Theses and Dissertations from UMD

Permanent URI for this communityhttp://hdl.handle.net/1903/2

New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a give thesis/dissertation in DRUM

More information is available at Theses and Dissertations at University of Maryland Libraries.

Browse

Search Results

Now showing 1 - 10 of 27

Aeroacoustic Implications of Installed Propeller Interactional Aerodynamics and Transient Propeller Motions
(2023) Jayasundara, Dilhara; Baeder, James; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
The emergence of advanced air mobility and sustainable aviation concepts have revived the interest in propeller-driven aircraft. A number of electric vertical take-off and landing (eVTOL) aircraft have been developed to cater to the demands of urban air mobility (UAM) and significant advancements have been made in unmanned aerial vehicles (UAV) equipped with vertical take-off and landing capabilities. However, the community acceptance of these new aircraft configurations highly depends on having a low noise footprint as they will operate in dense urban environments. Propeller noise is considered the major source of noise in these aircraft with the introduction of electric propulsion and it can significantly increase with the effects of installation and transient propeller motions. This study aims to comprehend the complex aerodynamic interactions within such aircraft that result from propeller installation and contribute to the generation of high noise levels. To understand the physics of propeller installation, a wingtip-mounted propeller was analyzed at several angles of attack using computational fluid dynamics (CFD) based on Reynolds-averaged Navier-Stokes (RANS) equations and computational aeroacoustics based on the Ffowcs Williams - Hawkings equation. The aeroacoustic implications of the propeller axis inclination and the propeller-wing aerodynamic interaction were studied in-depth. The propeller-wing interaction leads to a significant increase in propeller noise (~20 to 30 dB increase along the rotational axis) and causes the wing to generate a loading noise in the same order of magnitude as the propeller noise. To extrapolate the understanding of installation effects to a full aircraft, the aeroacoustic characteristics of a quadrotor biplane tailsitter were analyzed in both hover and forward flight focusing on the rotor-rotor and rotor-airframe aerodynamic interaction. The rotor-rotor interaction was found to be a significant source of loading noise in hover but having the fuselage as a physical barrier between the rotors largely reduces its effect. The airframe loading noise and rotor broadband noise are equally dominant as the rotor tonal noise when the aircraft is in forward flight. Moreover, the study evaluated the effectiveness of rotor synchrophasing in reducing the aircraft noise footprint and it showed promising results in hover, causing a reduction of aircraft noise by more than 10 dB. Furthermore, an efficient computational aeroacoustics framework was developed to facilitate the computations, ensuring optimal utilization of the computational resources. The CPU and GPU parallelization and other optimization techniques were able to achieve a 98% reduction in computation time for an isolated propeller case. This enabled the rapid aeroacoustic computations of periodic and non-periodic problems. This was used to analyze the aeroacoustics of an isolated propeller undergoing a transition from hover to forward flight. The aerodynamic and acoustic results of the unsteady case were compared with quasi-steady cases performed at intermediate tilt angles. The quasi-steady CFD simulations predicted the unsteady transition aeroacoustics with reasonable accuracy. A tilting quasi-steady approach was proposed to better capture the aerodynamics and acoustics of the unsteady transition.
DEVELOPMENT OF VARIABLE TUBE GEOMETRY HEAT EXCHANGERS USING ADJOINT METHOD WITH PERFORMANCE EVALUATION OF AN ADDITIVELY MANUFACTURED PROTOTYPE
(2023) Klein, Ellery; Radermacher, Reinhard; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Air-to-refrigerant heat exchangers are a key component for heating, ventilation, air conditioning, and refrigeration (HVAC&R) systems. The performance of these heat exchangers is limited by their air-side thermal resistance. Finless non-round bare tube designs have the potential to improve the air-side thermal-hydraulic performance over their finned counterparts and consequently improve the coefficient of performance (COP) of air-conditioning systems. Previous researchers have used heuristic methods such as multi-objective genetic algorithms (MOGA) with approximation-assisted optimization (AAO) methods utilizing computational fluid dynamics (CFD) based metamodels to shape and topology optimize non-round bare tube heat exchangers. A rather unexplored optimization technique used for heat exchanger optimizations is the gradient based adjoint method. CFD solvers utilizing discrete adjoint methods can be used to shape optimize bare tube heat exchangers and can reveal unintuitive, organic, and potentially superior designs. Additionally, additive manufacturing technology has the capability of building these previously unrealizable heat exchanger designs.The objectives of this dissertation are to experimentally evaluate the performance of shape and topology optimized compact bare tube heat exchangers with non-round bare tubes on a 1) component level, and 2) system level integrated into an air conditioner. Plus, 3) develop new shape optimized variable geometry compact bare tube heat exchangers using discrete adjoint methods for HVAC&R applications. First, a comprehensive experimental investigation of multiple shape and topology optimized compact non-round bare tube heat exchangers was conducted under dry and wet evaporator, condenser, and radiator conditions. For all heat exchangers, air-side pressure drop and heat transfer capacity were predicted within 37% and 15%, respectively. Next, an experimental test facility capable of evaluating the system level performance of a 7.03-8.79 kW commercial packaged air conditioning unit was designed and constructed. The performance of the air conditioning unit was evaluated before and after its conventional tube-fin evaporator was replaced with a shape and topology optimized bare tube evaporator. Results are presented and discussed. Lastly, an ε-constraint and penalty method optimization scheme was implemented which utilizes a commercial CFD software with a built-in discrete adjoint solver to perform multi-objective shape optimizations of 2D bare tube heat exchangers. Critical solver/mesh set-up to best optimize heat exchangers with 0.5-10.0 mm diameter bare tubes were identified and established. The optimized designs can achieve a 40-50% reduction in air-side pressure drop with at least the same heat transfer capacity compared to the initial circular bare tube geometry. An adjoint shape optimized 500 W bare tube radiator was additively manufactured in polymer and experimentally tested. Air-side pressure drop and heat transfer capacity were predicted within 15% and 10%, respectively. The experimental performance confirms the adjoint method shape optimized designs improve the thermal-hydraulic performance over the initial circular bare tube geometry.
CFD INVESTIGATION OF A PULSE JET MIXED VESSEL WITH RANS, LES, AND LBM SIMULATION MODELS
(2023) Kim, Jung; Calabrese, Richard V.; Chemical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Pulse Jet Mixed (PJM) vessels are used to process nuclear waste due to their maintenance free operation. In this study we model the turbulent velocity field in water during normal PJM operation to gain insight into vessel operations and to evolve a modeling strategy for process design and operator training. Three transient simulation models, developed using Large Eddy Simulation (LES), unsteady Reynolds-Averaged Navier-Stokes (RANS), and Lattice Boltzmann Method (LBM) techniques, are compared to velocity measurements acquired for 3 test scenarios at 3 locations in a pilot scale vessel at the US DOE National Energy Technology Laboratory (NETL). The LES and RANS simulations are performed in ANSYS Fluent, and the LBM simulations in M-STAR.The LES model well predicts the experimental data provided that the operational pressure profile within the individual pulse tubes is considered. While the RANS model failed to predict the data and exhibited significant differences from LES with respect to turbulence quantities, it is a useful comparison tool that can quickly predict averaged flow parameters. The LBM model’s rigid grid system is deemed unsuitable, as currently configured, for the NETL PJM vessel’s wide range of length scales and curved boundaries, resulting in the longest simulation time and least accurate velocity predictions. Predicted velocity and turbulence metrics are explored to better understand the strengths and failures of the three models. Because the LES model produced the most accurate predictions, it is exploited to generate animations and still images on various 2D planes that depict extremely complex flow patterns throughout the vessel with numerous local jets and mixing layer vortices The study concludes with recommendations for future research to improve the model development and validation strategy.
A Time Parallel Approach to Numerical Simulation of Asymptotically Stable Dynamical Systems with Application to CFD Models of Helicopter Rotors
(2023) Silbaugh, Benjamin Scott; Baeder, James D; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Modern High Performance Computing (HPC) machines are distributed memoryclusters, consisting of multi-core compute nodes. Engineering simulation and analysis tools must employ efficient parallel algorithms in order to fully utilize the compute capability of modern HPC machines. The trend in Computational Fluid Dynamcis (CFD) has been to construct parallel solution algorithms based on some form of spatial domain decomposition. This approach has been shown to be a success for many practical applications. However, as one attempts to utlize more compute cores, limitations in strong scalability are inevitably reached due to a diminishing compute workload per compute core and either fixed or increasing communication cost. Furthermore, spatial domain decomposition approaches cannot be easily applied to mid-fidelity structural dynamics or rigid body dynamics models. A significant majority of industrial fluid and structural dynamic models utilize some form of time marching. Thus, if the domain decomposition strategy may be extended to include the temporal dimension, additional opportunity for increased parallelism may be realized. A new form of periodic multiple shooting is proposed that ismatrix-free and may be applied to high-fidelity multiphysics models or other high dimensional systems. The proposed methodology is formulated entirely in the time domain. Therefore, existing time-domain simulation tools may utilize the proposed approach to achieve a high degree of distributed memory parallelism without requiring any reformulation. Furthermore, the proposed methodology may be combined with conventional space domain decomposition techniques and other forms of data parallelism to achieve maximal performance on modern HPC architectures. The proposed algorithm retains the iterative shoot-correct approach of conventational periodic shooting methods. However, the correction stage is formulated using a hierarchical evaluation strategy combined with an Arnoldi subspace approximation to eliminate the need for explicit formulation of Jacobian matricies. The local convergence of the proposed method is formally proven for the case of an asyptotically stable dynamical system. The proposed method is numerically tested for a 2D limit cycle problem, a rigid blade helicoper rotor model with quasi-steady aerodynamics and autopilot trim, and an OVERSET CFD model of a helicopter rotor with prescribed elastic blade motions. The method is observed to be convergent in all test cases and found to exhibit good scalability. The proposed periodic multiple shooting method is a practical means of reducingtime-to-solution for numerical simulations of asymptotically stable periodic systems on distributed memory parallel computers. Furthermore, the proposed method may be used to enhance the parallel scalability of OVERSET CFD models of helicopter rotors in steady periodic flight.
A Time Parallel Approach to Numerical Simulation of Asymptotically Stable Dynamical Systems with Application to CFD Models of Helicopter Rotors
(2023) Silbaugh, Benjamin Scott; Baeder, James; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Modern High Performance Computing (HPC) machines are distributed memoryclusters, consisting of multi-core compute nodes. Engineering simulation and analysis tools must employ efficient parallel algorithms in order to fully utilize the compute capability of modern HPC machines. The trend in Computational Fluid Dynamcis (CFD) has been to construct parallel solution algorithms based on some form of spatial domain decomposition. This approach has been shown to be a success for many practical applications. However, as one attempts to utlize more compute cores, limitations in strong scalability are inevitably reached due to a diminishing compute workload per compute core and either fixed or increasing communication cost. Furthermore, spatial domain decomposition approaches cannot be easily applied to mid-fidelity structural dynamics or rigid body dynamics models. A significant majority of industrial fluid and structural dynamic models utilize some form of time marching. Thus, if the domain decomposition strategy may be extended to include the temporal dimension, additional opportunity for increased parallelism may be realized. A new form of periodic multiple shooting is proposed that ismatrix-free and may be applied to high-fidelity multiphysics models or other high dimensional systems. The proposed methodology is formulated entirely in the time domain. Therefore, existing time-domain simulation tools may utilize the proposed approach to achieve a high degree of distributed memory parallelism without requiring any reformulation. Furthermore, the proposed methodology may be combined with conventional space domain decomposition techniques and other forms of data parallelism to achieve maximal performance on modern HPC architectures. The proposed algorithm retains the iterative shoot-correct approach of conventational periodic shooting methods. However, the correction stage is formulated using a hierarchical evaluation strategy combined with an Arnoldi subspace approximation to eliminate the need for explicit formulation of Jacobian matricies. The local convergence of the proposed method is formally proven for the case of an asyptotically stable dynamical system. The proposed method is numerically tested for a 2D limit cycle problem, a rigid blade helicoper rotor model with quasi-steady aerodynamics and autopilot trim, and an OVERSET CFD model of a helicopter rotor with prescribed elastic blade motions. The method is observed to be convergent in all test cases and found to exhibit good scalability. The proposed periodic multiple shooting method is a practical means of reducingtime-to-solution for numerical simulations of asymptotically stable periodic systems on distributed memory parallel computers. Furthermore, the proposed method may be used to enhance the parallel scalability of OVERSET CFD models of helicopter rotors in steady periodic flight.
NUMERICAL SIMULATION OF THE BLUE WHIRL: A REACTING VORTEX BREAKDOWN PHENOMENON
(2019) Chung, Joseph Dong il; Oran, Elaine S; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
The blue whirl is a small, stable, spinning blue flame that evolved spontaneously in recent laboratory experiments while studying turbulent, sooty fire whirls. It burns a range of different liquid hydrocarbon fuels cleanly with no soot production, presenting a new potential way for low-emission combustion. This thesis uses numerical simulations to present, for the first time, the flame and flow structure of the blue whirl. These simulations show that the blue whirl is composed of three different flames - a diffusion flame and a premixed rich and lean flame - all of which meet in a fourth structure, a triple flame which appears as a whirling blue ring. The results also show that the flow structure emerges as the result of vortex breakdown, a fluid instability which occurs in swirling flows. This thesis also presents the development and testing of the numerical algorithms used in the simulation of the blue whirl. This work is a critical step forward in understanding how to use this new form of clean combustion.
Hamiltonian Paths and Strands for Unified Grid Approach for Computing Aerodynamic Flows
(2019) Jung, Yong Su; Baeder, James D; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
A solution algorithm using Hamiltonian paths and strand grids is presented for compressible Reynolds-Averaged Navier–Stokes (RANS) formulation as a unified grid approach. The hidden line-structure is robustly identified on the general unstructured grid with mixed elements which provides a framework for line-solvers similar to that with a structured grid solver. A pure quadrilateral/hexahedral mesh is a prerequisite for the line identification and enables approximate factorization along the lines on the unstructured grid. Among various methods, subdivision is the easiest way to obtain a pure quadrilateral/hexahedral mesh from the general unstructured grid. Strand based grids have been employed to extend to three-dimension by extruding the surface mesh. As a result, Hamiltonian paths on the surface mesh represents two distinct surface coordinate directions and the strand grids represent the wall normal direction. These structures are analogous to the grid coordinate directions in a structured grid solver, therefore the current method is directly applicable on the typical structured grid. The numerical accuracy and convergence rate are investigated under various flow conditions. The numerical efficiency was improved with a line-implicit method compared to a point-implicit method on the unstructured grid. Both stencil- and gradient-based reconstructions are available on the unstructured grid and the numerical accuracy for each method was evaluated on both structured and unstructured grids. The combined reconstruction method was also proposed for the current mesh system which uses both stencil- and gradient-based reconstructions simultaneously but for different grid directions. The solution convergence rate has been improved further using Generalized Minimum Residual (GMRES) method. GMRES requires a preconditioning step which is performed using the line-implicit method. GMRES provides better convergence rate than the pure line-implicit method at various flow conditions. Both parts of mesh generation and flow solver are parallelized to be executed in parallel using METIS and MPI. During the mesh generation, two different domain partitioning methods are suggested for the strand grid and the unstructured volume mesh, respectively. The capability of the flow solver has been extended for rotary wing simulations: time-accurate method, turbulence model, moving grids, and overset meshes. The flow solver is also integrated into a multi-mesh/multi-solver paradigm through a Python framework which enables a more efficient solution algorithm than a single-solver. The integrated framework has been applied to various practical problems, such as wind turbine, rotor hub, and elastic rotor blades.
Integrated Field Inversion and Machine Learning With Embedded Neural Network Training for Turbulence Modeling
(2019) Holland, Jonathan Richard; Baeder, James D; Duraisamy, Karthik; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
A rich set of experimental and high fidelity simulation data is available to improve Reynolds Averaged Navier Stokes (RANS) models of turbulent flow. In practice, using this data is difficult, as measured quantities cannot be used to improve models directly. The Field Inversion and Machine Learning (FIML) approach addressed this challenge through an inference step, in the form of an inverse problem, which treats inconsistencies between the models and the data in a consistent manner. However, a separate learning algorithm is not always able to be learned from the generated inverse problem data accurately. Two new methods of incorporating higher fidelity data into RANS turbulence models via machine learning are proposed and applied for the first time in this thesis. Both build on the FIML framework by performing learning during the inference step, instead of considering the inference and learning steps separately as in the classic FIML approach. The first new approach embeds neural network learning into the RANS solver, and the second trains the weights of the neural network directly. Additionally, for the first time, the inverse problem can incorporate higher fidelity data from multiple cases simultaneously, promising to improve the generalization of the augmented model. The two new methods and the classic approach are demonstrated with a simple model problem, as well as a number of challenging RANS cases. For a 2D airfoil case, all three FIML augmentations are shown to improve predictions, with the new methods demonstrating increased regularization. Additionally, a model augmentation is generated by considering seven angles of attack of an airfoil in the inference step, and the augmentation is shown to improve predictions on a different airfoil. Additional cases are considered including a transonic shock wave boundary layer interaction and the NASA wall-mounted hump. In all cases, the inference is shown to improve predictions. For the first time, the inverse problem accounts for the limitations of the learning procedure, guaranteeing that the model discrepancy is optimal for the chosen learning algorithm. The results in this thesis prove that learning during the inference step provides additional regularization, and guarantees the inference produces learnable model discrepancy.
Advancing the Multi-Solver Paradigm for Overset CFD Toward Heterogeneous Architectures
(2019) Jude, Dylan P; Baeder, James; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
A multi-solver, overset, computational fluid dynamics framework is developed for efficient, large-scale simulation of rotorcraft problems. Two primary features distinguish the developed framework from the current state of the art. First, the framework is designed for heterogeneous compute architectures, making use of both traditional codes run on the Central Processing Unit (CPU) as well as codes run on the Graphics Processing Unit (GPU). Second, a framework-level implementation of the Generalized Minimal Residual linear solver is used to consider all meshes from all solvers in a single linear system. The developed GPU flow solver and framework are validated against conventional implementations, achieving a 5.35× speedup for a single GPU compared to 24 CPU cores. Similarly, the overset linear solver is compared to traditional techniques, demonstrating the same convergence order can be achieved using as few as half the number of iterations. Applications of the developed methods are organized into two chapters. First, the heterogeneous, overset framework is applied to a notional helicopter configuration based on the ROBIN wind tunnel experiments. A tail rotor and hub are added to create a challenging case representative of a realistic, full-rotorcraft simulation. Interactional aerodynamics between the different components are reviewed in detail. The second application chapter focuses on performance of the overset linear solver for unsteady applications. The GPU solver is used along with an unstructured code to simulate laminar flow over a sphere as well as laminar coaxial rotors designed for a Mars helicopter. In all results, the overset linear solver out-performs the traditional, de-coupled approach. Conclusions drawn from both the full-rotorcraft and overset linear solver simulations can have a significant impact on improving modeling of complex rotorcraft aerodynamics.
Drag and Noise Reduction of Flatback Airfoil by Span-wise Wavy Trailing Edge
(2019) Yang, SeungJoon; Baeder, James D; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
The flatback airfoil is a promising idea for future large wind turbine blade structure design; however, it causes notable drag increase and low frequency tonal noise due to the presence of span-wise coherent standing flow and Karman vortex shedding at the trailing edge. Current dissertation proposes a span-wise wavy trailing edge design as a solution to flatback airfoil drag and noise, and provides relevant CFD results. Proposed span-wise wavy trailing edge prevents the span-wise coherent standing flow and vortex shedding, and results in a decrease of the tonal noise and pressure drag of the airfoil. Delayed Detached Eddy Simulation is employed in HPC environments. In-house developed N-S solvers, OVERTURNS (CPU-based) and GPURANS3D (GPGPU) are used for the computation. A design parametric study for the span-wise wavy trailing edge is conducted in the first half of the dissertation. Aerodynamic and aeroacoustic performance of a particular flatback airfoil (FB3500-1750, tTE is 17.5% of a chord length) and various span-wise wavy trailing edge modifications are investigated, regarding the influence of the major wave design parameters. A total of 16 design variations of the wavy trailing edge are investigated. For a Reynolds number 666,000 and Mach number 0.3, the best trailing edge wave design is a combination of the less portion - more than 0.25c length - shallow wave depth; the best design reduces 60% (maximum) of the flatback airfoil with only 7% lift loss, and results in about 150% (maximum) of lift/drag ratio increase. Measured tonal noise is also reduced by about 20-25dB(SPL) with the best performance design. In the second half of the dissertation, the isolated rotor simulation is performed for a straight-flat trailing edge wind turbine blade and its wavy trailing edge modification. In the simulation, the Blunt-Wavy trailing edge (a combination of trailing edge augmentation and wavy modification) is proposed and applied at the inboard of baseline blade (SNL100-03FB). Applying the best performance wavy design, overall turbine power generation is increased by 2.62% and tonal noise is decreased by 5-15dB.

Theses and Dissertations from UMD

Browse

Filters

Settings

Sort By

Results per page

Search Results