature variables listed are utilized to calculate the overall temperature effect, T*.
It is found that the JC parameters utilized for analysis successfully defines IN718 subject to a high strain rate environment and a viscoplastic regime. The determination of JC parameters for additively manufactured IN-718 along with a comparison to the JC parameters described for wrought IN-718 will be part of ongoing work.
Abstract ID: 50DCASS-045
Experimental and Simulation Analysis of the Mechanical Properties of 3D Printed Octet Lattice Structures
Lucas Stevenson
University of Kentucky
Xingsheng Sun
University of Kentucky
Lattice structures offer a lightweight solution for applications where strength to weight ratio is crucial. One such application is the design of lattices that can provide effective protection against impact loads during spacecraft landings. NASA has completed research in the development and testing of lattices, but there is still a critical knowledge gap concerning the optimization of these structures and the effects of manufacturing defects on their behavior when loaded. A series of experiments were conducted on 3D printed PETG octet lattice structures with variance in unit cell density, loading orientation, number of fixed size unit cells, and aspect ratio for both compression and 3-point bend tests. The purpose of these experiments is to characterize the mechanical properties of 3D printed octet lattices and find optimal design characteristics to implement under various loading conditions. The data gained from the experiments is compared against data from simulation software utilizing limit analysis and the finite element method with a periodic boundary condition. The experimental data is also used to calibrate the software to produce accurate results allowing the software to be used to characterize the effects of manufacturing defects on the mechanical performance of the lattices. The experiments yielded several results including that the max load and corresponding strain have negative correlations with increased unit cell density in both compression and bending, the optimal loading orientation is inline with the z-axis of the structure, and the max load decreases exponentially with increasing aspect ratio. These results can be combined to create an optimized lattice based on loading conditions along with weight and size constraints. They also consider imperfections created during the 3D printing manufacturing process, making the lattice
Abstract ID: 50DCASS-048
Selective Laser Sintering Based Additive Manufacturing of Nylon 12 Novel Hybrid Lattice Structures for Improved Compressive Strength
Ahkar Min Thant
Miami University
Jianfeng Ma
St. Louis University
Muhammad P. Jahan
Miami University
Lattice structures (LS) have gained increasing interest among additive manufacturing researchers for their impressive compressive properties and energy absorption capabilities, reduced weight and high specific strength. Additive manufacturing allows creating lightweight structures for various applications including aerospace structures with the help of innovative designs and testing of lattice structures. In this context, this study explores design, printing and testing of novel hybrid lattice structures by combining two or more common lattice structures with known capacity to improve upon their individual compressive strength and energy absorption properties. In this study, four commonly known LS, such as, Auxetic (Aux), Body-Centered-Cube (BCC), Face-Centered-Cube (FCC), and Simple-Cube (SC) were used as the basic structures, whereas, three hybrid structures were designed by combining these structures
Abstract ID: 50DCASS-060
Evaluation of Additive Manufactured GRX-810 Alloys in the AFIT Burner Rig Facility
Marlee Ruiz
Air Force Institute of Technology
Marc Polanka
Air Force Institute of Technology
Matthew R. Gazella
Air Force Institute of Technology
Ryan A. Kemnitz
Air Force Institute of Technology
NASA recently developed a new oxide dispersion strengthened alloy, GRX-810, that survived nearly 2,000 times longer than traditional nickel-based superalloys under stress at 1093
Abstract ID: 50DCASS-076
Fatigue Behavior of two Advanced C/SiC Composites at 1200 C
Conner Adams
Air Force Institute of Technology
M. B. Ruggles-Wrenn
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-084
Fatigue Behavior af an Advanced SiC/SiC Composite with an Array of Small Cooling Holes at 1200 C
Tess Panchyshyn
Air Force Institute of Technology
M. B. Ruggles-Wrenn
Air Force Institute of Technology
Fatigue behavior of an advanced silicon carbide/silicon carbide (SiC/SiC) composite with an array of cooling holes was evaluated at 1200 C in air. The composite consisted of a SiC matrix reinforced with laminated SiC (Hi-Nicalon Type-S) fibers woven in a 5-harness satin weave. The Hi-Nicalon-S fiber preforms were coated with boron nitride (BN) fiber coating by chemical vapor infiltration (CVI). The melt-infiltrated (MI) matrix was consolidated by combining CVI-SiC with SiC particulate slurry and molten silicon infiltration. Some test specimens contained an array of 17 effusion holes with 0.5-mm diameter. The holes were drilled normal to surface of the specimens using electric discharge machining (EDM). Specimens without holes or notches were also tested. Tensile stress-strain behavior of the specimens without holes or notches and of the specimens with holes was investigated, and the tensile properties measured at 1200 C. Tension-tension fatigue behavior of specimens with and without holes was studied in cyclic tests at 1200 C with a minimum to maximum stress ratio of 0.1 and frequency of 1 Hz. Fatigue stresses ranged from 124 to 203 MPa. Fatigue run-out was defined as 200,000 cycles. All specimens that achieved run-out were then tested in tension to failure at 1200 C in order to determine their retained tensile properties. The presence of EDM holes reduced fatigue lifetimes and significantly altered the shape of the stress vs. cycles (S-N) curve. Composite microstructure, as well as damage and failure mechanisms were investigated using scanning electron microscopy and energy dispersive spectroscopy.
Abstract ID: 50DCASS-086
Elastic Wave Propagation in Bio-inspired Structures
Arun Nair
Air Force Institute of Technology
Joshua DeMorris
Air Force Institute of Technology
Carson Willey
Air Force Research Laboratory
Vincent Chen
Air Force Research Laboratory
Abigail Juhl
Air Force Research Laboratory
Awaiting public release.
Abstract ID: 50DCASS-087
Mechanical Properties of Low Dimensional Materials Based Nanocomposites
Arun Nair
Air Force Institute of Technology
Abigail Eaton
AFIT Contractor
Awaiting public release.
Abstract ID: 50DCASS-088
Deformation of Composite Plates Under High Strain Rates
Arun Nair
Air Force Institute of Technology
Jackson King
Air Force Institute of Technology
Wyatt Hodson
Air Force Institute of Technology
John Hansen
Air Force Institute of Technology
Awaiting public release.
Model-Based Systems Engineering
Abstract ID: 50DCASS-073
Universal Model Based Systems Engineering Rendezvous and Proximity Operations Mission Planner
Thomas Kelly
Air Force Institute of Technology
Curtis, David H.
Air Force Institute of Technology
Stern, Jordan L.
Air Force Institute of Technology
This research develops an integrated Model-Based Systems Engineering (MBSE) framework to address inefficiencies and fragmentation in Rendezvous and Proximity Operations (RPO) mission planning. The research combines the System Modeling Language (SysML) with MATLAB-based physics simulations to standardize mission requirement characterization, establish evaluation criteria, and implement iterative design methods for refining RPO strategies and satellite design. The framework leverages the Hill-Clohessy-Wiltshire (HCW) equations for trajectory planning and incorporates rigid-body attitude control with constraints derived from physical system models. Digital twins are employed for real-time validation and monitoring, enabling the dynamic capture of mission requirements. Validation through case studies of representative RPO scenarios demonstrates the framework
Abstract ID: 50DCASS-099
Model Based System Engineering Framework Applied to a Generic Hypersonic Vehicle
Charles Lampp
Air Force Institute of Technology
Jose Camberos
Air Force Institute of Technology
David Jacques
Air Force Institute of Technology
Awaiting public release.
Orbital Mechanics
Abstract ID: 50DCASS-006
Poincare Mapping and Trajectory Analysis in the Circular Restricted N-Body Problem (CRNBP) for the Jovian, Saturnian, and Plutonian Systems
Annika Gilliam
Air Force Institute of Technology
Robert Bettinger
Air Force Institute of Technology
The Circular Restricted N-Body Problem (CRNBP) is a novel dynamical model that has not been extensively analyzed in the context of trajectory determination. One method used to locate trajectories, Poincare mapping, is applied to the CRNBP here. These maps are examined in the context of three distinct complex environments to analyze the behavior of the CRNBP due to changing conditions. The Jovian-Galilean moons system, the Saturnian system for ten total system bodies, and the Plutonian system with Hydra as a minor secondary are all presented and analyzed using Poincare maps in the Circular Restricted 3-Body Problem (CR3BP) and the CRNBP. As the CRNBP varies over time and has a dependence on the initial phase angle of any added celestial body, Poincare map structures evolve based on the chosen system configuration. Additionally, maps are generated for a variety of Jacobi constant values. As a result, all CRNBP cases yield a more chaotic system, but some conditions may be less perturbed over a chosen period. For the final case of the Pluto-Hydra system, the impact of a much more massive added body, Charon, clearly demonstrates the ability of the CRNBP to reveal stable quasi-periodic trajectories via Poincare mapping and highlights potential applications of the CRNBP to initial mission planning. In that case, extensive differences from the traditional CR3BP are obvious. Growing scientific interest in missions to moons located in complicated and dynamic systems necessitates the use of the CRNBP in initial system analysis and trajectory determination.
Space Systems
Abstract ID: 50DCASS-013
Modeling, Simulation, and Design Optimization of Space Mobility and Logistics Architectures
Hunter Rausch
Air Force Institute of Technology
Awaiting public release.
Abstract ID: 50DCASS-015
Launching a Deputy to a Target Area From a High Degrees-of-Freedom Space Manipulator on a Free-Floating Base
Sarah Hudson
University of Cincinnati
Diego Quevedo
University of Cincinnati
Donghoon Kim
University of Cincinnati
Servicing space satellites presents complex challenges, especially as debris increases and
satellites become obsolete. A promising long-term solution involves using a high-degree-of
freedom space manipulator to deploy a deputy into specific orbits for tasks, such as maintenance,
inspection, and debris removal. The manipulator employs a throwing motion to transfer the
deputy to a target area, using a precise delta-v to minimize reliance on onboard thrusters. A key
consideration in this process is the dynamic coupling between the manipulator arm and the free
floating base. As the arm moves, it induces motion in the base, which in turn affects the state of
the deputy. To optimize the transfer, we analyze the forces and torques acting between the robot
arm and the base satellite, selecting the ideal motion to minimize base disturbance. Additionally,
we explore the use of different deputy sizes for various missions, testing how CubeSats of
different sizes impact trajectory performance and base spacecraft stability.
(Approved for Public Release; distribution is unlimited. Public Affairs release approval # AFRL-2025-0558)
Abstract ID: 50DCASS-016
Dynamic Modeling and Control of a High Degrees-of-Freedom Free-Flying Space Manipulator System Performing Robot-to-Robot Handover
Diego Quevedo
University of Cincinnati
Sarah Hudson
University of Cincinnati
Aaron Borger
University of Cincinnati
Donghoon Kim
University of Cincinnati
As in-space servicing, assembly, and manufacturing (ISAM) continues to expand, the need for advanced robotics in space grows. This work explores the relatively unexplored challenge of robot-to-robot handover within the context of in-space operations. The proposed space manipulator system is designed for autonomous in-space servicing and is capable of performing object handovers between robotic arms. The key contribution of this work is the development of an integrated kinematic and dynamic model, along with a trajectory generation technique and control law, that support efficient cooperative manipulation. With a 20-degrees-of-freedom free-flying platform, the system enhances the operation flexibility of ISAM spacecraft and provides a versatile solution for a variety of future missions.
(Approved for Public Release; distribution is unlimited. Public Affairs release approval # AFRL-2025-0556.)
Abstract ID: 50DCASS-017
Design and Evaluation of a Deployment Mechanism for a Space-Based Origami Mirror in CubeSat Missions
Arturo Luna
Air Force Institute of Technology
Robert Bettinger
Air Force Institute of Technology
Suboptimal lighting and eclipse periods in space hinder inspection and in-orbit servicing missions. To address this, AFIT launched the MIRRORS initiative to enhance the illumination of resident space objects (RSOs) by reflecting sunlight from a mirror satellite. Leveraging an origami-inspired design for compact packaging, AFIT developed a cube origami flasher mirror—constructed from aluminum panels and passively unfolded with Nitinol hinges. This research investigates the necessary deployment and panning mechanisms for a dynamic, space-based origami mirror to advance the MIRRORS initiative. Considering the stringent volume constraints of small satellites, the deployment mechanism was restricted to a 2U (2000 cm^3) volume within a 6U CubeSat. This research presents a hinge deployment mechanism developed through iterative design to satisfy mission requirements and comply with NASA's General Environmental Verification Standards (GEVS), which stipulate that resonant modes must exceed 70 Hz to prevent damage during launch. Computer-aided design (CAD) and finite element analysis (FEA) confirmed a fist-mode resonant frequency of 194.74 Hz, demonstrating the mechanism's structural integrity. An event tree analysis identified potential failure modes to guide risk mitigation. Functional prototype tests demonstrated successful deployment from a CubeSat chassis and achieved a panning range of up to 13 degrees to accurately direct reflected sunlight toward an RSO at distances up to 20 m. Torque and power draw calculations further informed future mission planning, while simulations verified that the RSO's irradiance meets the optical camera's capture requirements, with measures in place to prevent overexposure. Overall, this research provides a compelling proof-of-concept for a novel hinge deployment mechanism, advancing the capability for augmented illumination in in-orbit servicing and inspection of RSOs.
Abstract ID: 50DCASS-019
A Modular Robotic Testbed for In-Space Mission Simulation
Anirudh Chhabra
University of Cincinnati
Aaron Borger
University of Cincinnati
Donghoon Kim
University of Cincinnati
Recently, the number of space missions has increased rapidly, driving significant interest in in-space servicing, assembly, and manufacturing missions. As a result, Hardware-in-the-Loop (HIL) simulations of in-space missions have become a popular area of research. In previous works, the authors proposed a novel HIL spacecraft simulator design that offers several advantages over existing robotic simulators. However, certain issues were observed, and therefore, this research enhances the current HIL simulator by incorporating hardware and software upgrades that support spacecraft simulation using diverse approaches. These advancements aim to provide accurate, flexible simulations for robust testing and validation of spacecraft control algorithms under realistic conditions, improving the performance of autonomous spacecraft systems in complex environments. To validate the functionality of the HIL simulator, an experimental study is conducted in which the hexapod represents a servicer spacecraft and the arm represents a space manipulator aiming to insert a module into the target satellite, which is undergoing tumbling motion. Based on the experiment's results, certain aspects of the HIL simulator are discussed, along with potential future improvements.
Abstract ID: 50DCASS-032
Exploring Dynamic Coupling in Space Manipulator System Using Nonlinear Correlation Tools
Gargi Das
University of Cincinnati
Daegyun Choi
University of Cincinnati
Donghoon Kim
University of Cincinnati
Space manipulator systems (SMSs) are revolutionizing modern on-orbit servicing satellite systems, serving as indispensable tools for mission success and advancing space operations. Operating in microgravity without any fixed ground, these systems induce significant dynamic interactions with the base satellite, affecting its attitude and translational dynamics due to the conservation of momentum. This phenomenon, known as dynamic coupling, requires careful analysis to ensure precision and stability during complex orbital operations. Existing studies often rely on deterministic models based on rigid body dynamics, which struggle to capture the intricate dependencies and nonlinearities inherent in SMSs. To address this, nonlinear correlation tools are employed to uncover hidden dependencies and system sensitivities, resulting in more accurate models. Dynamic coupling involves interacting variables, such as joint torques, angular velocities, base attitude, and manipulator orientations. Nonlinear correlation measures can provide insights into their interdependencies for improved system design and control. These tools extract insights directly from experimental or simulation data, identifying critical operating regimes where dynamic coupling is most pronounced. In this work, the Spearman rank correlation is employed to identify nonlinear monotonic relationships, while mutual information (MI) captures both linear and nonlinear dependencies. High MI values indicate strong dynamic coupling, with heatmap visualizations revealing critical interaction patterns for deeper interpretation. Understanding these correlations facilitates the design of robust control strategies to mitigate undesired effects, enhancing mission safety and efficiency. This work highlights the potential of nonlinear correlation tools in advancing SMS analysis and control, laying the foundation for more reliable and autonomous space operations.
