Dayton Engineering
Sciences Symposium
List of Submitted Abstracts
* 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.
Oral Presentations
Design & Optimization
Abstract ID: DESS2019-004
Investigating mechanical response of heterogeneous lattice cell structures involving vertical and horizontal struts using numerical approaches
Tahseen Alwattar
Wright State University
Dr. Ahsan Mian
Wright State University
Application-specific mechanical properties of lattice cell structures (LCS) can be obtained by combining different unit cell patterns. The main idea of this research is to predict the characteristics of combining lattice cell structures (CLCS) and using equivalent solid material investigate the effect of the vertical and horizontal struts on compression characteristics of the CLCS. In this study, the mechanical behavior of two different lattice configurations are considered: the structure geometry is occupied with BCC lattice cell type called “InsideBCC”, which has BCC unit cell inside the frame and the structure geometry has a diagonal strut on each face of the brick called “FrameCross. First, the finite element analysis (FEA) based computational approach is used to simulate and calculate the mechanical responses of BCC unit cell, vertical struts in a unit cell and unit cell frames individually. The Kriging surrogate model is then used to collect data obtained from FEA simulations and predict the properties of the combination of individual unit cells. For example, the properties of BCCV is expected to be the combination BCC and vertical strut unit cells. To evaluate the applicability of the Kriging method, the predicted mechanical response is compared with experimental and FEA model. The analysis is repeated for different strut diameters, dimensions of a single unit cell sizes, and aspect ratios (diameter truss/ unit cell length). The LCS specimens are fabricated on a Fused Deposition Modeling uPrint SEplus 3D printer using Acrylonitrile Butadiene Styrene (ABS) and tested under compression. Experimental load-displacement behavior and the results obtained from both the FEA models and Kriging model are in good agreement within the elastic limit.
Abstract ID: DESS2019-012
Spacecraft Attitude Determination using Terrestrial Illumination Matching
Liberty Shockley
Air Force Institute of Technology
Major Robert A. Bettinger
Air Force Institute of Technology
An algorithm to conduct spacecraft attitude determination via terrestrial illumination matching (TIM) is presented consisting of a novel method that uses terrestrial lights as a surrogate for star fields. Although star sensors represent a highly accurate means of attitude determination with considerable spaceflight heritage, TIM provides a potentially viable alternative in the event of star sensor malfunction or performance degradation. The research defines a catalog of terrestrial light “constellations,” which are then implemented within the TIM algorithm for attitude determination of a generic spacecraft bus. With the algorithm relying on terrestrial lights rather than the establish standard of star fields, a series of sensitivity studies are showcased to determine performance during specified operating constraints, to include varying orbital altitude and cloud cover conditions. Also, a comparison of relative attitude determination accuracy will be presented for TIM with a generic star sensor.
The current research proposes to analyze using these novel image processing techniques for initial position and attitude determination, as well as a real-time system. The city lights data base will first be based off NASA’s “Black Marble,” given in Fig. 1, which used VIIRS images over a yearlong period that removed atmospheric conditions and undistorted them to create a flat map of all the city lights of the Earth at night [3].
The TIM algorithm will take in images from a nadir-pointing camera, employ the image processing algorithms, determine an initial position and attitude estimate, and then pass the resulting information into an extended Kalman filter (EKF). This process is shown in Fig. 2, where the options in similar OpenCV algorithms are given in parallel.
MATLAB and OpenCV (in Python) will be used to integrate terrestrial light data and create a real-time TIM algorithm for successful attitude determination. Since OpenCV works better with black and white images, a high contrast version of Black Marble that shows only city lights in black, as shown in Fig. 3, will be used for comparison with the unedited “Black Marble.” Both images with VIIRS samples have gotten many successful matches using OpenCV’s tools. The next step is to choose the best match between two images for a good measurement.
The resulting images will be processed in a computer separate from the ongoing simulation in order to create an unbiased test. TIM will be employed with the “Black Marble” database and position and attitude can be found. As research goes on, different image processing techniques, such as color correction, distortion, sunlight, and thresholding will be explored [15,4]. The goal of the work is to obtain the attitude matrix fast enough to be used in a real mission. Vision based sensors, whether for tracking urbanization and population densities in remote areas, or for absolute positioning on landers, are becoming increasingly important [16,17]. Every avenue of capability must be explored as we continue to see the world in new ways.
[pictures omitted, references omitted]
Abstract ID: DESS2019-016
Designing New Generations of Body-Centered Cubic (BCC) Lattice Structures Based on Strut Orientation and Length
Hasanain Abdulhadi
Wright State University
Dr. Ahsan Mian
Wright State University
Lattice structures (LSs) have been exploited for wide range applications including mechanical, thermal, and biomedical structures because of their unique attributes combining the light weight and high strength. The first main goal of this research is to investigate the effect of strut orientation and length on the mechanical characteristics of body centered cubic (BCC) LS subjected to a quasi-static axial compressive loading within linear elastic limit using finite element analysis (FEA). In this study, two lattice generations were built and analyzed in commercial finite element software, ABAQUS/STANDARD, version 6.16, using a “smart procedure”, which was developed for this research to reduce the computational time and increase the accuracy of results by creating hexahedral mesh elements. The first generation, called as constant weight models, comprises thirteen models having fixed strut length with strut angle variation from 40° to 100° with a step of 5°. The second, namely variable weight models, also includes thirteen models; however, having variant strut length, kept constant for a single unit cell and through the entire model but varied from one model to another, with the same strut angle variation as the first generation. Also, it is worthwhile to mention there is a common model between the two sets, called the reference out of which all other models in both sets were composed such that the total number of topologies adopted in the current study are twenty five (25). The reference topology represents the standard BCC configuration of 70.53° strut angle with 5mmx5mmx5mm dimensions for a single unit cell and the other models were tailored from it based on changing the strut angle and length with 3x3x3 unit cells in x, y and z directions. Furthermore, specimens of the reference model were fabricated by a fused deposition modeling (FDM) based 3D printer using Acrylonitrile Butadiene Styrene (ABS) material and tested experimentally under compression for the purpose of validating the employed boundary conditions.
Abstract ID: DESS2019-025
Experimental and computational investigation of the post-yielding behavior of 3D printed polymer lattice structures.
Abdalsalam Fadeel
Wright State University
Dr. Ahsan Mian
Wright State University
Sandwich structures are widely used due to its light weight, high strength, and higher energy absorption. The cores of the sandwich structure are typically fabricated by using high strength cellular materials such as aluminum and titanium alloys, or polymer foams, and honeycombs. Due to its design freedom, lattice cell structures (LCS) are currently being investigated as core material. Three-dimensional printing (3DP), one of the additive manufacturing techniques, is recently being explored in creating lattice cores. Additive manufacturing (AM) has rapidly become popular in recent years due to many reasons, easy to fabricate, low cost, capability to performing complex geometrical shapes. However, studying the mechanical response of LCS experimentally are costly in time and materials. As such, the finite element analysis (FEA) can be used to predict the mechanical behavior of LCS with many different design variations more economically. There has been some attempt in developing FEA simulation of LCS; however, the FEA modeling of post-yield behavior of LCS is not reported much in literature. Therefore, in this research, the study focuses on the response of different LCS at post-yielding stages. A total of six LCS configurations were tested and studied experimentally and computationally. The LCS configurations are body centered cubic (BCC), BCC with vertical struts at every node (BCCV), BCC with vertical struts added in alternate layers (BCCA), tetrahedron (Tet), tetrahedron with horizontal structs (TetH), and pyramidal (Pyr). It may be mentioned here that the Tet and TetH configurations are similar except that TetH have horizontal struts connecting all the base nodes. Finally, for the pyramidal the struts are connected and distributed pyramidal shape. All the six configurations were fabricated using a fused deposition modeling (FDM) based 3D printer with acrylonitrile-butadiene-styrene (ABS) thermoplastic. Specimens were then tested under compression in the z direction under quasi-static conditions. FEA was used to model to capture the post-yielding compressive behavior of the different LCS. A reasonably good agreement is observed between the experimental and FEA results.
Abstract ID: DESS2019-026
Application of an Event-Driven Logistics Network to Cislunar Space Operations
Alexander Collins
Air Force Institute of Technology
Kirk W. Johnson
Air Force Institute of Technology
PENDING PUBLIC RELEASE
Motivation:
As space becomes increasingly crowded and the world increasingly dependent
on satellite-provided services, both the risk and consequences of a collision rise
dramatically. In such an event, repair, refueling, and, if necessary, reconstitution
(R3) must be caried out in a timely and cost-effective manner. The idea of an on-
orbit logistics network of spare satellites has been proposed before [4], and others
a promising method for R3 of distressed satellites more quickly and reliably than
a launch from Earth. It is critical that whatever assets make up this network are
at low risk of damage themselves. One solution to this requirement is the use of
cislunar trajectories, which can keep replacement satellites and service vehicles
away from crowded orbits, reducing their risk of collision while they wait to be
needed. These trajectories also offer a potential for very low-cost inclination
changes that could enable the R3 of vastly different constellation orbits.
Background:
This investigation attempts to model a selection of cislunar logistics networks as
mixed-integer linear programming (MILP) problems. The method for creating
these MILP problems draws heavily from the Event-Driven Generalized Multi-
Commodity Network Flow (ED-GMCNF) method proposed by Jagannatha and
Ho and derived from Ishimatsu. The most promising cislunar candidates
are Lagrange point orbits and resonant orbits, which lend themselves well to
exploiting the low-traffic, low-cost transfer perks mentioned above.
Method of Solution:
For each network, the rst step is the assembly of a library of nodes, arcs,
constraints, and commodities. Nodes refer to speci c points of interest in the
logistics network, more accurately speci c state vectors in a coordinate frame.
Examples include a launch site on Earth and a parking orbit around an Earth-
Moon Lagrange point. In a time-expanded network, each node is also attached
to a discrete point in time. Arcs are the possible paths between nodes. Multiple
arcs can connect the same pair of nodes, and holdover arcs connect the same
node in two different time instances. Commodities are anything that travels
through the network, not just the actual payloads. The constraints ensure
that the network models realistic spacecraft behavior, for example, requiring
fuel to travel in a vehicle. Once this library is assembled, mixed integer linear programming-based optimization software can be used to determine the best
arcs, nodes, and commodities to use.
The prelimnary phase of this investigation, a simple network to test the
viability of the ED-GMCNF, has already been completed, and has generated
promising results, as well as revealing some of the limitations.
Abstract ID: DESS2019-031
Developing Scaling Laws to Predict Elastic Mechanical Characteristics and Geometrical Parameters of Modified BCC Lattice Structures
Hasanain Abdulhadi
Wright State University
Dr. Ahsan Mian
Wright State University
Predicting the mechanical characteristics of lattice structures (LSs) is of high importance in the field of lattice design. This is due to the fact that the fabrication of LSs might be challenging, expensive or time-consuming. Based on that, the objective of this study is to develop generalized empirical closed-form equations using scaling laws and finite element methods (FEMs) to predict not only the compressive elastic mechanical properties (CEMPs) but also the geometrical parameters (GPs) of modified BCC LSs, with considering both the effect of lattice cell tessellations and the influence of material distribution at strut intersections. For that purpose, the relative density (RD) is varied from 0.14-0.3 with a step of 0.02 by changing the strut diameter, corresponding to strut angle variation from 40° to 100° with a step of 10°. Then, design constraints regarding the strut length are applied on the entire set of RDs and strut angle variation to develop two generations of modified BCC LSs. The first comprises sixty three models (63) of fixed strut length and the other also consists of sixty three models (63) but with variant strut length. Due to the repetition of nine models with strut angles of 70.53° in both generations, the total number of models adopted to achieve this goal are one hundred seventeen (117), not (126), all built and analyzed using ABAQUS/STANDARD. The Data ensuing from FE simulation of the axial compression test within liner elastic limit are thereafter applied to the most important formula of Gibson and Ashby, which represents the relationship between the relative elastic modulus (RE) and relative density (RD), in order to determine Gibson and Ashby’s pre-factors, C1 & n. In addition, all other geometrical parameters are correlated with the RD based on the measurements of the geometries of 117 lattice models using ABAQUS diagnostic tools. The results showed that the generalized empirical closed-form equations can predict well both CEMPs and GPs. In addition, the RE increases with increasing both the strut angles and the RD. Besides, Gibson and Ashby’s coefficients are found to be approximately similar for both generations.
Abstract ID: DESS2019-035
A Machine Learning Framework for Drop-in Volume Swell Characteristics of Sustainable Aviation Fuel
Shane Kosir
University of Dayton
Joshua Heyne
University of Dayton
John Graham
University of Dayton Research Institute
A machine learning framework has been developed to predict volume swell of non-metallic materials commonly found in commercial aircraft fuel systems submerged in blends consisting of neat molecules of interest and zero-aromatic synthetic paraffinic kerosene. Volume swell, a material compatibility concern, serves as a significant impediment for the minimization of the environmental impact of aviation. Sustainable aviation fuels, the only near and mid-term solution to mitigating environmental impacts, are limited to low blend limits with conventional fuel due to material compatibility/O-ring swell issues. A neural network was trained to predict volume swell for neat molecules found in conventional jet fuel. Subsequent blend optimization incorporated swell predictions for cycloalkanes (saturated cyclic hydrocarbons) and iso-alkanes (branched linear hydrocarbons) to create a high-performance jet fuel within ‘drop-in’ limits.
The neural network was able to predict neat molecule volume swell with a cross-validated mean absolute error of 4.5% v/v and a test mean absolute error of 2.5% v/v. Optimization considering nitrile rubber volume swell achieved median specific energy [MJ/kg] and energy density [MJ/L] increases of 1.9% and 5.1% respectively relative to conventional jet fuel while maintaining a median volume swell of 6.2% v/v, 68% higher than the lower swell limit. Optimized solutions were heavily biased toward monocycloalkanes, indicating that they are a suitable replacement for aromatics (unsaturated cyclic hydrocarbons). This study concludes that cycloalkanes can replace aromatics in jet fuel considering volume swell and other operability requirements while also significantly reducing soot and particulate matter emissions, which are largely associated with aromatics in conventional jet fuel.
Abstract ID: DESS2019-036
Neural-Fitted Integral Reinforcement Learning in Simulink
Andrew Ellicott
Wright State University
Dr. Rory Roberts
Wright State University
Integral reinforcement learning is a numerical technique for solving the optimal control problem of controlling a plant in order to minimize a cost function. It extends reinforcement learning from discrete to continuous state spaces. Reinforcement learning requires the learning of an unknown value function unique to each plant. The use of a neural network to fit this value function was tested in Simulink on simple linear test models and compared against analytical results.
Abstract ID: DESS2019-038
Bio-Inspired Evolutionary Design of a 2D Morphing Wing Section in Unsteady Supersonic Flow
Joshua Hodson
Hodson Aerospace LLC
Gregory Reich
Air Force Research Laboratory
Joshua Deaton
Air Force Research Laboratory
Alexander Pankonien
Air Force Research Laboratory
Philip Beran
Air Force Research Laboratory
A bio-inspired aeroelastic topology optimization method is applied to an aerostructure design problem. Evolutionary design processes are applied programmatically to a population of randomly-generated design candidates using a coupled multiphysics objective function. The specific design problem solved here is that of a morphing 2D airfoil in supersonic flow maneuvering along a prescribed flight path. Aerodynamic control of the airfoil is achieved through actuation of distributed active elements embedded within the airfoil topology. The aerodynamic efficiencies of the airfoil under specified lift conditions are maximized subject to constraints on pitching moment, actuation force, and actuator stroke length. Design candidates capable of achieving the prescribed lift conditions under actuation are further evaluated on their ability to transport a payload along a prescribed flight path.
Each design candidate is represented by a set of values called a genome. A genetic algorithm is used to randomly generate an initial population of design candidates and then evolve the population through the application of evolutionary algorithms that mimic the biological processes of mutation, reproduction, and natural selection. A graph-based Lindenmayer system is used to convert each genome into a topology by iteratively applying a set of rules to an initial axiom, the rules and axiom being encoded within the genome. This allows for a relatively small number of design variables in the optimization problem when compared to strictly gradient-based optimization methods for exploring topological spaces of similar size and complexity.
Gradient-based optimization is used to determine actuator settings that maximize lift-to-drag ratio while satisfying the specified lift requirements and other constraints. Low-fidelity analysis tools for supersonic aerodynamics and structures are used to efficiently process large populations of design candidates. Multidisciplinary coupling of the codes is achieved using an explicit formulation, and an implicit time integration scheme is used for the unsteady maneuvering problem.
Pareto-optimal designs from the final population are presented and specific characteristics that allow these designs to outperform their peers are identified. Limitations in the design process are discussed and suggestions for improving the process are made. Future applications for the topology optimization method are also discussed.
This abstract has been cleared for public release, case number 88ABW-2019-4192.
Abstract ID: DESS2019-039
Characterizing the Strength and Rigidity of Tensegrity Aircraft Wings
Austin Mills
University of Dayton
Dr. David Myszka, Dr. Andrew Murray
University of Dayton
Dr. James Joo, Dr. Daniel Woods
Air Force Research Laboratory
The suitability of tensegrity aircraft wing concepts and a comparison with their simulated structural performance to a baseline conventional wing structure is presented. Tensegrity systems, which consist of arrangements of struts and cables, are appealing for their structural efficiency. With each member loaded in tension or compression, tensegrity systems enable lightweight structures. Of specific interest, tensegrity systems may provide a pathway to morphing aircraft structures through the actuation of cables. Aircraft with morphing features can provide multi-role and multi-mission capabilities by adapting for different in-flight requirements, improving aerodynamic efficiency and performance. With an eye towards morphing applications, tensegrity-based wing designs were compared, both using engineering design judgement and through structural topology optimization methods, to a conventional wing structure as a baseline performance case. The baseline conventional wing model, selected to be representative of the Van’s RV-4 aircraft, consists of a rib/spar structure, composed of aluminum alloy 2024. This conventional wing model was subjected to aerodynamic loading conditions that simulate a 2g pullup maneuver. The first tensegrity concept, developed with design judgment, is configured by merging established unit cells. The second tensegrity design, in contrast, was developed by application of a topology optimization algorithm, minimizing the weight with maximum stress constraints. The tensegrity wing concepts were sized to yield deflections and strain energies comparable to the conventional wing at a fraction of its weight. Additionally, a static prototype of a tensegrity wing was designed and fabricated to gain insight into the development process. This investigation is intended to further the understanding of tensegrity-based designs in the context of aerospace structures, potentially enabling future tensegrity aircraft wings that permit morphing capabilities.
Abstract ID: DESS2019-040
Orthogonal Reference Surrogate Fuels for Operability Testing
Harrison Yang
University of Dayton
Robert Stachler, Joshua Heyne
University of Dayton
Sustainable aviation fuel is being sought as an opportunity to mitigate carbon emissions in the aviation community. To increase confidence in safety and performance, engine operability tests on combustor figures of merit (FOM) pertaining to fuel effects are needed, specifically on lean blowout, high altitude relight, and cold start ignition. Key fuel performance properties (surface tension, viscosity, density) were determined as critical properties for ignition probability in multiple experimental rigs, an essential measurement in cold start ignition experimental testing. In this effort, a previous surrogate calculator and relative bulk property blending rules were integrated into a new, robust Jet Fuel Blend Optimizer (JudO). A surface tension blending rule was validated and incorporated into this developed tool. Four jet fuel surrogates were developed using JudO to create the greatest variance in the key fuel properties, which would increase the variance in the results from experimental efforts. With those scenarios tested experimentally, we can further understand the influence on the key properties relative to ignition probability.
Fluid Dynamics / CFD
Abstract ID: DESS2019-001
Propeller Partial Ground Effect
Jielong Cai
University of Dayton
Sidaard Gunasekaran
University of Dayton
Anwar Ahmed
Auburn University
Michael OL
Folderol, LLC
We extend our recent propeller ground-effect study to consider a circular ground-plate, instead of planar-surface of assumed infinite extent. Parameter studies include propeller to plate diameter ratio, propeller diameter to ground-offset ratio, and propeller pitch to diameter ratio. As with classical ground effect, benefits of thrust-augmentation and/or power-reduction with proximity to the ground, depend on the propeller pitch to diameter ratio. Flow visualization suggests that for larger pitch to diameter ratio, the lack of conclusively large ground-effect benefits can be attributed to stalled flow about the blade, and spatially more diffuse tip-vortices. A circular ground-plate of half of the propeller diameter was found to have almost no distinction from that of an unimpeded free-stream, while when the plate and propeller have the same diameter, the resulting ground-effect already resembles that of the infinite-plate.
Abstract ID: DESS2019-037
Investigation of Scaled Down Doppler Lidar for Velocity Measurements in Wind Tunnels
Samuel Barnhart
University of Dayton
Dr. Sidaard Gunasekaran
University of Dayton
Particle Image Velocimetry (PIV) and Laser Doppler Velocimetry (LDV) are two common techniques used for velocity measurements in wind tunnels, but both methods require seeding the tunnel with tracer particles. The disadvantages of seeding particles are well known in the industry. The seed doesn’t follow the flow in the areas where Stokes number is greater than unity. This creates issues in high speed applications, maintaining uniform seeding in vortical flows, and flows involving mixing and chemical reactions. One potential solution to avoid seeding of the flow is the use of Doppler light detection which relies on Rayleigh scattering from air molecules instead of Mie scattering from particulates. A system level analysis of the transmitter, receiver and optics revealed that a 200 mW laser combined with a linear mode APD detector can be used to measure the Doppler shift in the freestream velocity range between 0 and 50 m/s for a target distance of 1.5 m. The constructed lidar system was successfully used to measure the vibration frequency of an ultrasonic piezoelectric transducer at 1.4 MHz as a proof of concept and as a validation of the test setup. Work is currently being done to then measure the corresponding Doppler shift associated with a known velocity of seeded flow. This data is being analyzed by employing enhanced post-processing techniques to identify the weak signals among background noise.
Abstract ID: DESS2019-041
Design and Analysis of an Additive Manufactured Supersonic Wind Tunnel
Scott Chriss
University of Dayton
Austin Abel, Undergraduate Student
University of Dayton
Matthew Gazella, Aerospace Engineer
Air Force Research Laboratory
Sidaard Gunasekaran, Assistant Professor
University of Dayton
For decades, supersonic wind tunnels have been designed and fabricated using common metallic materials for structural integrity. With advancements in additive manufacturing technology, high-performance thermoplastics are being developed as an alternative to steel that offer excellent strength and thermal stability. Utilizing the latest advancements in additive manufacturing, an additive manufactured supersonic wind tunnel with Mach 2+ capability was designed and fabricated using ULTEMTM 1010 resin for academic and research purposes at the University of Dayton Aerospace Laboratory. The supersonic wind tunnel features an open loop, blowdown design utilizing a high-pressure nitrogen cylinder with a settling chamber, modular C-D facility nozzle, modular isolator, modular 1" x 1" test section with optical access for flow visualization, and a supersonic step diffuser. High fidelity CFD analysis was performed using CFD++ software on an AFRL supercomputer which verified the flowpath design estimated from method of characteristics and ideal starting conditions. The CFD analysis was performed with over 10.5 million cells and assumed compressible perfect gas while using a RANS 2-equation cubic k-ε turbulence model with viscous terms. Experimental tests were conducted using a Schlieren flow visualization system to validate the CFD analysis and characterize operating conditions for future educational and research activities.
Abstract ID: DESS2019-052
Computational Assessment of Aortic Valve Function and Mechanics Under Hypertension
Saurav Kadel
Wright State University
Philippe Sucosky
Wright State University
The aortic valve (AV) is the semilunar valve located between the left ventricle (LV) and the aorta which controls the unidirectional flow of blood. The most common AV disease is calcific aortic valve disease (CAVD), which is present in 0.4% of the general population and 1.7% of geriatric population. CAVD is a slow progressive disorder that ranges from mild valve thickening to severe calcification of valve leaflets. CAVD leads to increased workload on the LV and reduced cardiac output resulting in left ventricular hypertrophy, aortic dilation, and other cardiovascular complications. Hypertension, which is characterized by persistent increase in mean arterial blood pressure, is considered an independent risk factor for CAVD. However, the mechanisms by which hypertension may promote CAVD are still unknown. Previous studies have highlighted the sensitivity of AV tissue to hemodynamic stress abnormalities, while others have suggested the existence of valvular hemodynamic alterations under hypertensive conditions. Therefore, the hypothesis of this study is that hypertension generates wall shear stress (WSS) abnormalities on AV leaflets, which may activate pathological cascades leading to tissue remodeling and inflammation. In order to assess this hypothesis, the objective of this study is to quantify computationally the WSS environment on AV leaflets subjected to normotensive and hypertensive conditions. A realistic geometrical model of a human AV and aortic root was constructed, and fluid structure interaction simulations were performed in ANSYS to quantify the mechanical interactions between the deforming valve leaflets, the compliant aortic root and the pulsatile blood flow under physiological (120/80 mmHg) and hypertensive (150/100 mm Hg) aortic pressure. This presentation will discuss the impact of hypertension on leaflet dynamics in terms of fluid WSS, structural strains and leaflet deformation.
Abstract ID: DESS2019-055
Fundamental Flowfield Production using a Swirl Generator Rig
Marcus Acton
Wright State University
Dr. Mitch Wolff
Wright State University
Dr. Mike List
Air Force Research Laboratory
Inlet swirl can negatively impact the performance and operability of a gas turbine compressor which will result in reduced thrust. As a leader in discovery and development of air technologies, the Air Force Research Laboratory is investigating inlet swirl effects on a high-speed, transonic fan stage. The research follows practices set forth by the Society of Automotive Engineers (SAE) S-16 Committee on Turbine Engine Inlet Flow Distortion. Realizing complex swirl patterns can be described by a blend of more basic distortions, the SAE S-16 Committee has proposed 4 basic types of swirl patterns. The basic patterns include bulk swirl, a single-direction rotating flow; paired swirl, a set of counter-rotating swirls; cross-flow induced swirl, a swirl induced from a flow parallel to the compressor face; and tightly-wound vortex, a small swirl that effects a local area of the flowfield. For describing these flow components while considering a use in quantifying fan performance, the SAE S-16 Committee has also formulated a methodology to describe the constituent swirls with a set of swirl descriptors. These descriptors will be used in expressing the resultant swirl of a “Swirl Generator.” The “Swirl Generator” was designed and built by the Arnold Engineering Development Complex to produce a mass of different swirl distortions using 16 independently-controlled vanes each with 5 independently-controlled flow turning flaps and an adjustable leading edge. This device drastically reduces time between studies and can even allow transient effects to be researched. The swirl generator will initially be tested in the Compressor Aero Research Laboratory’s Annular Cascade Facility where swirl patterns will be measured and characterized before moving to the compressor facility for fan performance measurements.
Heat Transfer / Thermal Sciences
Abstract ID: DESS2019-007
Modeling Nonlinear Heat Transfer for a Pin-on-Disc Sliding System
Brian Boardman
Air Force Institute of Technology
Dr. William Baker
Air Force Institute of Technology
Dr. Anthony Palazotto
Air Force Institute of Technology
A numerical method is developed to characterize the conduction of heat and analysis of wear rates for samples of Maraging-300 steel (Vascomax 300) subject to frictional heat generated from a pin-on-disc type experiment under high-pressure, high-velocity, sliding contact conditions. A two-dimensional, nonlinear heat transfer equation in cylindrical coordinates is discretized and solved via a second-order, implicit, finite-difference scheme to model the temperature distribution of the pin. This schematic is then used to predict material removal from the specimen over time based off the temperature profile of the pin. The results of the model are compared with temperature and wear rate data collected from experimental tests under various velocity and force profiles.
Abstract ID: DESS2019-032
Multi-Mode Rankine Cycle for Power and Thermal Management
Nathaniel Payne
Wright State University
Dr. Mitch Wolff
Wright State University
Dr. Rory Roberts
Wright State University
Levi Elston
Air Force Research Laboratory
Hypersonic vehicles represent a classification of next generation aerospace vehicles, capable of obtaining speeds greater than Mach 5. Most of the work to date regarding hypersonic vehicles has concentrated on vehicle aerodynamics, materials, structures, and propulsion systems improvements, with little effort on an integrated sub-systems approach. Two sub-systems that present a significant challenge for hypersonic vehicles are the power generation and thermal management sub-systems. The air friction that is experienced at high speeds, particularly around the engine, generate large thermal loads that need to be managed. Traditional jet engines also do not operate at speeds greater than Mach 3, therefore eliminating the possibility of a traditional rotating power generator. This research is currently investigating the potential for using the fuel as a method of transferring thermal energy from the vehicle to a Rankine cycle. The use of the Rankine cycle will reduce the thermal load on the vehicle while meeting the on board power requirements. Experimental testing will be used along with transient computer modeling to study the system dynamics.
Abstract ID: DESS2019-049
Numerical modeling of turn-down ratio for a dynamic spacecraft radiator functionalized with electrochromic surfaces
Rydge Mulford
University of Dayton
Nicholas Debortoli
University of Dayton
Calvin Callahan
University of Dayton
Spacecraft thermal radiators, which are currently limited to static geometries and static radiative properties, are incapable of adjusting to fluctuations in spacecraft thermal loads, requiring the use of power consuming heaters to achieve spacecraft temperature control. The required heater power might be reduced through real-time manipulation of radiator emission via control of surface properties or emitting area. Several technologies have been proposed by which either the surface properties or emitting area of the spacecraft radiators might be manipulated in real time, including variable emissivity surfaces (electrochromic or thermochromic) and extendable radiating surfaces. Combining radiative control technologies to operate in parallel, resulting in real-time control of both surface properties and emitting area, would augment the radiative heat transfer behavior of the resulting device.
In this work, the turn-down ratio of a re-deployable radiator coated with an electrochromic surface is numerically determined using the Segmented Fin Algorithm (SFA). The radiator panels are discretized into finite elements and a series of governing equations is written for these elements. The emissivity of the panels varies as a function of radiator extension, simulating the behavior of an electrochromic surface. The non-dimensional system of equations is solved iteratively using the Thomas Algorithm. The turn-down ratio (maximum heat transfer divided by minimum heat transfer) is given by calculating the heat transfer for a fully extended radiator with the panel emissivity set to a maximum value divided by the heat transfer for a fully retracted radiator with the panel emissivity set to a minimum value. Several combinations of maximum and minimum emissivity values are evaluated based on published electrochromic surface behavior. Initial results for the combined technology indicate a turn-down ratio of 26.5 as compared to 6 for a re-deployable radiator with static panel emissivity and 9 for the electrochromic surface operating individually. Electrochromic surfaces with the largest turn-down ratio and the lowest minimum value result in the greatest combined technology turn-down ratio. These results demonstrate the utility of combining complimentary radiative control technologies to operate in parallel.
Abstract ID: DESS2019-056
An Analysis of Well-Stirred Reactors for Combustion Applications
Robert Stachler
University of Dayton
Joshua Heyne, PhD
University of Dayton
Scott Stouffer, PhD
University of Dayton Research Institute
Knowledge of simple reactor modeling provides a fundamental framework to the understanding of complex combustion systems. A well-stirred reactor, or perfectly stirred reactor (PSR), is one of the four basic reactors in combustion modeling that has aided in the comprehension of chemical kinetics of hydrocarbons. The use of these reactors in an array with plug flow reactors provides an alternative method to simplify combustion simulations and reduce computation time compared to computational fluid dynamics and other numerical approaches. Historically, gas turbine blowoff measurements, a performance parameter for engine operability testing, correlated with PSR theory. This effort provides a review of previous and current experimental configurations to emulate a PSR. Supplemental theory of PSRs is discussed in relation to premixed turbulent combustion theory.
Abstract ID: DESS2019-066
Analyzing the Ignition Differences Between Conventional Spark Discharges and Nanosecond-Pulsed High-Frequency Discharges
Katherine Opacich
University of Dayton
Joshua S. Heyne
University of Dayton
Timothy Ombrello, Robert J. Leiweke
Air Force Research Laboratory
Joseph K. Lefkowitz
** Other (please contact webmaster)
A significant challenge to engine design is the development of combustion systems that meet increasingly strict efficiency, performance, and emissions demands. One way of achieving these aims is through fuel-lean combustion which, makes ignition difficult due to higher minimum ignition energy (MIE) and greater potential for misfires. New methods of energy deposition using nanosecond-pulsed high-frequency discharges (NPHFD) have shown promise in igniting under fuel-lean conditions. Despite these observations, questions remain regarding how the NPHFD ignition system will perform against the conventional ignition system on shorter timescales and in a more engine-relevant flowing environment. This paper compares the NPHFD ignition system to a conventional, capacitive discharge, system in a flowing environment under matching total energy and average power conditions.
Human Factors / Biomedical
Abstract ID: DESS2019-017
Influence of Factors on the Stability of Older Adults Performing an Overhead Reaching Task
Kyra Twohy
University of Dayton
Dr. Kimberly Bigelow
University of Dayton
Reaching into an overhead cabinet is a task that may be particularly challenging for older adults’ balance, putting them at increased risk of fall. Research suggests that while reaching overhead for items is common with 99% of older adults completing it, 43% of older adults have an increased fear of falling while completing the task. Despite the noted concern related to performing this task, little research has been done to examine balance and stability while reaching into a cabinet. We have focused on three factors that have been identified in research as influencing stability during this task: whether or not the individual is wearing shoes; how the feet are positioned; and whether or not the individual rises onto their toes in completing the task. The aim of our study was to provide a recommendation as to which conditions would maximize stability when reaching overhead at a 45° angle, holding a canned good. We hypothesized that balance would be the best with shoes on over off; feet in bipedal stance versus staggered; and feet flat on the floor over heels lifted. Twenty older adults (8 male and 12 female) aged 65 and older (average age of 76 ± 7.30) participated in the study. The three factors were each tested at two levels, shoes on or off (footwear), feet in a bipedal or staggered stance (foot placement), and feet flat or heels lifted (foot contact). 2 trials for each possible combination of the factor levels were conducted on a force plate for 30 seconds, for 16 total trials. In each trial, the study participants held a 15 oz. canned good overhead at a 45° angle and tried to maintain a static position. AP sway range, ML sway range, and mean velocity postural sway measures were calculated and analyzed using an ANOVA with significance p<0.05. The largest average AP sway range occurred with shoes off, bipedal stance, with heels lifted (110.7 ± 24.7 mm). The same condition but with feet flat created the lowest average AP and ML sway range (34.9 ± 23.2 mm). The largest average ML sway range occurred with shoes off, feet staggered, and heels lifted (94.1 ± 27.3 mm, 88.4 ± 17.8 mm/s). The main effect of footwear was reduced impact of changes both stance and foot contact, which produced significantly lower sway ranges. However, the lowest overall values of sway range were observed in a shoes off condition. Less foot contact in the heels lifted condition resulted in the largest way ranges and mean velocity, a significant difference from feet flat on the floor. The results of this study suggest that the safest way to reach into a cabinet is to have your shoes on, your feet side by side, and to keep your feet flat on the floor. While it is unknown whether this may prevent falls, it could be helpful for older adults who exhibit postural instability, or task-related fear of falling.
Abstract ID: DESS2019-006
Detecting Vasodilation to Prevent Heat Strokes: Transducing Liquid Flow Rate to Temperature for Sensing
Neeti Prasad
Dayton Regional STEM School
Several years ago, I had my first experience with feeling symptoms of heat exhaustion during the Dayton Air Show. Athletes, infants, and obese adults are vulnerable to heat strokes due to exertion or high ambient temperature. Vasodilation, the dilation of blood vessels to increase blood flow to cool off the body, is a symptom of heat strokes. I embarked on a journey to create a heat stroke preventive method by detecting vasodilation using a wearable sensor.
Current sensor-based methods estimate core body temperature approaching 40C to warn of a potential heat stroke, but sweat can introduce errors. Alternatively, measuring the increase in wrist size or pressure due to vasodilation is unreliable. I conducted experiments to simulate and measure vasodilation by modeling the blood flow through the blood vessels using water pumped through various rubber and copper tubing, to understand the relationship between flow rate and heat dissipation. I also developed a conceptual design of a wearable device that can be worn to transduce varying flow rate to temperature change. I then explored the use of piezoelectric sensors to transduce temperature change to voltage.
I concluded that the heat dissipation rate increased when the water flow rate is increased, even when the segmented temperature difference decreases. I characterized piezoelectric sensors to transduce temperature to voltage that can power an LED. In the future, I would like to build a wearable device that can detect heat stress, alert a patient about deteriorating condition, and ultimately prevent heat strokes.
Abstract ID: DESS2019-009
Integration of Human Factors, Cognitive Ergonomics, and Artificial Intelligence in the Human-Machine Interface for Additive Manufacturing
Sharon Bommer
University of Dayton
Eric Doran
University of Dayton
Adedeji Badiru
Air Force Institute of Technology
As additive manufacturing transitions from a technology of manufacturing prototypes to rapid manufacturing, more human factors considerations must be assessed and integrated for improved work design. This presentation provides a literature survey regarding human-machine integration for human factors, cognitive ergonomics, and artificial intelligence to improve the performance output in the additive manufacturing process. Also, research on the Design-Evaluation-Justification-Integration (DEJI) model, a systems engineering model, was performed. This work addresses how the DEJI model can be deployed to integrate modern technologies in additive manufacturing to improve the performance of both a human operator and the technology.
Abstract ID: DESS2019-010
Using DEJI Systems Model to Develop and Integrate AI-Based Technology for People with Disabilities: A Human Factors Framework
Adedeji Badiru
Air Force Institute of Technology
Sharon Bommer
University of Dayton
The recent resurgence of artificial intelligence (AI) in operational settings has led to a heightened interest in developing AI tools for common human-centric applications. This is a good thing. However, caution and prudence must be exercised to ensure that the tools developed are, indeed, relevant and effective for the intended applications. This is particularly critical where the envisioned tools are targeted for people with disabilities. The proper integration of tools for appropriate applications is the premise of this paper, which focuses on a systems framework for human factors considerations in using AI tools. Specifically, the DEJI® systems model is proposed as a methodology for aligning and integrating tools to the needs of the disabled users. The basis for the DEJI® model is structured around Design, Evaluation, Justification, and Integration. The model has been applied to a variety of problem areas. It is expected that it could facilitate a more effective development of AI tools for people with disabilities.
Abstract ID: DESS2019-046
Design of a Trike for Paraplegic Use with FES
Anthony Bazler
University of Dayton
Dr. Andrew Murray
University of Dayton
Dr. David Myszka
University of Dayton
Joe Bernicke
University of Dayton
Bennett Synder
University of Dayton
The goal of this project is to design a performance tricycle for paraplegics whose leg muscles are stimulated to pedal via Functional Electrical Stimulation (FES). FES stimulates muscle contraction with small electrical currents and has proven useful in building muscle in patients while relieving soreness and promoting cardiovascular health. An FES-stimulated cyclist produces approximately 25 Watts of power, nearly 20 times less than a typical rider. At these reduced power levels, the challenges of pedaling are amplified. For example, as the pedal follows the traditional circular path, there are portions referred to as dead zones where neither FES-stimulated leg actively propels the bike forward. One possibility for reducing or eliminating dead zones is to redesign the circular path of the pedaling motion. Bicycles have recently been marketed that feature pedaling mechanisms that employ alternate pedaling motions. In addition to addressing dead zones, these bikes also optimize the muscle capacity of the rider to deliver torque to the wheels. These new bikes achieve alternate pedaling paths through the introduction of more complicated mechanisms including four-bar and ratchet-and-pawl linkages. Such alternates are being considered for the redesign of the performance tricycle piloted by FES-stimulated riders. To investigate possible changes to the tricycle, quasi-static models have been developed for traditional and alternate cycling mechanisms. This allows for a comparison of torque generation between the mechanisms which facilitates selecting the optimal design. Such a tricycle is viewed as beneficial due to health advantages, improved mobility, and independence created for the end user.
Abstract ID: DESS2019-048
Lower body joint kinematics following an anterior cruciate ligament (ACL) reconstruction surgery
Vinayak Vijayan
University of Dayton
Shanpu Fang
University of Dayton
Dr. Allison Kinney, Dr. Megan Reissman
University of Dayton
A review paper examining anterior cruciate ligament (ACL) injuries and ACL reconstruction surgeries, revealed a trend of higher rates of ACL injuries and ACL reconstruction surgeries in pediatric subjects (ages 5 to 19), when compared to adults (ages 20 to 45). The paper suggests an increased involvement in sports as the reason for higher rates of ACL injuries in the pediatric population. When compared to adults, the skeletally immature population also had significantly higher incidences of concomitant meniscal and cartilage injuries. Despite the higher rate of ACL injuries in skeletally immature population, the research on kinematics after ACL reconstruction surgery has been skewed towards studying the joint kinematics of adults. The kinematic results from adults, who underwent ACL reconstruction surgery, might not strictly apply to pediatric patients, as the change in bone dimensions of a growing pediatric patient could affect the kinematics of the subject’s motion. The ACL reconstruction surgery itself, often, follows a slightly altered procedure for pediatric patients, to spare the epiphyseal plate. Extensive damages to the epiphyseal plate could result in growth arrest, and other growth deformities like genu valgum and tibial recurvatum. This work is a pilot study of a broader study, in collaboration with Dayton Children’s hospital, which hopes to fill the gap in knowledge about lower-body kinematics and kinetics of pediatric patients after ACL reconstruction surgery. This pilot study, examines and compares the lower-body joint kinematics of a control subject to a subject who underwent an ACL reconstruction surgery. The protocol followed in this study involves a variety of tasks (walking, running, cutting, squatting, single-legged lateral hopping, sidestepping, and forward jumps) that would help comprehensively define the lower-body joint kinematics, and identify the differences in kinematics between the control subject and the ACL patient. Motion capture data and ground reaction forces were collected for at least 10 repetitions of each task for both subjects, and this data were processed in Visual3D and OpenSim. Some of the movements, examined in this study, involved higher degrees of hip and knee flexion. So, an appropriate model, created by Catelli et alia, capable of providing accurate kinematic data at extreme angles of hip and knee flexion was chosen for this study. The scaling tool in OpenSim was used to scale the generic model, to match the dimensions of individual subjects. The bigger segments in the model were scaled nonuniformly, by defining the scaling factors along 2 or 3 axes, and the smaller segments were scaled uniformly, by using only one scaling factor. Inverse kinematics was then used to match the position of the scaled model with the experimental data, and this simulated motion of the scaled model was used to calculate the lower body joint kinematics. Insights into lower-body joint kinematics of pediatric patients, obtained from the broader study, could help in improving the rehabilitation of skeletally immature patients, and could reduce the chances of reinjury and other concomitant knee injuries in pediatric patients.
Abstract ID: DESS2019-053
Virtual Train Rides Will Challenge Your Standing Balance
Leah O'shea
University of Dayton
Megan Reissman
University of Dayton
Virtual reality was used to expose healthy younger and older adults to environments anticipated to challenge standing balance. Challenges can occur due to difficult terrain, dynamic perturbations, and visual or auditory stimuli. As these environments are difficult to replicate experimentally, this study investigates balance response to different environments using Virtual Reality (VR). If specific VR environments influence balance response, VR may become a new physical therapy tool for those with balance impairments. This study seeks to determine which environments elicit the highest response in standing sway, if short VR exposures influence balance, and potential differences in response between younger and older adults.
Ten healthy young adult participants aged 19.9 ± 0.7 years (mean ± std) and 12 healthy older adult participants aged 60.6 ± 12.4 years were recruited. Participants stood on a force plate for 1 minute during each trial. The first two trials defined the baseline stance behavior prior to VR exposure. Each participant then viewed seven VR environments in a randomized order. Boat Day, Boat Night, Train, and Busy Road had dynamic (moving) visual stimuli. We instructed participants to look forward with their hands at their side (test position) for 1 minute. Center of pressure (COP) trajectories from the force plate were analyzed in the anterior-posterior (AP) plane to determine range and variability (std).
In the VR to Baseline comparison, age was not a significant factor. COP range in the AP plane was significantly increased for environments that featured moving surfaces and dynamic visual stimuli. COP variability was significantly different between Eyes Open and Boat Night (p=0.002). COP range was significantly lower in the Eyes Open condition compared to Boat Day, Boat Night, and Train (all p less than 0.05) and Busy Road suggested a trend (p less than 0.06). Features of the Boat Night condition (low light, moving platform, dynamic visuals) may have contributed to changes in both the range and variability responses. In the non-VR comparison, the Age and Time factors were not significant. This suggests that short exposures to VR do not generate any changes in balance for either age group. The Closed eyes condition had significantly higher COP variability and range compared to Open eyes (both p less than 0.02) which is consistent with normal vestibular function.
VR representation of dynamic environments can induce an increased sway response during quiet stance. VR may benefit individuals with balance impairments by providing safe conditions under which to practice balance responses.
Abstract ID: DESS2019-054
Influence of Added Mass on the Kinematics of Over-Ground Walking: Preliminary Results
Shanpu Fang
University of Dayton
Vinayak Vijayan, Ellen Lucchesi, Peter St Amand
University of Dayton
Megan Reissman, Allison Kinney, Timothy Reissman
University of Dayton
Recently there has been increased research and development within wearable robotics, or exoskeletons, for the purposes of gait rehabilitation. The mechanical design of such systems has largely been based on previous biomechanics research, which have investigated treadmill-based gait changes in response to added mass on different leg segments. These studies have shown that added mass can influence metabolic energy consumption, but does not change the joint kinematic patterns during walking. As treadmill walking can influence joint kinematics, there is value to analyze how overground walking is influenced by added mass, so as to provide better predictions of the effects of exoskeleton masses and their locations on the body. The current study presents preliminary data collected as part of a larger study using healthy male and female subjects aged 18 – 45 years old. All participants gave written informed consent to take part in this study, which was approved by the University of Dayton’s IRB. Motion capture and ground reaction force data were collected while a single subject performed overground walking under seven conditions: Baseline, Added Shank Mass (2 and 4 lbs), Added Thigh Mass (2 and 4 lbs), and Added Trunk Mass (4 and 8 lbs). The experimental data were processed in Vicon Nexus and imported into OpenSim for analysis. In OpenSim, a generic musculoskeletal model with 23 degrees-of-freedom and 92 muscle-tendon actuators was scaled to match the geometry of the subject. Then, joint kinematics were determined using the Inverse Kinematics Tool for five trials for each of the seven conditions. Comparisons between the baseline condition and each added mass condition were made to determine if added mass influenced sagittal plane joint kinematics. Initial results with one subject indicate the knee and ankle joint kinematics are altered with respect to baseline walking for nearly all add mass conditions. Examining the knee first, at early to mid stance the knee flexion increased for nearly all conditions, and the most change, approximately 4-5 degrees, occurred with both mass levels of the “Added Shank Mass” and “Added Trunk Mass” conditions. At late stance (toe off), knee flexion increased as well for nearly all conditions and the most change, approximately 7-8 degrees, occurred with the lower mass level of the “Added Shank Mass” condition. At mid swing phase, knee flexion decreased for nearly all conditions and the most change, approximately 7-10 degrees, occurred with both mass levels of the “Added Shank Mass” condition. Examining the ankle, changes occurred during both stance and swing phases, however all changes were minor at less than 4 degrees. The only exception was for late stance, in which plantarflexion increased by 4-5 degrees for the lower mass level of the “Added Shank Mass” and the higher mass level of the “Added Thigh Mass”. Thus, preliminary results indicate that joint kinematics may be altered in addition to metabolics when wearing added masses. If such preliminary results are proven statistically true by the larger study, then the design of such exoskeletons must compensate for such altered gait.
Abstract ID: DESS2019-058
Smart Cane: A Pilot Study Examining How People Walk with Canes in the Wild
Sydney Lundell
University of Dayton
Ben Berry
University of Dayton
Dr. Megan Reissman
University of Dayton
Dr. Timothy Reissman
University of Dayton
Roughly 4 million patients in the United States utilize weight bearing canes for balance and mobility aid. Unfortunately, 84% of these users have not been professionally trained and exhibit habits of incorrect usage which increases the probability of preventable falls. While physical therapists are trained to notice physical abnormalities of patient cane usage in clinical settings, long term community usage of these walking aids cannot currently be observed in an effective and noninvasive manner. Past research has shown that gait deterioration and weight transfer habits can indicate the need for a change in device. To provide a more detailed image of cane usage over a long period of time outside of a clinical setting, we have developed a clip on device for canes which monitors a patient's cane use habits over a 5 day period.
Our monitoring device consists of a suite of sensors including a 6 dimensional orientation sensor (IMU), a 50kg load cell, two distance sensors, a real time clock, and an SD card data logger. Data is logged from each sensor fifteen times a second, monitoring the cane orientation and providing the user and clinician with a quantitative record of the cane position throughout the gait cycle. To achieve a 5 day data collection period, a rechargeable battery, capable of 48 hours of continuous data collection, is placed distal to the cane handle to minimize perceived added weight. To conserve energy, the monitoring device is designed to turn off when no motion is detected, extending the operational usage before recharging is necessary.
We will present data detailing the validation of each sensor within the monitoring device, and how this is used to provide an accurate report of gait speed, frequency, and force. A representative day of cane use data will be presented. We will also present on our current data collections utilizing Xsens to record human movement in relationship to the cane data collected. This data will be used to train a neural network to identify characteristic gait patterns by relating the Xsens data to the data collected from the cane. Our future goals include creating a database of community cane use which we can utilize to reference common errors when they occur outside of a laboratory environment. Long term application of these datasets will utilize advanced statistical methods to calculate an increased chance of fall or predict when a change in device is appropriate.
Abstract ID: DESS2019-068
Optimizing a Brain-Machine-Interface to Create Artworks using EEG Technology
Ashley Martin
Carroll High School
The purpose of this project was to design a Brain Machine Interface (BMI) that uses Electroencephalography (EEG) technology and a robotic arm to paint via mental commands so that those with disabilities can express themselves creatively. The engineering goals included the prototype accomplishing all given tasks, each individual piece of the system functioning, and the painted points being within 5 cm of the 5 different target circles. Custom-programmed C++ software was developed. Utilized by the prototype, the program progresses through a series of stages based on user input. Another key aspect of the system is the Emotiv headset and its software. The user trained the software to recognize certain thoughts, or mental commands, through the use of the EEG headset. In the program, arrow keys are used to navigate through the software stages. In the completed prototype, the trained mental commands are bound to the keys so that a ¨push¨ thought is recognized as a down key press and a ¨pull¨ thought is recognized as an up key press. Data showed that the completed prototype was able to progress through the programmed stages and paint points within 5 cm of the targeted area. The points had a tendency to be along the edge of the circle or slightly above it. Since all engineering goals were met, the project provides a proof of concept for a new BMI that can be used to help those with physical disabilities express their artistic creativity in a method that was previously unavailable to them.
Manufacturing
Abstract ID: DESS2019-005
Thermal Modeling of Multi-Beam Additive Manufacturing
Rachel Evans
Wright State University
Joy Gockel
Wright State University
In additive manufacturing (AM), it is necessary to know the influence of processing parameters in order to have better control over the mechanical performance of the part. Laser powder bed fusion (LPBF) is a metal AM process in which thin layers of powdered material are selectively melted to create a three-dimensional structure. This manufacturing process is beneficial for many reasons; however, it is limited by the thermal solidification conditions achievable in the available processing parameter ranges for single-beam processing methods. Therefore, this work investigates the effect of multiple coordinated heat sources, which are used to strategically modify the melting and solidifying in the AM process. The addition of multiple heat sources has the potential to provide better control of the thermal conditions, thus having better control of the microstructure of the additively manufactured parts. To model this, an existing thermal model of the LPBF process has been modified to predict the thermal effects of multiple coordinated laser beams. The model uses the Rosenthal equation to calculate melt pool dimensions and thermal conditions throughout the LPBF process. Furthermore, the results of this model are used to determine influence of the distance between the coordinated laser beams. The predictive method used in this research provides insight to the effects of using multiple coordinated beams in LPBF, which is a necessary step in increasing the capabilities of the AM process.
Abstract ID: DESS2019-013
Application of the Theory of Critical Distances to Coupled Defects in Additive Manufacturing
Wesley Eidt
Wright State University
Craig Baudendistel
Wright State University
Joy Gockel
Wright State University
Numerous pathways exist for predicting the fatigue life of a part, and all of them become increasingly complex when the part in question is an additively manufactured (AM) one. While many strategies have displayed merit for analyzing traditionally built components, it is not yet determined which are best suited for translation to AM applications. One method for performing failure prediction analysis is known as the theory of critical distances (TCD). This work serves to create a predictive model for applying the TCD to additively manufactured parts. TCD formulas are equated to notch and pore formulas for use on the microscale. Additionally, AM parts often contain many competing defects, and the interactions between them can potentially be better understood using the TCD. In this work, stresses are aggregated and plotted according to the theory of critical distances, and these stress maps are compared with finite element analyses of the same geometry and loading scenarios. It is shown that the TCD frequently predicts failure from pairs of defects even when the individual defects alone are not abnormally severe. This could indicate that preventing surface-subsurface defect localization is more important to the fatigue life of an AM part than the quantity of defects present, potentially influencing strategies for deciding parameters of an additive build.
Abstract ID: DESS2019-022
Direct Printing of nanosilver (UTDAg) ink on Kapton Substrate using 〖Jetlab〗^® 4xl
Aamir Hamad
Wright State University
Dr. Ahsan Mian
Wright State University
Inkjet printing is an additive manufacturing (AM) technique dealing with liquid phase material by dropping on demand (DOD) various kinds of liquids or conductive nanoparticle inks. This technology has been used in producing everyday smart printed electronics such as conformal antennas (planner and non-planar antennas), sensors, conductors, and solar cells mounted on flexible and rigid substrates. The performance of printed electronics strongly depends on the behavior of generated droplet from the nozzle and the behavior of ejected droplet on the substrate. In this work, the effect of jetting parameters on generated droplet will be studied to generate optimal waveform voltage (bipolar) that drives the piezoelectrical actuator to generate ideal droplet and the effect of high stage velocities on the behavior of ejected ideal droplet on Kapton substrate will be investigated using nanosilver ink (UTDAg). Lines will be printed at different waveforms, stage velocities and drop spacings using fly mode printing with burst. The profile thickness and the resistance of printed lines will be measured at different locations and curing temperature.
Abstract ID: DESS2019-027
In Situ Defect Detection using a Multi Sensor Approach in Laser Powder Bed Fusion
Andrew Drieling
Wright State University
Joe Walker
Wright State University
John Middendorf
Universal Technology Corp.
Nathan Klingbeil
Wright State University
Joy Gockel
Wright State University
Additive Manufacturing (AM) provides a way to create parts that would be extremely difficult or impossible with conventional manufacturing processes. However, AM also introduces defects, which are detrimental to the mechanical performance. Some defects can be identified using non-destructive post-processing inspection and testing. If defects exceed predetermined mechanical limitations, they cause time and material waste. Other defects cannot be detected non-destructively, potentially initiating unexpected failure. In situ monitoring using a multi-sensor approach for fabrication of Alloy 718 on a laser powder bed fusion system is performed in this work. Single bead and simple geometry builds with varied processing parameters are performed to detect variances in sensor outputs and to correlate to part parameters and properties. Using multiple sensors allows for a more thorough recording of the building process and comparisons between different data collection types. A multi-sensor approach can provide near real-time monitoring, allowing defects to be predicted, and potentially corrected before the completion of the part, saving time and resources.
Materials
Abstract ID: DESS2019-020
Role of carbide coarsening in mechanical properties of Nickel Alloy 718 manufactured by laser beam powder bed fusion
David Newell
Air Force Institute of Technology
Anthony N. Palazotto
Air Force Institute of Technology
Ryan P. O'Hara
Air Force Institute of Technology
Researchers are currently pursuing modified heat treatment methods to overcome the directional grain patterns found in powder-bed fused metal alloys. Existing heat treatments for Nickel Alloy 718 (IN718) were developed for wrought and cast productions and carry some limitations from these manufacturing methods. By using laser beam powder bed fusion (LB-PBF), the microstructure of the fabricated parts presents a different morphology than the wrought and cast counterparts. A higher solution treatment is necessary to recrystallize the grain structure and generate equiaxed grains. This results in a decrease in the anisotropy of the tensile properties but also causes a severe decrease in the directional creep properties of the same material. This research investigates the influence of niobium carbides on the creep behavior of vertically and horizontally fabricated LB-PBF IN718 creep specimens. The microstructure and performance of the existing solution treatment of 1010 °C, 1 h with aging is compared against a modified supersolvus solution treatment of 1160 °C, 4 h with aging.
Abstract ID: DESS2019-034
Toughened Zirconia as Dental Implant Material
Abdullah Al Saad
University of Toledo
Prabaha Sikder
University of Toledo
Devin Dinh Ta
University of Toledo
Sarit B Bhaduri
University of Toledo
The demand for developing biomaterials with the optimum mechanical and corrosion properties for dental prostheses is increasing. Zirconia, a ceramic material, is getting more attention for its favorable physical, biological and corrosion resistance attributes. Zirconia demonstrates polymorphism and upon cooling, it exhibits martensitic transformation. This results in the uncontrolled expansion of the volume of about 5% and makes it suitable for clinical load bearing application. However, it can result in catastrophic material failure. By introducing a dopant into its crystal structure, this transformation can be stabilized and controlled as a result the material becomes fracture-resistant and reliable under load-bearing situations. Ceria, amidst other dopants, is isovalent with zirconia and does not create any vacancies upon substitution. In this study, 20 mol % Ceria Stabilized Zirconia (20Ce-TZP) has been characterized upon sintering at three different temperatures such as 1400℃, 1500℃ and 1600℃ for 2 hours. X-ray Diffraction Pattern and Raman spectroscopy confirmed the tetragonal phase and while analyzing the Scanning electron microscopy images, the increase in grain size is confirmed with sintering temperature. After autoclaving, the tetragonal phase was confirmed with a very small change in the grain size distribution and surface roughness which means Ce-TZP is not prone to LTD. Besides, the bio-inertness of Ce-TZP was suggested from the biocompatibility test. In conclusion, analyzing all the results, Ce-TZP can be considered for further investigation as a prospective dental implant material.
Abstract ID: DESS2019-072
A Simplified Investigation into Fatigue Viability of Additively Manufactured IN-718
Austin Schoening
Wright State University
Luke Sheridan
Air Force Research Laboratory
Onome Scott-Emuakpor
Air Force Research Laboratory
Tommy George
Air Force Research Laboratory
Material characterization by way of fatigue testing is a common practice in materials research. This research is then applied to many different engineered devices, one of which includes the gas turbine engine. In the environment of a turbine engine the fatigue life plays a critical role in the design, operation, and maintenance of the engine. This study uses the fatigue life of two different metals (Titanium 6Al-4V, Aluminum 6061-T6) with different cross-sectional measurements to assess the fatigue viability of another material (additively manufactured Inconel 718, or AM IN-718). The assessment of AM IN-718 fatigue life is done by normalizing all the fatigue data against respective ultimate tensile strength results obtained from monotonic tests. The data from two different gage section types for Titanium (Ti) 6Al-4V and Aluminum (Al) 6061-T6 specimens show that the comparability of the normalized fatigue results fit within a 99% prediction interval. The viability assessment of IN-718 highlighted concerns in the material integrity, and this finding, guided by the normalized fatigue data of Ti 6Al-4V and Al 6061-T6, led to the identification of flaws which were artifacts of poor AM process controls.
Other
Abstract ID: DESS2019-028
Policy and Geopolitical Implications of Launch-on-Demand Capabilities
Robert Bettinger
Air Force Institute of Technology
Liberty Shockley
Air Force Institute of Technology
With the growing capability and frequency of spacecraft launch operations, as well as the accelerated research and development of high-altitude hypersonic vehicles, the prospect of rocket-based cargo mobility requires new legal and policy strategies to contend the implications of projecting air power to any global theater within one hour via a vehicle based in the continental U.S. This research will explore the challenges posed by a sub-orbital “launch-on-demand” capability by not only air and space law, but also national policy and geopolitical perceptions. Citing contemporary legal and process-based requirements for space launch and atmospheric reentry operations, this research will advocate an evolution of national policy to enable a launch-on-demand capability to deliver personnel and cargo to contested theaters of operation in support of U.S. core competencies of global reach and rapid global mobility.
For the purposes of this research, a notional launch-on-demand system is not based on any existing launch vehicle concept, but it is assumed to have a final stage that lands safely in a designated location. The vehicle is restricted to two stages based on its intercontinental mobility mission, in which injection into low Earth or higher Earth orbits with the use of three, or possibly four stages, is not required. In terms of classification, the vehicle is a vertical-takeoff, vertical-landing (VTVL) system. The vehicle’s first stage represents a “boost” or ascent stage that will be reusable rather than expendable. The vehicle’s second stage, containing of the passenger and/or cargo payload compartment, features minimal aerodynamic properties, thereby restricting the landing profile to a powered vertical descent instead of an unpowered glide (e.g., Space Shuttle). The non-commercial implementation of a launch-on-demand system could occur in one of two modes: (1) emergency mobility; or, (2) routine mobility. For the “emergency mobility” mode, vehicle operation only occur in the event of a geopolitical situation necessitating a rapid deployment of personnel and/or cargo to a specified area of operation or theater.
Abstract ID: DESS2019-057
Chemical and Physical Effects on Lean Blowout in a Single-Cup Swirl-Stabilized Combustor
Jennifer Colborn
University of Dayton
Joshua Heyne
University of Dayton
Tyler Hendershott, Scott Stouffer
University of Dayton Research Institute
Edwin Corporan
Air Force Research Laboratory
Abstract awaiting public release.
Abstract ID: DESS2019-060
Development of Universal Autopilot Translator
Nicholas Degroote
University of Cincinnati
Evan Barnes
University of Cincinnati
Anthony Lamping
University of Cincinnati
Dr. Kelly Cohen
University of Cincinnati
As the world is primed for a large increase in the number of autonomous systems, new solutions are
required to address the collision avoidance and air traffic management challenges for both manned and unmanned craft. In response, the University of Cincinnati has been developing RouteMASTER—an air
traffic management digital infrastructure that allows effective operation of vehicles flying beyond visual line of sight (BVLOS) and helps integrate these vehicles into the national airspace. Such a system would be useful in a variety of situations, such as during an emergency response scenario. As part of RouteMASTER, a universal autopilot translator (UAT) has been developed with the SkyVision system, located at Springfield-Beckley Municipal Airport in association with the Ohio Federal Research Network. The UAT addresses some of the limitations of the SkyVision system by providing it the ability to track small UAS platforms that would not normally be picked up by radar. Most modern communication
systems for UAS feature a single ground control station (GCS) which can display telemetry data from one or more systems but is often limited to vehicles using the same flight controller. The UAT opens the door for UAS operations across different platforms by converting telemetry data from each autopilot to a common air traffic control protocol. Currently, the UAT has been tested with SkyVision using MAVLink (Pixhawk), Piccolo, and MicroPilot autopilots.
Abstract ID: DESS2019-070
Path Planning and Obstacle Avoidance with UAV's
Matthew Terry
University of Cincinnati
Anthony Lamping
University of Cincinnati
Bryan Brown
University of Cincinnati
Justin Ouwerkerk
University of Cincinnati
Dr. Kelly Cohen
University of Cincinnati
As the demand and use of unmanned aerial vehicles(UAVs) grow, so does the need for collision prevention and obstacle avoidance. Researchers at the University of Cincinnati have been working on formulating innovative and collaborative strategies to mitigate these challenges. The first step to achieving a solution was to test an obstacle avoidance system inside a simulated environment. The simulation environment included Robot Operating System (ROS), Gazebo, and PX4 Firmware. Inside the simulation, a UAV, LiDAR sensor, and a world with obstacles were created and used to develop and test different avoidance solutions. Following successful simulations, work began to fuse a 360 LiDAR to a custom UAV platform such that we could test our proposed solution in real-world environments. The fused UAV system combined custom endurance UAV platforms with off-the-shelf LiDAR components and incorporated the PX4 collision avoidance implementation. The current systems specifications are capable of detecting and preventing collisions during manually piloted flight and avoiding obstacles while in autonomous flight. Currently, research is being performed on how to use additional sensing devices to increase the spatial awareness about various axes of the vehicles, thus providing a higher fidelity avoidance base in the entire environment.
Abstract ID: DESS2019-071
A Near-Real-Time Near-Optimal Shortest Path Solution for an Unmanned Aerial System (UAS) in a Highly Constrained Environment
Kyle Matissek
Air Force Institute of Technology
Dr. Richard G. Cobb, Dr. David R. Jacques
Air Force Institute of Technology
Dr. David J. Grymin
Air Force Research Laboratory
Lt Col Michael D. Zollars
Air Force Life Cycle Management Center
A current challenge in path planning is the ability to efficiently calculate a near-optimum path solution to a highly constrained problem in near-real-time. In addition, computing performance on a Small UAS is typically limited due to size and weight restrictions. This method determines a solution quickly by first mapping a highly constrained three-dimensional environment to a two-dimensional weighted node surface in which the weighting represents the terrain gradient and vehicle performance. The surface is then discretized into triangles which are sized depending on the vehicle maneuverability and the terrain complexity. The shortest feasible path between the nodes of the two-dimensional triangulated surface is determined by using an A* algorithm. An optimal path is then chosen through the unconstrained corridor to yield a quick near-optimal path solution in three-dimensional space. This technique requires prior knowledge of the terrain and vehicle performance. The cost to traverse each increment of the map is agnostic to a change in the start state and can be precalculated once the goal state is known. This proposed method allows for a quick path solution to any start state.
Renewable and Clean Energy
Abstract ID: DESS2019-043
Improving Cooling Energy-Efficiency: A Case Study of Kettering Labs
Andrea Mott
University of Dayton
Patrick Fitzgerald, Abinesh Selvacanabady, Amanda Alvarado
University of Dayton
Cooling is an important part of many buildings. Energy efficiency opportunities can be identified by sequentially investigating the end uses, the distribution, and the primary energy conversion components of a cooling system. This research identifies seven recommendations to improve cooling efficiency at the University of Dayton Kettering Laboratories building. The approach involves collecting and analyzing data from the building’s automation system and separately installed current transducers. The recommendations are: 1) increasing the temperature of the water leaving the chiller, 2) eliminating flow through inactive chillers, 3) staging chillers based on part-load efficiencies, 4) eliminating an unnecessary bypass loop, 5) improving control of variable speed chilled water pumping, 6) staging cooling towers to provide lowest temperature condensate water possible, and 7) eliminating condensate water flow through inactive chillers. Best practices and methods identified in this case study will be applied to buildings across campus.
Abstract ID: DESS2019-047
An Investigation into the Potential for Fusel Alcohol Mixtures from Biomass Derived Feedstocks to Improve the Efficiency of Gasoline Blends
Lily Behnke
University of Dayton
Eric Monroe, Ryan W. Davis, Anthe George
Sandia National Laboratories
Biofuels additives to gasoline are a promising way to both reduce the greenhouse gas emissions associated with petroleum products and improve fuel properties such as octane. Fusel alcohol mixtures generated from complex biomass feedstocks (including waste streams) have are a promising 2nd generation biofuel with desirable fuel properties relevant to spark ignition engines. In this work experimentally generated fusel blends composed of isobutanol, 3-methyl-1-butanol, 2-methyl 1 butanol, 2-phenyl ethanol, and ethanol, are shown to demonstrate comparable efficiency gain, higher energy density, and lower vapor pressure than ethanol when blended into commercial gasolines. Since the composition of the fusel alcohol blends can be tuned based on the desired properties of the mixture, two approaches, based on experimental data and kinetic modeling, were utilized to computationally predict the properties of a multitude of fusel blends at various blending percentages with gasoline to determine the optimal blend composition for each property of interest. The results of the two approaches were compared and analyzed on the basis of the total efficiency increase of the blended fuel, while also considering properties beyond those contributing to efficiency. These data suggest that fusel alcohol blends are a promising biofuel candidate for increasing engine efficiency without sacrificing high energy content or causing disruptions to the existing refining infrastructure.
Abstract ID: DESS2019-062
Automated Residential Energy Audits and Savings Measurements Using a Smart WiFi Thermostat and Data Mining Approach
Abdulrahman M Alanezi
University of Dayton
Kefan Huang
University of Dayton
Residential buildings account for 40% of the US energy demand. The heating and cooling systems constitute the biggest portion of this (about 31% of the total).
WiFi thermostats such as Google’s Nest, Ecobee, and Emerson Climate Technologies’ Sensi have become commonplace in residential buildings. In 2017 more than 82 million smart thermostats were in use in North America according to a study by Berg Insight. Also, the same study projected that more than half (51%) of North America homes would be smart homes by 2022. Overall, smart WiFi thermostats have enabled the companies providing these devices to residences to have access to overwhelming amounts of data regarding cooling and heating patterns. Most prominently, this data has been used to recognize residents heating and cooling patterns and, through geofencing technologies, to recognize when residents have left their homes. As a result, the thermostat manufacturers have been able to help residents reduce energy consumption. For example, Google’s Nest thermostats have been recognized as having help realized savings equal to about 10%-12% of heating usage and electric savings equal to about 15% of cooling usage on average in single-family residences.
This research is predicated on the fact that there remain far greater opportunities to leverage the thermostat data to save energy, particularly when such data is combined with historical weather data, building geometry characteristics, historical energy consumption data, building characteristics (R-values), and energy characteristic data. Moreover, when such types of data are combined, there is strong potential to leverage machine learning to analyze such data in order to extract actionable information.
Specifically, this research aims to demonstrate the value of the thermostat, in combination with the other data described above, in effectively rendering virtual energy audits of residences. To do so, thermostat data is collected from two groups of single-family residences – one associated with approximately 450 Midwestern U.S. university and another associated with a group of approximately 84 low income subsidized homes in an East Dayton Ohio neighborhood. These homes provide a rich spectrum of energy characteristics needed to develop a robust data-based predictor of the energy characteristics in these homes.
Abstract ID: DESS2019-067
Performance and Proximity Investigations on Small Scale Lensed Turbines
Neal Novotny
University of Dayton
Sidaard Gunasekaran
University of Dayton
Lensed turbines increase the power output of a conventional turbine by a factor of 2 to 5. The changes in performance of a 1ft diameter lensed turbine is experimentally investigated at the University of Dayton Low Speed Wind Tunnel (UD-LSWT) along with proximity tests. This investigation is a precursor for developing a novel energy harvesting technique using grid lensed turbines at much lower altitudes than a conventional turbine. An iterative design/optimization process is presented for the development of the 1ft wind-lens turbines used in this investigation. This process was necessary to maximize the power output given the low Reynolds number conditions. The effect of atmospheric boundary layer and wind direction changes is ignored in the current analysis. Force-based testing was conducted on circular disks (surrogate to lensed turbines) in a 1-D, 2-D and 3-D grid to gain insight into lens-lens aerodynamic interactions. The results of the side-by-side (1D) disk testing indicated a favorable increase in coefficient of drag at a normalized lens-lens distance of 0.25 x/D. Depth testing of disks indicated that the optimum downstream placement of turbine is 6 diameters from the upstream disk.
Abstract ID: DESS2019-073
Machine-Learning Enabled Accurate Prediction of Energy Savings from Thermostat Setpoint Schedule Changes Using Smart WiFi Thermostat Data
Justin Ehren
University of Dayton
Kefan Huang
University of Dayton
Lu Hao
University of Dayton
A machine learning algorithm has been developed to predict the daily cooling, heating, and fan usage for residential buildings in Dayton, OH. Residential buildings account for 40% of the US energy demand, while the heating and cooling systems constitute the biggest portion of this (about 31% of the total). WiFi thermostats such as Emerson Climate Technologies’ Sensi are capable of recording when your heating/cooling equipment turns on and off as well as your instantaneous setpoint temperature in real time. In combination with the Sensi thermostat data, local weather data and house characteristics can be used to identify realistic energy/financial saving from reducing your equipment’s heating and cooling. More specifically, the algorithm enables the data described above to accurately predict the daily percent cooling/heating if a different setpoint schedule was in effect. The research was limited to low income subsidized homes in an East Dayton Ohio neighborhood; however, a similar algorithm could be developed to any house regardless of region.
Structures / Solid Mechanics
Abstract ID: DESS2019-008
3D-Printed NinjaTek Cheetah: Influence of Process Parameters on Tensile Properties
Brad Hripko
University of Dayton
London Ayton, Will Parker, Timothy Reissman, Robert Lowe
University of Dayton
Soft thermoplastic elastomers with over 500% strain have recently become available for use with desktop FDM printers. With exciting functional applications, the effects of FDM process parameters on final mechanical properties must be well defined. This study presents a full-factorial design of experiments that investigates the impact of extrusion temperature and layer height on the quasi-static (0.1 s-1) tensile properties of the commercial elastomer NinjaTek Cheetah. Statistical results, based on an analysis of variance, indicate that increasing extrusion temperature exhibits significance for all three properties: a 12% increase in initial Young’s modulus, a 14% decrease in ultimate tensile strength, and a 5% increase in fracture strain. In contrast, increasing layer thickness exhibits significance only for initial Young’s modulus, with a 5% increase. Dynamic tension testing performed at elevated strain rates (1 s-1, 10 s-1 and 100 s-1) indicates strong rate dependence, with lower fracture strains observed at higher strain rates.
Abstract ID: DESS2019-011
The Helical Sphere – A Near-Vacuum Lighter-than-Air Envelope
Ruben Adorno
Air Force Institute of Technology
Dr. Anthony N. Palazotto
Air Force Institute of Technology
Lighter-than-Air (LTA) systems have been developed throughout history for a multitude of applications, taking several forms and names. Airships, aerostats, blimps, and balloons are all part of this family of systems, which use Archimedes principle to achieve neutral and positive buoyancy in air by replacing an air volume with LTA gases, commonly known as 'lifting gasses'. Helium being the most common nowadays, these lifting gases are contained within a LTA system's envelope, i.e. the outermost membrane acting as the barrier between the lifting gas and the atmosphere. To date, LTA systems have only been realized with the use of lifting gases, as these stiffen the otherwise gossamer envelopes, allowing them to sustain the pressure brought by the displaced air. The compliance of these structures is driven by the large density disparity between air and materials exhibiting low air permeability, which in turn tends to result in highly void geometries containing dimensions in the order of 1/10,000 of their characteristic length. These small dimensionalities can exhibit low or virtually non-existent in-plane bending stiffness, which, combined with the need for structural integrity to achieve positive buoyancy, make structural stability a primary design consideration for vacuum LTA envelope design.
With this in mind, the Helical Sphere design is introduced and its potential as a vacuum LTA envelope argued on the basis of the stress-control inherent to the geometric configuration, enabling stress limits consistent with commercially-available polymeric membranes. Design characterization is developed sequentially, starting from preliminary studies that established initial feasibility, and continuing with structural finite element analysis using the SIMULIA Abaqus package.
Abstract ID: DESS2019-014
Analysis of the Behavior of a 3D Printed Celestial Icosahedron Structure Under Compressive Loading
Kevin Greenoe
Air Force Institute of Technology
Dr. Anthony Palazotto
Air Force Institute of Technology
The Vacuum Lighter Than Air Vehicle (VLTAV) is a new and innovative platform within the Unmanned Aerial Vehicle (UAV) design space. It’s unique capability and design provides a method for the safe measurement of environmental conditions within a hostile atmosphere. Hostile atmospheres of interest include those in which a violent storm is active or corrosive gases are present. The development of a VLTAV relies on a unique structural design that enables the vehicle to be light enough to float in air and strong enough to withstand atmospheric conditions such as pressure loading. The current design is comprised of an internal vacuum packed within a celestial icosahedron structure. It is composed of nine intersecting rings, offset at different angles from one another, to form an overall spherical profile. The rings are composed of solid tubes with circular cross-sections. Optimal sizing of the circular tubes is critical to ensure maximum strength while maintaining a low overall weight.
A thorough analysis of the mechanical behavior and instability of the structure under compressive loading is required to ensure the design is suitable for real-world environments. The structure will be manufactured with ULTEM 9085, an aviation-grade thermoplastic, using fused deposition modeling (FDM). This material exhibits anisotropy, which introduces complexity in accurately modeling the design’s behavior due to loading. The mechanical properties are dependent on the build orientation and direction. The celestial features a unique geometry and requires rings to be printed in various orientations. Therefore, the mechanical properties, such as the modulus of elasticity, will vary throughout the build. The degree of variation is of interest along with a means to accurately characterize the properties throughout the design. The same additive manufacturing process and instability testing has been conducted for an icosahedron frame in previous research. The icosahedron was manufactured out of VeroBlue material. It features straight beam members as opposed to the celestial’s curved members. Although the shape and material differs, the overall experimental process will aim to closely follow the process conducted for the icosahedron. Ultimately, the purpose of the analysis is to measure the response of a 0.2032m-diameter, 3D printed celestial under sea-level pressure loading.
The structure is to be wrapped with a membrane to protect the internal vacuum and measuring devices onboard. The exact material and wrapping method is to be investigated. The entire frame, with an appropriate membrane, will be evaluated under sea-level pressure using the finite element analysis (FEA) via Abaqus. The deformed structure, generated from FEA, will be utilized for aerodynamic analysis using computational fluid dynamics (CFD). Finally, the development of an exhaust mechanism is required to evacuate the air for the creation of an internal vacuum.
Abstract ID: DESS2019-019
Mechanical Properties of Additively Manufactured Periodic Cellular Structures
Derek Spear
Air Force Institute of Technology
Dr. Anthony Palazotto
Air Force Institute of Technology
Advances in new materials have the potential to drastically change how engineers view and address problems. Imagine a highly porous metal cellular structure or foam that is significantly lighter than its solid counterpart, but provides improved energy absorption and dispersion characteristics. A major effort is being developed in which the impact effects of a projectile incorporating advanced materials is carried out and compared to a baseline traditional projectile. The initial stage of this effort is, through experimentation, to evaluate the properties of various metal foam-like structures (pseudo-foams), looking at how the porosity level and void structure affect the global material properties. Additive manufacturing techniques will be utilized in constructing both open and closed cell manufactured metal pseudo-foams. These properties will be compared to the baseline solid material to determine change in specific stiffness, or the stiffness to weight ratio, along with a comparison between the structural designs of energy absorption.
Abstract ID: DESS2019-024
Optimization-Based Laser Shock Pressure Impulse Determination
Colin Engebretsen
Air Force Institute of Technology
Anthony Plazotto
Air Force Institute of Technology
Kristina Langer
Air Force Research Laboratory
Laser shock peening (LSP) is a method of work hardening by laser initiated pressure impulse. Finite element modeling of LSP currently relies on an assumed temporospatial pressure impulse shape because the direct, in situ measurement of the ~200 nanosecond event is not possible. Optimization methods were used to match measured residual stress and surface displacement data with a "best-fit" pressure impulse for aluminum specimens. This pressure impulse shape was then validated on titanium. This process could pave the way for the cataloging of pressure impulse shapes for given laser system settings in order to more accurately predict imparted residual stresses.
Abstract ID: DESS2019-042
Progressive Failure in bolted Hybrid Composite Joints
John Brewer
Air Force Institute of Technology
Dr. Anthony Palazotto
Air Force Institute of Technology
John Feie
Air Force Research Laboratory
Michael Gran
Air Force Research Laboratory
Composite materials are commonly used in aerospace and automotive applications due to their high strength to weight ratio, stiffness, and ease of manufacturing in complex geometries. However, joining composite materials to other structures can prove problematic. Bonding agents may have difficult failure predictions, so many manufacturers employ fasteners for more predictable joints. Furthermore, fasteners are often required in joints that need to be serviceable or provide access to other components. Despite their common use, fasteners cause stress concentrations, which may lead to localized failures and crack initiation.
Common practices to combat failure near holes in composite materials include adding layers of composite at fastener sites, known as pad‐ups, and adding metallic inclusions between plies. While these do improve strength locally, composite pad‐ups still suffer from the same failure characteristics as the primary composite, and both methods locally thicken the material and create potential for defects and resin concentrations.
The intent of this research program is to include metallic foil insertions between and in place of composite layers to distribute the bearing load evenly through the structure with more predictable failure. Specifically, the materials employed in this study are IM7/977-3 pre-impregnated carbon fiber and full-hard 301 stainless steel. In contrast to other Fiber Metal Laminates, such as GLARE (GLass Aluminum REinforced-fiberglass/aluminum hybrid), the hybrid considered in this study only includes metal near stress concentrations such as bolts.
The focus of this researcher is to characterize the progressive failure nature of this hybrid composite material as compared to a control material (pure composite). Alongside this experimentation, the goal is to evaluate the efficacy of Abaqus finite element software in modelling this hybrid material in bearing loading and failure.
To date, engineers have performed preliminary Mode I and Mode II delamination tests and bearing tests on hybrid composite materials. The Mode I/II experimentation proved the requirement for a co-cured adhesive between the composite and metallic layers. Preliminary bearing tests proved the efficacy of the hybrid design. In-depth progressive failure experimentation is in progress and preliminary finite element models are complete. More complex modelling is underway.
This work will present completed efforts in design and manufacturing of the hybrid composite and the statistically designed experimental procedure. Bearing testing in single-shear, double-shear and countersunk configurations will be discussed in both quasi-static and fatigue loading scenarios. Test results and analysis from completed to date will be discussed. Finally, finite element development and progress will be discussed and compared with experimental results.
Derived from documents cleared for public release: Case Numbers 88ABW-2018-2972, 88ABW-2019-3162
Abstract ID: DESS2019-044
Instrumented Impact Behavior of ULTEM 9085 Panels Produced by Additive Manufacturing
Alex Elsbrock
University of Dayton
Robert L. Lowe
University of Dayton
Thomas J. Whitney
University of Dayton
Instrumented impact testing was conducted on ULTEM 9085 panels fabricated using fused deposition modeling (FDM). Panels of various thickness, print direction, and environmental conditioning were examined. Statistically significant differences in peak load and absorbed energy were noted among the tested panels. Finite-element simulations of the impact tests were performed in LS-DYNA using an elastic-plastic constitutive model, calibrated to quasi-static tensile data. Numerical simulations showed good agreement with experimental load-displacement curves up to peak load. Our results suggest that continuum constitutive models and commercial finite-element analysis tools show promise for first-order predictions of the pre-failure mechanical performance of as-built FDM components.
Abstract ID: DESS2019-059
An Efficient Iterative Approach for Determining the Post-Necking True Stress-Strain Response of Aerospace Metals
Luke Hoover
University of Dayton
Christopher A. Negri, Robert L. Lowe
University of Dayton
Jeremiah T. Hammer, Jeremy D. Seidt, Amos Gilat
The Ohio State University
To numerically simulate the plastic deformation of aerospace metals during extreme events (e.g.,
turbine engine blade-out/rotor-burst events and automotive crashworthiness assessment),
accurate experimental knowledge of the metal’s hardening behavior at large strains is requisite.
Tensile tests on thin (plane stress) specimens are frequently used for this purpose, with the
metal’s large-strain plasticity ultimately captured by an equivalent true stress vs. equivalent true
plastic strain curve. It is now well known that if axial strain is measured using an extensometer
(either physical or virtual), the equivalent true stress-strain curve is valid only up to the onset of
diffuse necking, when the strain field heterogeneously localizes in the specimen gage. A number
of approaches have been proposed to correct the post-necking strain hardening response. Perhaps the most widely used technique involves inputting a suite of candidate post-necking true
stress-strain curves into finite-element software. A tensile test simulation is run for each
candidate curve, and the curve that produces the best agreement between simulation and
experiment is adopted. In this talk, a novel variation of this iterative approach is presented that
addresses some of its key deficiencies. Notably, we use local/pointwise in-plane Hencky (true)
strain data from digital image correlation to generate an optimized initial guess for the iterative
simulation process, resulting in an efficient and computationally inexpensive post-necking
correction procedure. Our approach is successfully demonstrated using experimental data for
two aerospace metals, Ti-6Al-4V titanium alloy and 310 stainless steel.
Undergraduate Project
Abstract ID: DESS2019-018
Effect of Airfoil-Preserved Undulations on Wing Performance
Faith Loughnane
University of Dayton
Sidaard Gunasekaran, Rachael Supina, Michael Mongin
University of Dayton
This bioinspired study investigates the effect of the humpback whale's leading edge tubercles as applied to wings with airfoil-preserved undulations, or uneven surface contours. These undulations are hypothesized to affect the spanwise flow, thereby affecting the induced drag, roll-up of the wingtip vortex, and the parasite drag of the wing. Sensitivity study was done on the number on undulations along the span (6, 9, and 12) and undulation placement (leading edge, trailing edge, and both leading edge and trailing edge) by performing force-based experiments and Particle Image Velocimetry (PIV) at the University of Dayton Low-Speed Wind Tunnel (UD-LSWT).
Abstract ID: DESS2019-063
UCAV at 2019 AUVSI SUAS Competition
Austin Wessels
University of Cincinnati
Nicholas DeGroote, Nicholas Little, Dr. Kelly Cohen
University of Cincinnati
Every year, the Autonomous Unmanned Vehicle Systems International (AUVSI) hosts a competition to foster interest in Unmanned Aerial Systems (UAS) and UAS technologies. The University of Cincinnati Aerial Vehicles (UCAV) team designed a custom Unmanned Aerial Vehicle (UAV) for the 2019 AUVSI SUAS (Student Unmanned Aerial Systems) competition. Using prior knowledge gained from previous competitions, the UAV in tandem with the Ground Control Station (GCS) were designed to accomplish all the tasks required by AUVSI. Following a crawl, walk, run approach, the GCS was tested in simulation, and the UAV was flight tested. Once the GCS performed in simulation and the flight controller of the UAV was properly tuned, the subsystems required by the competition were implemented individually. Finally, the entire UAS was tested to validate its performance and test reliability to ensure safety for the competition. The UCAV team participated in the 2019 AUVSI SUAS competition and placed fourth overall out of seventy-five registered teams.
Undergraduate Student Presentation Competition
Abstract ID: DESS2019-002
Additive Manufacturing: Porosity in Laser Powder Bed Fusion
Sabrina D'alesandro
Wright State University
Andrew Harvey
Wright State University
Joy Gokel
Wright State University
Additive manufacturing is shaping the manufacturing world through simplistic household printers’ to more complex metal printers used for a variety of applications. Specifically, laser powder bed fusion (LPBF) is an additive process that deposits metal powder over the build plate and melts it with a laser in the shape of the build part. In order to make LPBF more efficient with higher quality material, an experiment was done using in-situ sensors to observe the LPBF process as it printed nickel super-alloy 718. The focus of this experiment was to observe defects in the printing process such as pores and inclusions. The printer printed two separate parts, one small coupon and one larger coupon. Changing the geometry for LPBF parts will create different outcomes microscopically, because of the different thermal histories, that result in different defect characteristics. The pores in the material were analyzed in order to understand the relationship between the geometry and the number, size and shape of pores created. The implications of this research highlight the impact to the structural integrity of printed LPBF parts, which will help ensure that future materials have less defects, are stronger, and have a higher level of quality.
Abstract ID: DESS2019-015
Identifying a Heat Transfer Ratio for Steady State High-Speed Sliding Contact
Elloria Shaw
Wright State University
Dr. William Baker
Air Force Institute of Technology
Dr. Anthony Palazotto
Air Force Institute of Technology
In sliding contact, the thermal flux is generated due to friction flow into both objects at an unknown ratio. In order to better understand the thermal effects of sliding contact, this research examines the effects of steady state sliding contact on the distribution of thermal energy between two surfaces. The thermal distribution of an object sliding down a rigid rail at constant velocity is evaluated by deploying an iterative finite difference approximation. The gradient term at the interface between the two surfaces characterizes this thermal flux. This methodology allows for the numerical determination of heat partitioning between the two surfaces.
Abstract ID: DESS2019-029
Profiling Optical Turbulence using Dual-Camera Time Lapse Imagery
Benjamin Wilson
Air Force Institute of Technology
Dr. Santasri Bose-Pillai
Air Force Institute of Technology
Dr. Jack McCrae
Air Force Institute of Technology
For effective turbulence compensation, especially in highly anisoplanatic scenarios, it is useful to know the turbulence distribution along a path for performance assessment of optical systems operating in real environments and for designing systems to mitigate turbulence effects. Irradiance-based techniques suffer from saturation when profiling turbulence over long ranges and hence alternate techniques are currently being explored. We have developed a technique to measure the distribution of turbulence along an experimental path using the time-lapse imagery of a target from multiple cameras. The approach uses an LED array as target on one end of the path and two cameras separated by a few centimeters at the other end of the path imaging the LED board. By measuring the variances of the difference in wavefront tilts sensed by a single camera and between the two cameras due to a pair of LEDs with varying separations, turbulence information along the path can be extracted. The mathematical framework is discussed and the technique has been applied on experimental data collected over a 511m horizontal path, half of which is concrete and half grass. The estimates were compared to derived measurements from 3D Sonic Anemometers placed every 100m along the path. A potentially significant advantage of the method is that it is phase based, and hence can be applied over longer paths. The ultimate goal of this work is to profile turbulence remotely from a single site using targets of opportunity. Imaging elevated targets over slant paths will help in better understanding how turbulence varies with altitude in the surface layer.
Abstract ID: DESS2019-045
Development of a Computational Framework for Estimating Knee Joint Contact Forces in Walking and Running
Sean Kapp
University of Dayton
Dr. Joaquin Barrios
University of Dayton
Dr. Allison Kinney
University of Dayton
Running is a common form of exercise and walking is a simple form of locomotion that many individuals perform. Both running and walking require coordination of muscle forces within the body. As a result of the muscle forces, the joints in the body experience loading, which fluctuates with changes in motion. One potential method to alter joint loading is changing foot strike pattern. When running, individuals typically choose either a rearfoot strike (i.e., contacting the ground with the heel first) or non-rearfoot strike (i.e., contacting the ground with the ball of the foot first) pattern. Runners often develop knee pain and knee injuries due to the high impact or ground reaction forces experienced during rearfoot strike running. Previous research has shown that a non-rearfoot strike pattern influences the biomechanics of running, with reductions in the vertical ground reaction forces and in the moments at the knee joint. Also, it may be helpful to understand the compressive knee joint forces that may have adverse effects on tissues in the knee such as articular cartilage or menisci. However, few studies have compared the influence of strike pattern on knee joint contact loading because these forces are physically impossible to obtain in vivo. Given the complexity of the motion of running and the forces involved, is it useful to apply biomechanical modeling and simulation to study the underlying mechanical aspects of walking and running. The OpenSim modeling software provides many of the resources required to create and test such actions. For this preliminary study, OpenSim was used to analyze biomechanical data collected from one female performing walking and running with both rearfoot and non-rearfoot strike patterns. A generic musculoskeletal model was scaled to patient specifications, and using motion capture marker data and ground reaction force data, kinematics of the captured movement and individual muscle forces were calculated with OpenSim's Static Optimization Tool. Finally, the joint reaction forces in the knee joint were calculated and compared to previous estimates of knee joint loading. Data from rearfoot strike running trials have produced promising results that agree with the literature, with a peak compressive force of 4,500 Newtons during stance phase. Further research includes the processing of forefoot strike running and both rear- and forefoot strike walking data.
Abstract ID: DESS2019-061
Wing Performance Changes Due to Trailing Edge Extensions
Rachael Supina
University of Dayton
Michael Mongin
University of Dayton
Sidaard Gunasekaran
University of Dayton
Wing trailing edge shapes can significantly affect the magnitude and distribution of vorticity in the wake which in turn influences the parasitic drag coefficient and acoustic tones. It is also hypothesized that the trailing edge extensions at the wingtip can influence the momentum transfer between the free shear layer and the wingtip vortex which in turn has an impact on the induced drag coefficient. Two different categories of trailing edge extensions, one with the extensions along the span and the other with extensions at the wingtip were experimentally investigated at the University of Dayton Low Speed Wind Tunnel (UD-LSWT). Force based experiments and Particle Image Velocimetry (PIV) was conducted on five different types of wingtip trailing edge extensions and five different types of full span trailing edge extensions to characterize the changes in aerodynamic efficiency and to specifically identify the changes in parasite and induced drag coefficient of the wing. Tests were conducted at a Reynolds number of 300,000. The results from the force based testing for the 10 different wing cases will be presented.
Poster Presentations
Design & Optimization
Abstract ID: DESS2019-021
Developing and Implementing a Viability Framework to Evaluate 3D-Printed Construction
Jenee Jagoda
Air Force Institute of Technology
Steven Schuldt
Air Force Institute of Technology
3D-printed construction is an advanced, additive construction process capable of producing a wide range of complex structures and geometries without formwork using a layer-by-layer material deposition approach. The method has been used to successfully construct residential homes, apartment buildings, hotels, office buildings, and bridges. While still in the early stages of development, 3D-printed construction displays the potential to address issues and challenges faced by conventional construction by lowering total costs, decreasing labor requirements, eliminating the need for formwork, reducing material utilization, increasing customization, shortening construction duration, and enhancing sustainability. This research examines the viability of 3D-printed construction by focusing on the benefits and challenges that labor requirements, material considerations, structural design, construction efficiency, supply and transportation requirements, and environmental impact pose to overall cost and viability. With continued investment in research and development, 3D printing could foreseeably become a viable and accepted method of construction in the near future; transforming the way the industry manages costs, labor, materials, scheduling, customization, and sustainability.
One way to assess the viability of existing 3D-printed construction technology is through proof-of-concept case studies. In addition to providing a broad overview of the viability of 3D-printed construction, this research also investigates a recent demonstration of 3D-printed construction capability that took place in an expeditionary environment at Camp Pendleton, California in December 2018 between the U.S. Marines, Navy, Air Force, and Army Corps. The tri-service exercise culminated in the construction of a 10-meter concrete bridge – the first of its kind to be both printed and placed in a field environment. Despite challenges with weather, materials, hardware, and power, the U.S. military successfully demonstrated the potential of 3D-printed construction in the expeditionary environment by proving it is possible to print and place a bridge on-site using locally sourced materials. The U.S. military also exhibited the potential of 3D printing to reduce the labor, materials, and logistics required for military construction. Furthermore, the exercise revealed additional, future opportunities to automate the 3D printing process and lessen the manpower demand.
Abstract ID: DESS2019-023
Development of a Forward Operating Base Assessment Model Quantifying the Environmental and Economic Performance of Site Infrastructure
Jamie Filer
Air Force Institute of Technology
Steven J. Schuldt
Air Force Institute of Technology
Forward operating bases (FOB) are often detached from established infrastructure grids, requiring a constant resupply of resources. In one instance, a 600-person FOB required 22 trucks per day to deliver necessary fuel and water and remove generated wastes. This logistical burden generates negative environmental impacts and increases operational costs. To minimize these consequences, construction planners can implement sustainability measures such as renewable energy systems, improved waste management practices, and energy-efficient equipment. However, integration of such upgrades can increase construction costs, presenting the need for a tool that identifies tradeoffs among conflicting criteria. To assist planners with this challenge, this research effort presents the development of a novel FOB sustainability assessment model capable of quantifying the environmental and economic performance of a set of infrastructure alternatives. Through field data and literature estimates, a hypothetical FOB is designed and evaluated over three site durations to demonstrate the model’s distinctive capability to accurately and efficiently assess construction alternatives. The model demonstrates that over time, high initial investments may be offset by low operating costs and minimal emissions. The proposed model will enable construction planners to maximize the sustainability of FOBs, creating sites that are more self-sufficient and produce less environmental impacts.
Abstract ID: DESS2019-033
Insulation Sensitivity Analysis for an Optimized Hybrid Energy System Powering a Fabric Shelter
Jay Pearson
Air Force Institute of Technology
Torrey Wagner
Air Force Institute of Technology
Steven Schuldt
Air Force Institute of Technology
During military and disaster relief operations, connecting to an established electrical grid is rarely an option. In these situations, camps consisting of poorly insulated fabric shelters are predominantly powered by inefficient diesel generators that require frequent fuel resupply. In order to reduce the fuel demand of these generators, camps may utilize photovoltaic-battery systems. This paper presents an innovative cost-performance model capable of optimizing solar array size, battery backup system, and shelter insulation type to minimize the operating cost of powering a single fabric shelter. Model performance was evaluated using one year of insolation, weather and energy requirement data from a shelter located in Southwest Asia. For a shelter with R-4.7 insulation, the model generated an optimal system configuration consisting of a 251 m2 solar array and an 86 kWh lithium-ion battery. Over one year, this system would reduce the fuel consumption by 97% and save $1.1 million, including system purchase price, compared to a diesel generator. The results of the case study analysis demonstrate the model’s unique capability to optimize photovoltaic-battery system size and shelter insulation material in order to minimize annual operating costs. These capabilities will enable base planners to construct cost-effective military and disaster relief sites that have a reduced reliance on fuel resupplies.
Human Factors / Biomedical
Abstract ID: DESS2019-074
University of Dayton Go Baby Go! Program
Timothy Reissman
University of Dayton
Megan Reissman
University of Dayton
Sophia Chirumbole and Mary Clare Corrigan
University of Dayton
Will provide an overview of the program and its impact on the Dayton community.
Renewable and Clean Energy
Abstract ID: DESS2019-003
A Sustainable Prototype for Renewable Energy: Optimized Prime-Power Generator Solar Array Replacement
Nathan Thomsen
Air Force Institute of Technology
Lt Col Torrey Wagner
Air Force Institute of Technology
Lt Col Andrew Hoisington
Air Force Institute of Technology
Reza Salavani
Air Force Civil Engineering Center
Maj Steven Schuldt
Air Force Institute of Technology
Remote locations such as disaster relief camps, isolated arctic communities, and military forward operating bases are disconnected from traditional power grids forcing them to rely on diesel generators with a total installed capacity of 10,000 megawatts worldwide. The generators require a constant resupply of fuel, resulting in increased operating costs, negative environmental impacts, and challenging fuel logistics. To enhance remote site sustainability, planners can develop standalone photovoltaic-battery systems to replace existing prime power generators. This paper presents the development of a novel cost-performance model capable of optimizing solar array and Li-ion battery storage size by generating tradeoffs between minimizing initial system cost and maximizing power reliability. A case study for the replacement of an 800 kilowatt generator, the U.S. Air Force’s standard for prime power at deployed locations, was analyzed to demonstrate the model and its capabilities. A MATLAB model, simulating one year of solar data, was used to generate an optimized solution to minimize initial cost while providing over 99% reliability. Replacing a single diesel generator would result in a savings of 1.9 million liters of fuel, eliminating 100 fuel tanker truck deliveries annually. The distinctive capabilities of this model enable designers to enhance environmental, economic, and operational sustainability of remote locations by creating energy self-sufficient sites, which can operate indefinitely without the need for resupply.
Keywords: renewable energy, photovoltaic, solar array, optimization, energy storage, diesel generator, battery, standalone, isolated sites.
Abstract ID: DESS2019-065
Data Mining for Residential Buildings using Smart WiFi Thermostats
Kefan Huang
University of Dayton
Abdulrahman Alanezi
University of Dayton
Smart WiFi thermostats are not just a device for controlling heating and cooling comfort in buildings, they also can learn from occupant behaviors and permit occupants to control their comfort remotely. This research seeks to go beyond this state of the art by utilizing smart thermostat WiFi data from detached residences to develop dynamic models to predict room temperature and cooling/heating demand and then apply these models to new thermostat temperature scheduling scenarios, associated with lower energy cooling/heating. The ultimate objective of this effort is to reduce energy use in residences and demonstrate the ability to respond to peak utility demand events while maintaining thermal comfort within a minimally acceptable range. Deep learning neural network, Long-Short Term Memory (LSTM) and Encoder-Decoder LSTM approaches are used to develop these dynamic models. These algorithms are assessed using training and testing accuracy, and fine-tuning performance to seek the best learning algorithm. Preliminary results demonstrate that LSTM outperforms DNN and Encoder-Decoder LSTM approach, yielding MAPE values for features of less than 0.5. Additionally, the models developed are shown to be highly accurate (within 90%) in predicting savings from aggressive thermostat setpoint schedules aimed at yielding deep reduction (up to 17.3%) in heating and cooling energy.
Keywords: Smart WiFi thermostats; Deep learning neural network; Long-short term memory; Encoder-Decoder LSTM; Demand management; Energy savings
Abstract ID: DESS2019-064
A Smart WiFi Thermostat Data-Based Neural Network Model for Controlling Thermal Comfort in Residences Through Estimates of Mean Radiant Temperature
Yisheng Lou
University of Dayton
Indoor thermal comfort is usually achieved by tenants manually adjusting fixed temperature set-points. Prior research has explored automated control of thermal comfort based on the concept of a Predicted Mean Vote (PMV) index, which has been developed to provide a model of perceived human comfort. However, Mean Radiant Temperature (MRT), a highly dominating parameter to calculate the PMV index has been 1) inaccurately assumed to be the same as indoor air temperature and 2) costly to implement due to the need for numerous additional sensors. This research paradigm poses to leverage prior work in R-value estimates and the usage of smart WiFi thermostats in the prediction of desired indoor air temperature for the purpose of reduced energy consumption.
Harnessing smart WiFi thermostat technology, the proposed model collects data including the measured air temperature and humidity of the indoor environment, the cooling/heating/fan status, as well as the outdoor weather conditions. In parallel, the model implements an improved accuracy of MRT using estimated R-values and outdoor temperatures. Prior research done by Alinezi et al. has shown the ability to estimate R-values through WiFi thermostats data, thereby eliminating the need for additional sensors and measurements. The proposed model implements a NARX Neural Network to predict indoor air temperature as a function of previous temperature set-point schedules and external weather conditions. The results show the model has the ability to provide accurate forecasting with alternative set-point schedules. With such a model, it is now possible to simulate different temperature set-point due to comfort level change. Moreover, in this research, it is assumed that at each time there is a desired setpoint temperature to maintain minimum human comfort based on the PMV index with estimated MRT without further measurements. Application of this ideal temperature set-point for minimum human comfort to historical weather data and indoor weather conditions can yield an estimate for minimum cooling energy. The initial results show a 1% energy savings for cooling and 50% energy savings with fan usage during a given time period, when comparing actual cooling and minimum cooling at minimum comfort level. Based on this research, it is proposed that the approach to estimate MRT can be used to calculate a more accurate PMV value and a better representation of human comfort, without anything more than a smart WiFi thermostat. Thus, a control strategy based on this paradigm can both achieve thermal comfort in residential buildings and less energy consumption.