Doctor of Philosophy, The Ohio State University, 2023, Mechanical Engineering

Decades of successful research and development on automotive silencers for engine breathing systems have brought about significant reductions in engine noise emissions. Much of this research has pursued airborne noise at relatively low frequencies, which typically involves planar wave propagation. However, with the increasing demand for downsized turbocharged engines in passenger cars, high-frequency compressor noise has become a challenge in engine induction systems. Elevated frequencies on the order of 10 kHz promote multi-dimensional wave propagation rendering at times conventional silencer treatments ineffective due to the underlying assumption of one-dimensional wave propagation in their design. Multi-dimensional waves also cause experimental in-duct acoustic measurements to be sensitive to the angular and axial location of pressure transducers, therefore making it more challenging to characterize the acoustics of modern turbocharger compressors. The present study focuses on developing a reflective high-frequency silencer for turbocharger compressors to mitigate tonal noise at the blade-pass frequency (number of main impeller blades times the compressor shaft rotational speed in revolutions per second) within the compressor inlet duct for a wide range of rotational speeds. The approach features a novel “acoustic straightener” that creates exclusive planar wave propagation near the silencing elements, hence improving overall acoustic attenuation. An analytical treatment using acoustic lumped impedance models is combined with three-dimensional (3D) acoustic finite element method (FEM) to conceptualize a silencer configuration comprising quarter-wave resonator (QWR) arrays. The effects of mean flow and nonlinearities on acoustics are captured by 3D computational fluid dynamics (CFD) simulations, which are also utilized to introduce geometry modifications that reduce flow losses. The CFD simulations reveal noise generation due to flow-acoustic coupling: a phenomeno (open full item for complete abstract)

Committee: Ahmet Selamet (Advisor); Rajendra Singh (Committee Member); Jung-Hyun Kim (Committee Member) Subjects: Acoustics; Automotive Engineering; Design; Engineering; Experiments; Fluid Dynamics; Mechanical Engineering

Doctor of Philosophy, The Ohio State University, 2023, Mechanical Engineering

Since its inception in the 1930s, the evolution of the jet engine has largely been dictated by the need to improve propulsion and fuel economy while reducing noise and environmental impact. Although rapid engineering advancements have enabled progress towards these design objectives, they have also yielded increasingly complex air flowpaths, and further improvements have been inhibited by a lack of understanding of the fluid dynamics. This is because the flowfields are typically characterized by interacting turbulent jets, shear layers, boundary layers, and wakes which evolve in the presence of supersonic expansion and shock waves. Additionally, they are rife with fluid dynamic instabilities that may pose a challenge to the structural, thermal and acoustic performance of the engine. To address this knowledge gap, the present work builds upon a collaborative numerical and experimental campaign between The Ohio State University and Syracuse University, that investigates a supersonic, multistream, airframe-integrated, rectangular nozzle architecture that is representative of emerging industry designs. The configuration consists of a contoured, single-sided expansion nozzle on one side, and a flat aft-deck surface representing the airframe on the other. Within the nozzle, two rectangular streams - a supersonic (Mach 1.6) "core" stream and sonic (Mach 1) "deck" stream - interact after being initially separated by a thick splitter plate. The flow conditions and geometry of the nozzle render a complex 3D flowfield which has been extensively examined in previous work. It is also comprised of a shear layer instability which is initiated at the splitter plate trailing edge (SPTE) and is associated with large vortical structures and a strong tone that have a deleterious effect on the acoustic and structural characteristics of the nozzle. Although previous research works attempting to mitigate the instability by thinning out the SPTE have been promising, such (open full item for complete abstract)

Committee: Datta Gaitonde (Advisor); Jen-Ping Chen (Committee Member); Seung Hyun Kim (Committee Member) Subjects: Aerospace Engineering; Mechanical Engineering

Doctor of Philosophy, The Ohio State University, 2022, Aerospace Engineering

Laminar-to-turbulent transition estimation is one of the key challenges in designing hypersonic vehicles, primarily due to uncertainties in the disturbance environment and the enormous parametric space involved. A comprehensive understanding of the underlying physical processes is lacking, hindering the development of efficient control strategies. In particular, the role of leading-edge bluntness during various stages of transition remains unresolved, which is of practical relevance. This thesis documents a series of five complementary numerical studies aimed at understanding the laminar-to-turbulent transition process on hypersonic flat plates with varying leading edge bluntness. Direct numerical simulations are used to accurately resolve the spatio-temporal scales of the flows. These are complemented with physics and data-driven techniques to gain insight into the flow fields. The first study investigates the role of leading-edge bluntness in the receptivity of broadband freestream disturbances and their linear amplification mechanisms. At small nose radii, the disturbances trigger Mack modes, whose growth rate reduces with increasing bluntness. After a critical radius, waves of a different class with predominant support in the entropy layer are amplified. With increasing bluntness, the amplification rates of these disturbances increase. In the second study, the laminar-to-turbulent transition on a sharp flat plate is examined with stochastic freestream forcing. Second-mode waves are most amplified, followed by their fundamental resonance and finally the onset of turbulence. During various stages of the transition, acoustic and vortical dissipation dominate the wall heating, and near-wall streaks enhance the skin-friction. The third study scrutinizes the characteristics of the late stages of transition by triggering an isolated bypassed turbulent spot on a hypersonic flat plate, and examines its evolution with the momentum potential theory. Mack (open full item for complete abstract)

Committee: Datta Gaitonde Prof (Advisor); Jack McNamara Prof (Committee Member); Lian Duan Prof (Committee Member) Subjects: Aerospace Engineering

Master of Science, The Ohio State University, 2022, Aerospace Engineering

The effect of varying submergence on the transonic flow past canonical wall-mounted hemispheres is investigated at a freestream Mach number, M = 0.8 using Delayed Detached Eddy Simulations (DDES). Four submergence levels are considered ranging from a full hemisphere (100% exposed) to a highly submerged case where the equator is well below the waterline (40% exposed). Analysis of the mean characteristics indicates a reduction in strength and extent of dominant flow topology, including the horseshoe vortex and the counter-rotating vortices in the wake. Additionally, it is found that the mean line of surface separation moves downstream with submergence. This finding is validated with analysis of the unsteady streamwise shock position, which indicates the mean shock foot position also moved downstream with submergence. However, the frequency associated with the streamwise motion of the shock remains consistent between the cases in terms of a suitably non- dimensionalized Strouhal number, StD ∼ 0.26. The unsteady surface forces and modal analysis are used to quantify the effect of submergence on two correlated shock-wake modes, as these are dominant unsteady features in flow over full hemispheres. The first mode is the “breathing” mode, corresponding to spanwise symmetric wake shedding and correlates to streamwise shock oscillations. The second mode is the “shifting”, corresponding to spanwise anti-symmetric wake shedding and correlates with the spanwise rocking of the shock. Proper Orthogonal Decomposition (POD) is used to isolate and rank the different modes; as the hemisphere is submerged, there is an evident change in prominence from the anti-symmetric shifting to the symmetric breathing mode. Dynamic Mode Decomposition (DMD) is used to investigate the spectral content of these modes; the results show that the breathing mode collapses at a frequency of StD ∼ 0.26 for all cases, while the shifting mode extends over a broad frequency range of StD between 0.13 and 0.21.

Committee: Jack McNamara (Committee Member); Datta Gaitonde (Advisor) Subjects: Aerospace Engineering

Master of Science, The Ohio State University, 2018, Computer Science and Engineering

High-fidelity observations of non-linear dynamical systems that are of practical interest lead to massive data sets which do not fit on a single computing node. Therefore, modal decomposition techniques must be able to exploit the capability of high-performance computing (HPC) facilities. Proper Orthogonal Decomposition and Sparse Coding are two of the commonly used modal decomposition techniques to obtain reduced order models. The goal of the research is to parallelize and implement these algorithms so that they can be used on high-performance computing clusters in order to expedite the process of modal decomposition from massive data sets. However, computation on various machines is associated with high memory usage and significant communication cost. Moreover, the overall computational cost is sensitive to the type of data set and various parameters of the algorithm. Therefore, several strategies are discussed and implementations are developed to address these constraints to perform expedient modal decomposition. Furthermore, a systematic study is performed over multiple data sets to assess the performance and scalability of the implementations.

Committee: Jack McNamara (Committee Member); Sadayappan P (Advisor) Subjects: Aerospace Engineering; Computer Science

Doctor of Philosophy, The Ohio State University, 2016, Aero/Astro Engineering

Galerkin projection is a commonly used reduced order modeling approach; however, stability and accuracy of the resulting reduced order models are highly dependent on the modal decomposition technique used. In particular, deriving stable and accurate reduced order models from highly turbulent flow fields is challenging due to the presence of multi-scale phenomenon that cannot be ignored and are not well captured using the ubiquitous Proper Orthogonal Decomposition (POD). A truncated set of proper orthogonal modes is biased towards energy dominant, large-scale structures and results in over-prediction of kinetic energy from the corresponding reduced order model. The accumulation of energy during time-integration of a reduced order model may even cause instabilities. A modal decomposition technique that captures both the energy dominant structures and energy dissipating small scale structures is desired in order to achieve a power balance. The goal of this dissertation is to address the stability and accuracy issues by developing and examining alternative basis identification techniques. In particular, two modal decomposition methods are explored namely, sparse coding and Dynamic Mode Decomposition (DMD). Compared to Proper Orthogonal Decomposition, which seeks to truncate the basis spanning an observed data set into a small set of dominant modes, sparse coding is used to identify a compact representation that span all scales of the observed data. Dynamic mode decomposition seeks to identity bases that capture the underlying dynamics of a full order system. Each of the modal decomposition techniques (POD, Sparse, and DMD) are demonstrated for two canonical problems of an incompressible flow inside a two-dimensional lid-driven cavity and past a stationary cylinder. The constructed reduced order models are compared against the high-fidelity solutions. The sparse coding based reduced order models were found to outperform those developed using the dynamic mode and (open full item for complete abstract)

Committee: Jack McNamara (Advisor); Datta Gaitonde (Committee Member); Ryan Gosse (Committee Member); Joseph Hollkamp (Committee Member); Mohammad Samimy (Committee Member) Subjects: Aerospace Engineering

MS, University of Cincinnati, 2014, Engineering and Applied Science: Mechanical Engineering

The vast field of rotational systems – including measurement capabilities, analytical tools and observability – is still evolving. Spectral maps and order tracks are the most popular tools for analyzing the behavior of various components subject to a rotational characteristic. Although various forms of these tools are well researched and implemented, they are still susceptible to improper sensor location on the structure and to measurement noise. This thesis attempts to bridge the gap between properly located sensors and effective analysis based on their measurements. The concept of singular value decomposition (SVD), which is already well used in modal analysis, forms the basis of observability improvement. By using response data acquired from multiple sensors on a structure, it is possible to calculate and plot the singular values obtained from the entire frequency domain response data at each point in the spectral map graph. The resulting singular value plots will depict the magnitude of contribution of the sensor assembly which can form a noise-free and reliable basis for further analytical tools.

Committee: Randall Allemang Ph.D. (Committee Chair); David L. Brown Ph.D. (Committee Member); Allyn Phillips Ph.D. (Committee Member) Subjects: Engineering

Master of Science, University of Akron, 2012, Civil Engineering

Conventional concentrically braced frame (CBF) systems have limited drift capacity prior to structural damage, often leading to brace buckling under moderate earthquake input, which results in residual drift. Self-centering CBF (SC-CBF) systems have been developed to maintain the economy and stiffness of the conventional CBFs while increasing the ductility and drift capacity. SC-CBF systems are designed such that the columns uplift from the foundation at a specified level of lateral loading, initiating a rocking (rigid body rotation) of the frame. Vertically aligned post tensioning bars resist column uplift and provide a restoring force to return the structure to its initial state (i.e., self-centering the system). Friction elements are used at the lateral-load bearings (where lateral load is transferred from the floor diaphragm to the SC-CBF) to dissipate energy and reduce the peak structural response. Previous research has identified that the frame geometry is a key design parameter for SC-CBFs, as frame geometry relates directly to the energy dissipation capacity of the system. This thesis therefore considered three prototype SC-CBFs with differing frame geometries for carrying out a comparative study. The prototypes were designed using previously developed performance based design criteria and modeled in OpenSees to carry out nonlinear static and dynamic analyses. The design and analysis results were then thoroughly investigated to study the effect of changing frame geometry on the behavior of SC-CBF systems. The rocking response in SC systems introduces large higher mode effects in the dynamic responses of structure, which, if not properly addressed during design, can result in seismic demands significantly exceeding the design values and may ultimately lead to a structural failure. To compare higher mode effects on different frames, proper quantification of the modal responses by standard measures is therefore essential. This thesis proposes three normalized q (open full item for complete abstract)

Committee: David Roke Dr. (Advisor); Kallol Sett Dr. (Committee Co-Chair); Qindan Huang Dr. (Committee Member) Subjects: Civil Engineering