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
33 results
Search Results
Item Impact of Polymeric Drops on Drops and Films of a Different but Miscible Polymer(2024) Bera, Arka; Das, Siddhartha; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The fluid mechanics of a liquid drop impacting on another stationery (or spreading) liquid drop or on a liquid film (of thickness comparable, or smaller, or larger than the impacting drop) has attracted significant attention over the past several years. Such problems represent interesting deviations from the more widely studied problems of liquid drops impacting on solid surfaces having different wettabilities with respect to the impacting drops. These deviations stem from the fact that the resting liquid (in the form of the drop or the film) itself undergoes deformation on account of the drop impact and can significantly affect the overall combined drop-drop or drop-film dynamics. The problem becomes even more intriguing depending on the rheology of the drop(s) and the film as well as the (im)miscibility of the impacting drop with the underlying drop or the film. Interestingly, the majority of such drop-impact-on-drop or drop-impact-on-film studies have considered Newtonian drop(s) and films, with little attention to polymeric drop(s) and films. This thesis aims to bridge that void by studying, using Direct Numerical Simulation (DNS) based computational methods, the impact-driven dynamics of one polymeric drop on another (different but miscible) polymeric drop or film. As specific examples, we consider two separate problems. In the first problem, we consider the impact of a PMMA (poly-methyl methacrylate) drop on a resting PVAc (polyvinyl acetate) drop as well as the impact of a PVAc drop on a resting PMMA drop. In the second problem, we consider the impact of a PMMA drop on a PVAc film as well as the impact of a PVAc drop on a PMMA film. For the first problem, the wettability of the resting drop (on the resting surface), the Weber number of the impacting drop, the relative surface tension values of the two polymeric liquids (PVAc and PMMA), and the miscibility (or how fast the two liquids mix) dictate the overall dynamics. PVAc has a large wettability on silicon (considered as the underlying solid substrate); as a result, during the problem of the PMMA drop impacting on the PVAc drop, the PVAc drop spreads significantly and the slow mixing of the two liquids ensures that the PMMA drop spreads as a thin film on top of the PVAc film (formed as the PVAc drop spreads quickly on silicon). Depending on the Weber number, such a scenario leads to the formation of transient liquid films (of multitudes of shapes) with stratified layers of PMMA (on top) and PVAc (on bottom) liquids. On the other hand, for the case of the PVAc drop impacting on the PMMA drop, a combination of the weaker spreading of the PMMA drop on silicon and the “engulfing” of the PMMA drop by the PVAc drop (stemming from the PVAc having a smaller surface tension than PMMA) ensures that the impacting PVAc drop covers the entire PMMA drop and itself interacts with the substrate giving rise to highly intriguing transient and stratified multi-polymeric liquid-liquid structures (such as core-shell structure with PMMA core and PVAc shell). For both these cases, we thoroughly discuss the dynamics by studying the velocity field, the concentration profiles (characterizing the mixing), the progression of the mixing front, and the capillary waves (resulting from the impact-driven imposition of the disturbance). In the second problem, we consider a drop of the PMMA (PVAc) impacting on a film of the PVAc (PMMA). In addition to the factors dictating the previous problem, the film thickness (considered to be either identical or smaller than the drop diameter) also governs the overall droplet-impact-driven dynamics. Here, the impact, being on the film, the dynamics is governed by the formation of crown (signifying the pre-splashing stage) and a deep cavity (the depth of which is dictated by the film thickness) on the resting film. In addition to quantifying these facets, we further quantify the problem by studying the velocity and the concentration fields, the capillary waves, and the progression of the mixing front. For the PMMA drop impacting on the thin film, a noticeable effect is the quick thinning of the PMMA drop on the PVAc film (or the impact-driven cavity formed on the PVAc film), which gives rise to a situation similar to the previous study (development of transient multi-polymeric-liquid structures with stratified polymeric liquid layers). For the case of the PVAc drop impacting on the PMMA film, the PVAc liquid “engulfs” the deforming PMMA film, and this in turn, reduces the depth of the cavity formed, the extent of thinning, and the amplitude of the generated capillary waves. All these fascinating phenomena get captured through the detailed DNS results that are provided. The specific problems considered in this thesis have been motivated by the situations often experienced during the droplet-based 3D printing processes (e.g., Aerosol jet printing or inkjet printing). In such printing applications, it is commonplace to find one polymeric drop interacting with an already deposited polymeric drop or a polymeric film (e.g., through the co-deposition of multiple materials during multi-material printing). The scientific background for explaining these specific scenarios routinely encountered in 3D printing problems, unfortunately, has been very limited. Our study aims to fill this gap. Also, the prospect of rapidly solidifying these polymeric systems (via methods such as in-situ curing) can enable us to visualize the formation of solidified multi-polymeric structures of different shapes (by rapidly solidifying the different transient multi-polymeric-liquid structures described above). Specifically, both PMMA and PVAc are polymers well-known to be curable using in-situ ultraviolet curing, thereby establishing the case where the present thesis also raises the potential of developing PMMA-PVAc multi-polymeric solid structures of various shapes and morphologies.Item HEALTH IMPACTS OF THERMAL RUNAWAY EVENTS IN OUTDOOR LITHIUM-ION BATTERY ENERGY STORAGE SYSTEM INSTALLATIONS(2024) Zhao, Zelda Qijing; McAllister, Jamie; Fire Protection Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This study aimed to develop a methodology for characterizing health impacts of large-scale, outdoor, lithium-ion battery energy storage systems (BESS) thermal runaway events. A literature review was conducted to identify toxic gas yields produced during flaming and non-flaming thermal runaway, as well as mass loss rates, gas temperature, typical BESS unit capacity and dimensions, and event durations. Lithium-iron-phosphate and nickel-manganese-cobalt cell chemistries were assessed. The BESS unit thermal runaway events were modeled in Fire Dynamics Simulator with a bounding analysis for wind and ambient temperature. Concentrations were evaluated using Immediately Dangerous to Life or Health values for occupational exposure and the Protective Action Criteria for Chemicals hierarchy values (Acute Exposure Guideline Levels- Level 1, Emergency Response Planning Guidelines- Level 1, Temporary Emergency Exposure Limits- Level 1) for community exposure. Through application of the methodology, a relationship between exposure limit distance and wind speed, ambient temperature, event duration, cell chemistry, and toxic gas species can be assessed. Under the conditions modeled in this project, exposure limits were exceeded at longer distances in the non-flaming scenarios when compared to the flaming scenarios. Wind speed, ambient temperature, event duration, cell chemistry, and toxic gas species were the controlling factors for non-flaming exposure limit distances. Wind speed was the primary controlling factor for flaming exposure limit distances; however, event duration had some influence.Item The Lives and Times of Stars and Black Holes in the Disks of Active Galactic Nuclei(2024) Dittmann, Alexander Joseph; Miller, Michael C; Astronomy; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Enormous disks of gas are thought to feed the supermassive black holes at the centers of active galaxies; these disks may capture stars from nuclear clusters, or form stars in situ after collapsing under their own gravity. Such stellar populations may enrich these accretion disks with fusion byproducts, cause giant flares in these active galaxies, and leave behind compact remnants detected on earth through gravitational waves emitted as they merge with one another. This dissertation charts a theoretical expedition into these phenomena, from studying the implications of star-forming accretion disks for the growth of black holes in the early universe, to simulating the flow of gas around black hole binaries to ascertain their orbital evolution. After a brief observational and theoretical overview of stars and active galactic nuclei, this dissertation delves into the development of simplified models of accretion disk structure, including the effects of stars and black holes embedded within accretion disks. The ultimate goal of this chapter was to determine if gravitational instability in the outer regions of these accretion disks might lead to the formation of large numbers of black holes, which might go on to merge with the central supermassive black hole; this process might decrease the effective radiative efficiency of accretion onto supermassive black holes, facilitating the rapid growth of black holes in the early universe, which defies conventional explanation. Along the way, this work developed a new flavor of model to describe these disks, accounting for the pressure support provided by feedback from disk-embedded stellar-mass black holes, developed a number of semi-analytical estimates for how stars might evolve within these accretion disks, and estimated the typical timescales for objects to move through the disk. Together, these estimates showed that accelerated supermassive growth in the early universe was indeed feasible, although this estimate hinged on a number of yet-untested assumptions. Subsequently, this dissertation advances to the question of how stars evolve when embedded within hot, dense disks of gas accreting onto supermassive black holes. Moving beyond the semi-analytical models of the preceding section, the third chapter reviews simulations of stellar evolution subject to the extreme conditions within these accretion disks. Stellar evolution calculations, due to the enormous spatial and time-scales involved, are virtually always restricted to one spatial dimension. This chapter investigates a number of the ways to account for the deviations in spherical symmetry inherent to accretion disks in these calculations, before reviewing how stellar rotation and the chemical composition of these accretion disks can affect the evolution of stars embedded therein. This work developed analytical criteria governing different regimes in stellar evolution, such as the balance between the stellar accretion and nuclear burning timescales, the relationship between gas composition and gas opacity, and the limiting effect of the central supermassive black hole's gravity on stellar accretion as the two compete for gravitational influence on the gas within the disk. Ultimately, the precise, quantitative details of these simulations depend on the specific 3D-inspired prescriptions implemented, but the overall trends identified are robust. The final study presented in this dissertation investigates the feasibility of these accretion disks as the host sites for the stellar-mass black hole mergers detected by the Laser Interferometer Gravitational-Wave Observatory. One of the primary uncertainties of this scenario is whether binaries formed within the disk will tend to spiral inward after formation, or instead be driven via hydrodynamic interactions to spiral outward to the point where chaotic three-body interactions would separate the binary. To address the feasibility of this gravitational wave progenation channel, we conducted three-dimensional hydrodynamical simulations of black hole binaries embedded within these accretion disks, at orbital separations slightly smaller than the limit for dynamical instability. This chapter focused on initially circular binaries over a range of orbital inclinations with respect to the midplane of the disk, finding that binaries with orbits at all misaligned with the disk midplane are gradually realigned, and that retrograde binaries can inspiral appreciably faster than prograde ones. Although the simulations were physically incomplete, in particular neglecting magnetohydrodynamic and radiative effects, they suggest that AGN disks could indeed host the binary black hole mergers detected via gravitational waves.Item Analyzing the Dynamics of Biological and Artificial Neural Networks with Applications to Machine Learning(2024) Srinivasan, Keshav; Girvan, Michelle; Biophysics (BIPH); Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The study of the brain has profoundly shaped the evolution of computational learning models and the history of neural networks. This journey began in the 1940s with Warren McCulloch and Walter Pitts’ groundbreaking work on the first mathematical model of a neuron, laying the foundation for artificial neural networks. The 1950s and 60s witnessed a significant milestone with Frank Rosenblatt’s development of the perceptron, showcasing the potential of neural networks for complex computational tasks. Since then, the field of neural networks has witnessed explosive growth, and terms like “Artificial Intelligence” and “Machine Learning” have become commonplace across diverse fields, including finance,medicine, and science. This dissertation explores the symbiotic parallels between neuroscience and machine learning, focusing on the dynamics of biological and artificial neural networks. We begin by examining artificial neural networks, particularly in predicting the dynamics of large, complex networks—a paradigm where traditional machine learning algorithms often struggle. To address this, we propose a novel approach utilizing a parallel architecture that mimics the network’s structure, achieving scalable and accurate predictions. Shifting our focus to biological neuronal networks, we delve into the theory of critical systems. This theory posits that the brain, when viewed as a complex dynamical system, operates near a critical point, a state ideal for efficient information processing. A key experimental observation of this type of criticality is neuronal avalanches—scale-free cascades of neuronal activity—which have been documented both in vitro (in neuronal cultures and acute brain slices) and in vivo (in the brains of awake animals). Recent advancements in experimental techniques, such as multi-photon imaging and genetically encoded fluorescent markers, allow for the measurement of activity in living organisms with unparalleled single-cell resolution. Despite these advances, significant challenges remain when only a fraction of neurons can be recorded with sufficient resolution, leading to inaccurate estimations of power-law relationships in size, duration, and scaling of neuronal avalanches. We demonstrate that by analyzing simulated critical neuronal networks alongside real 2-photon imaging data, temporal coarse-graining can recover the critical value of the mean size vs. duration scaling of neuronal avalanches, allowing for more accurate estimations of critical brain dynamics even from subsampled data. Finally, we bridge the gap between machine learning and neuroscience by exploring the concept of excitatory-inhibitory balance, a crucial feature of neuronal networks in the brain, within the framework of reservoir computing. We emphasize the stabilizing role of inhibition in reservoir computers (RCs), mirroring its function in the brain. We propose a novel inhibitory adaptation mechanism that allows RCs to autonomously adjust inhibitory connections to achieve a specific firing rate target, motivated by the firing rate homeostasis observed in biological neurons. Overall, this dissertation strives to deepen the ongoing collaboration between neuroscience and machine learning, fostering advancements that will benefit both fields.Item Mechanical evolution of small solar system bodies(2023) Marohnic, Julian Charles; Richardson, Derek C; Astronomy; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This dissertation presents a series of studies that develop and apply numerical modeling techniques to small bodies in the solar system. We are particularly interested in low-energy deformations, collisions, and disruptions, and our subjectsrange from near-Earth asteroids to Kuiper belt contact binaries in the farthest reaches of the solar system. We use the N-body code pkdgrav to investigate these processes and describe our significant additions to its capabilities. Our first subject is the Kuiper belt object Arrokoth. On January 1, 2019, the New Horizons spacecraft flew within 3,550 km of Arrokoth, returning the first in-situ images of a small body in the classical Kuiper belt. Arrokoth was found to be bilobate, with a distinctive contact binary structure. We use pkdgrav to investigate the origins of Arrokoth's striking shape and find that plausible formation mechanisms are quite limited. We rule out the possibility of a direct impact between two unbound objects and put forward an alternate scenario in which two cometesimals in a close, synchronous orbit gradually spiral in toward one another before meeting in a gentle merger. We conclude by exploring implications for the formation of small Kuiper belt objects more generally. Next, we describe our work modifying pkdgrav to accommodate non-spherical particles. Prior work in granular physics has established that particle shape is an important factor governing the behavior of granular bodies like small solar system objects. Irregular particles tend to interlock with one another, inhibiting bulk motion and adding to the shear strength of a medium. We adapt pkdgrav's existing soft-sphere, discrete element contact physics model to allow for modeling of non-spherical grains. We then apply this new capability in three, small-scale proof of concept studies of spin-up, tidal disruption, and the Brazil nut effect. We find a significant difference in behavior when comparing small rubble-pile bodies composed of spherical particles and those composed of non-spherical particles. Finally, we apply our newly-developed tools to a more comprehensive investigation of particle shape in tidal disruption simulations. We construct small rubble piles from a range of differently-shaped constituents and subject them to simulated tidal encounters with the Earth. We conduct a parameter sweep across different encounter geometries and constituent shapes and conclude that particle shape is a significant contributor to tidal encounter outcomes. The role of particle resolution is also investigated.Item NUMERICAL ACOUSTICS FOR PHYSICAL AND SIMULATED ENVIRONMENTS(2023) Kaneko, Shoken Eckhart; Duraiswami, Ramani; Gumerov, Nail A; Computer Science; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Computer modeling and numerical analysis of acoustical phenomena have important applications including manufacturing, audio technologies in immersive multimedia, and machine learning systems involving audio. The focus of the present dissertation is the exploration of numerical methods for modeling, simulating, synthesizing, estimating, processing, controlling, and analyzing acoustical phenomena in the physical world as well as its applications to the virtual world, i.e. immersive technologies for creating virtual, augmented, and extended realities.The dissertation is structured as follows. In chapter 1, I introduce some fundamentals and basic concepts of numerical acoustics and discuss existing practical problems in acoustics. In chapter 2 and chapter 3, I propose two novel techniques for three-dimensional sound field capturing end encoding for immersive audio applications, which are both based on (semi-)analytical cancellation of scattering caused by microphone arrays mounted on acoustic scatterers. In chapter 4 and chapter 5, I introduce a fast algorithm for synthesizing acoustic impulse responses in large-scale forests, and use it to predict the performance of acoustic wildlife monitoring systems based on large-scale distributed microphone arrays. In chapter 6, I propose a novel general-purpose individual-agnostic binaural localizer which supports sound source localization from arbitrary directions without a priori knowledge of the process generating the binaural signal. In chapter 7 and chapter 8, I develop frameworks for regularized active sound control, using either point- or mode-control and using either distributed or local worn loudspeaker and microphone arrays with applications including speech privacy, personal active noise control, and local crosstalk cancellation with limited noise injection into the environment. In chapter 9, chapter 10 and chapter 11, three numerical methods for evaluating integrals arising in the (fast multipole accelerated) boundary element method are introduced. In chapter 9, a recursive algorithm is developed which allows efficient analytical evaluation of singular and nearly singular layer potential integrals arising in the boundary element method using flat high-order elements for Helmholtz and Laplace equations. In chapter 10, a differential geometry-based quadrature algorithm is developed which allows accurate evaluation of singular and nearly singular layer potential integrals arising in the boundary element method using smooth manifold boundary elements with constant densities for Helmholtz and Laplace equations. In chapter 11, an algorithm for efficient exact evaluation of integrals of regular solid harmonics over high-order boundary elements with simplex geometries is developed. In chapter 12, I discuss future research directions and conclude the dissertation.Item MULTI-FIDELITY PARAMETRIC SENSITIVITY FOR LARGE EDDY SIMULATION(2023) Oberoi, Nikhil; Larsson, Johan Prof.; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Designing engineering systems involving fluid flow under uncertainty or for optimality often requires performing many computational fluid dynamics (CFD) calculations. For low-fidelity turbulence modeling simulations such as Reynolds-averaged Navier-Stokes (RANS), such a framework has been established and is in use. However, for high-fidelity turbulence-resolving simulations such as large eddy simulations (LES), the relatively high computational cost of even a single calculation hinders the development of such a framework. The overarching goal of this work is to aid LES in becoming a usable engineering design tool. In this thesis, a computationally affordable approach to estimate parametric sensitivities of engineering relevant quantities of interest in an LES is explored. The method is based on defining a RANS problem that is constrained to reproduce the LES mean flow field. The proposed method is described and assessed for a shock/boundary layer interaction problem, where the shock angle and wall temperature are considered variable or uncertain. In the current work, a proof-of-concept of the proposed method is demonstrated. The method offers qualitative improvements to the sensitivity prediction of certain flow features as compared to standalone RANS simulations, while using a fraction of the LES cost. Different cost functions to infer auxiliary RANS variables are also examined and their influence on the sensitivity estimation is assessed. Overall, the results serve as an important proof-of-concept of the method and suggests the most promising path for future developments.Item Development and Application of Solid-Liquid Lattice Boltzmann Model for Phase Change Material in Heat Exchanger(2022) Chen, Dongyu; Radermacher, Reinhard; Riaz, Amir; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Phase change materials (PCMs) are widely used in thermal energy storage systems, as they can absorb and release a large amount of heat during the phase change process. Numerical simulations can be used for parametric studies and analysis of the thermal performance of the PCM heat exchanger (HX) to produce an optimal design. Among various numerical methods, the lattice Boltzmann method (LBM), a mesoscopic approach that considers the molecular interactions at relatively low computation costs, offers certain key advantages in simulating the phase change process compared with the conventional Navier-Stokes-based (NS-based) methods. Moreover, LBM is ideal for parallel computing, by which numerical analysis can be efficiently performed. Therefore, a comprehensive solid-liquid phase change model is developed based on LBM which is capable of accurately and efficiently simulating the process of convective PCM phase change with and without porous media in both Cartesian and axisymmetric domains. Double distribution functions (DDF) coupled with a multi-relaxation-time (MRT) scheme are utilized in the LBM formulation for the simulation of the fluid flow and the temperature field. A differential scanning calorimetry (DSC) correlated equation is applied in LBM to model enthalpy, by which the solid-liquid interface can be automatically tracked. The source term in the MRT scheme is modified to eliminate numerical errors at high Rayleigh numbers. Moreover, the conjugate thermal model is adopted for the consideration of heat transfer fluid (HTF) flow and conducting fins. The new model is verified and validated by various case studies. The results indicate that the new model can successfully predict the process of PCM phase change with errors confined to less than 10\%. Parametric studies are then performed using the validated model to quantitatively evaluate the effect of convection on PCM melting, from which the acceleration rates (\(a_c\)) of PCM melting and the threshold Rayleigh numbers (\(Ra_{dc}\)) at various aspect ratios are defined and quantified. Furthermore, PCM melting in porous cylindrical HX is also investigated. The results indicate that the acceleration of melting could reach 95\% compared to that in pure PCM at 60\% energy storage. Moreover, the negative effect of uneven temperature distributions on thermal performance of the HX caused by convection is quantified and analyzed. A modified cylindrical HX that offsets this negative effect by varying the geometry is also evaluated. The results indicate that the modified geometry can successfully enhance heat transfer and balance the uneven temperature distributions.Item Why Do Rebels Split? Examining The Causes Of Rebel Group Fragmentation(2022) Stern , Moran; Telhami, Shibley; Government and Politics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Why do rebel groups undergo fragmentation? While extensive research about the consequences of rebel fragmentation exists, research on the process of fragmentation remains relatively nascent. This dissertation collects three papers on the causes of rebel group fragmentation. In the first paper, I develop a junior cadres-based explanation of fragmentation. I argue that in a centralized rebel group, factions will emerge when leaders block junior cadres’ access to senior decision-making bodies. Junior cadres who want to influence the organization’s politics therefore face a choice between remaining within the rebel group and exiting it. Factionalizing is a way to redress grievances by aggrieved junior cadres who deem peaceful mechanisms for upward mobility ineffective. Using original datasets and personal interviews, I find strong evidence supporting my argument in the case of Palestinian Fatah. In the second paper, I argue that the solution to the question of fragmentation lies in rebel socialization—specifically, military training (MT). MT increases group cohesion by strengthening horizontal bonds among combatants; vertical bonds between combatants and commanders; and members’ institutional bonds to the organization’s overall mission and esprit de corps. Members become mutually dependent, thus making splintering more costly and fragmentation less likely. I test this argument on a global sample of 83 rebel groups active between 1989 and 2010. I find that rebel groups that have recently conducted MT are less likely to fragment by about 75 percent. In the third paper, I explore the effect of foreign fighters (FFs) on rebel fragmentation, examining a number of mechanisms derived from previous research. First, I explore how reduced group dependency on local fighters, preference divergence, strategic disagreements, and member segregation increase the likelihood of fragmentation for rebel groups that recruit FFs. Second, I posit that if the foreignness of FFs in relation to local insurgents makes fragmentation more likely, then rebel groups that recruit coethnic FFs will be less likely to experience fragmentation. I test these arguments on a global sample of 227 rebel groups active between 1989 and 2011. I find that rebel groups that recruit FFs are significantly more likely to fragment, even after accounting for the endogenous choice of rebel groups to recruit FFs. Against my expectations, I find that the recruitment of coethnic FFs does not diminish the probability of fragmentation. This finding raises questions about the value of ethnic homogeneity in the context of FFs in particular.Item BUILDING KINETIC MODELS FOR COMPLEX SYSTEMS WITH ARBITRARY MEMORIES(2022) Tsai, Sun-Ting; Tiwary, Pratyush; Physics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Analyzing time series from complex dynamical systems in nature is a common yet challenging task in scientific computation since these time series are usually high-dimensional. To apply our physics intuitions to these dynamical systems often requires projecting these time series to certain low-dimensional degrees of freedom, which often introduces complicated memory effect. A simplest and classic example can be a 2-dimensional coupled differential equation. When one only looks at one of the Cartesian coordinates, one loses the predictability to predict what will happen next given the current 1-dimensional coordinate. The well-known solution is to describe the solution using the eigenvector, and the coupled equation is decoupled into a constant and a 1-dimensional memoryless equation. However, it can be imagined in a more complicated system we may have to look back to more time steps in the past, and it can be impossible to obtain a simple 1-dimensional eigenvector. In this work, we examine such memory effect within time series generated from Langevin dynamics, Molecular Dynamics (MD) simulations, and some experimental time series. We also develop computational methods to minimize and model such memory effects using statistical mechanics and machine learning. In recent years, MD simulation has become a powerful tool to model complex molecular dynamics in physics, chemistry, material science, biology, and many other fields. However, rare events such as droplet formation, nucleation, and protein conformational changes are hard to sample using MD simulations since they happen on the timescales far away from what all-atom MD simulation can reach. This makes MD simulation less useful for studying the mechanism of rare event kinetics. Therefore, it is a common practice to perform enhanced sampling techniques to help sample rare events, which requires performing dimensionality reduction from atomic coordinates to a low-dimensional representation that has a minimal memory effect. In the first part of this study, we focus on reducing the memory effect by capturing slow degrees of freedom using a set of low-dimensional reaction coordinates (RCs). The RCs are a low-dimensional surrogate of the eigenvector in the example of coupled equations. When describing the system using RCs, other dimensions become constant except fast randomly fluctuating noise. These RCs can then be used to help reproducing correct kinetic connectivity between metastable states using enhanced sampling methods such as metadynamics. We demonstrate the utility of our method by applying them to the droplet formation from the gaseous phase of Lennard-Jones particles and the conformational changes of a small peptide Ace-Ala3-Nme. The second part of the study aims at modeling another type of memory coming from intrinsic long-term dependency induced by ignored fast degrees of freedom wherein we utilize one of the fundamental machine learning techniques called the recurrent neural network to model non-Markovianity within time-series generated from MD simulations. This method has been shown to work not only on the molecular model of alanine dipeptide but also on experimental time series taken from single-molecule force spectroscopy. At the end of this second part, we also improve this method to extrapolate physics that the neural network had never seen in the training dataset by incorporating static or dynamical constraints on the path ensemble it generates.