Browsing by Author "Hsu, Tian-Jian"

Now showing 1 - 7 of 7
  • Thumbnail Image
    Item
    Causality between fluid motions and bathymetric features
    (Journal of Fluid Mechanics, 2022-02-07) Hsu, Tian-Jian
    Can morphodynamic problems be solved using a first-principles approach in multiphase fluid mechanics? This is the holy grail for many sediment transport researchers but has yet to be achieved in practice. Using a fully resolved direct numerical simulation for turbulent flow over a bed of spheres, the study of Scherer et al. (J. Fluid Mech., vol. 930, 2022, A11) investigates the onset of morphodynamics from a statistically flat bed. The study shows that the formation of streamwise-aligned sediment ridges is due to large-scale turbulent streaks in the logarithmic layer, which drives local sediment sweeps and bursts. The study provides a solid physical justification for introducing initial perturbations in other reduced-complexity models and opens up new perspectives for simulating sediment transport and morphodynamic problems using high-fidelity models.
  • Thumbnail Image
    Item
    Entrainment and Transport of Well-Sorted and Mixed Sediment Under Wave Motion
    (Journal of Geophysical Research: Oceans, 2022-08-08) Rafati, Yashar; Hsu, Tian-Jian; Calantoni, Joseph; Puleo, Jack
    Entrainment and suspension of sediment particles with the size distribution similar to a range of natural sands were simulated with a focus on the vertical size sorting and transport dynamics in response to different wave conditions. The simulations were performed using a two-phase Eulerian-Lagrangian model by combining the LIGGGHTS discrete element method solver for sediment and SedFoam solver for the fluid phase. The model was first validated for a range of sand grain sizes from 0.21 to 0.97 mm having well-sorted and mixed (bimodal) size distributions using laboratory oscillatory flow data. Three sediment bed configurations were studied under a wide range of velocity-skewed waves with different wave intensity and skewness. It was found that the bimodal distribution having only 30% of coarse fraction and 70% of medium fraction responds similar to a well-sorted coarse sand configuration. Sediment fluxes of the bimodal distribution were slightly higher than those of well-sorted coarse sand because of the pronounced inverse grading in the bimodal distribution. Furthermore, for the bimodal distribution the medium fraction acted as a relatively smooth foundation underneath the coarse fraction which facilitated the mobilization of the coarser particles. Under high energy wave conditions, the smoothing feature was exacerbated and further caused the formation of plug flow where a thick layer of intense sediment flux was observed. Model results also showed that under high skewness waves, phase-lag effect occurred in well-sorted medium sand which caused lower net onshore sediment transport rates but the effect was significantly reduced for mixed sediments. Key Points: - Transport rates of mixed sand with bimodal distribution are similar to those of well-sorted coarse sand - Plug flow formation depends on the particle size distribution and occurs for the bimodal distribution - Phase lags in sediment entrainment and sediment settling are important for predicting net transport rates Plain Language Summary: Sediment transport driven by shoreward propagating waves depends on the sediment particle size. Generally, coarse particles (greater than 0.5 mm diameter) respond directly to the wave motion due to being entrained and transported near the bed with faster settling whereas medium particles (smaller than 0.3 mm diameter) do not respond directly to the flow field due to sediment entrainment away from the bed and slower settling. Natural sediment in coastal zones has a variety of sediment sizes often classified as well-sorted (nearly uniform sizes) or poorly sorted (mixed sediment sizes). The response of well-sorted sediment particles can be characterized and predicted with a representative sediment diameter. However, the response of mixed sediment depends on the size fractions and the interaction of different size fractions with each other and with the flow field. Well-sorted and mixed sediment particles were simulated using a computational model with conditions representative of normal and storm waves. Mixed sediment with only 30% of the coarse fraction (70% of the medium fraction) responded similar to the well-sorted coarse sediment with slightly higher sediment fluxes due to the inverse vertical sorting (upward coarsening). Additionally, the medium particles serve as a smooth bed underneath coarse particles enhancing sediment entrainment.
  • Thumbnail Image
    Item
    Modeling Lobe-And-Cleft Instabilities on a River Plume
    (Journal of Geophysical Research: Oceans, 2024-05-13) Shi, Fengyan; Simpson, Alexandra; Hsu, Tian-Jian
    Abstract The lobe-and-cleft instability is a widely recognized mechanism leading to along-front structure on density current fronts. Early studies based on laboratory and numerical simulations suggested that the lobe-and-cleft instability is due to convective instability in the nose of gravity currents traveling over a nonslip boundary. Horner-Devine and Chickadel (2017, https://doi.org/10.1002/2017gl072997) reported the presence of lobe-and-cleft instabilities at the Merrimack River, which are generated at the river front in the absence of a no-slip boundary. Hence, the observed lobe-and-cleft instabilities must be due to other mechanisms. In this study, we carried out non-hydrostatic large eddy simulations of a riverine outflow into an idealized 3D domain. With a fine grid resolution of 0.15 × 0.31 m in two horizontal directions and about 0.125 m in the vertical direction, the model reproduced the lobe-and-cleft feature, with the magnitude and size of lobes consistent with the field observation. The model results revealed that instabilities start from the primary Kelvin-Helmholtz instability, followed by the secondary instability through stretching and tilting, generating counter-rotating streamwise vortices in the plume and at the plume head. The upwelling associated with streamwise vortex cells brings a slower flow to the plume surface, resulting in lobe-and-cleft patterns at the front and positive and negative vertical vorticity at the plume surface. The model also predicted a lobe width of about two to three times the plume thickness, consistent with the field observation and the lobe/cleft spacing associated with pairs of counter-rotating streamwise vortices. Modeled turbulent dissipation rate shows a trend of exponential decay from 10−4 to 10−3 m2/s3 at the frontal head to 10−7 to 10−6 m2/s3 behind the front, similar to the findings in the previous field studies. Key Points A non-hydrostatic large eddy simulation model is applied to reproduce lobe-and-cleft instabilities observed at the Merrimack River Model results reveal that instabilities originate from the Kelvin-Helmholtz instability, followed by the secondary instability, generating counter-rotating streamwise vortices at the front Modeled turbulent dissipation rate shows an exponential decay with increasing distance away from the front, consistent with field measurements Plain Language Summary Density currents are ubiquitous in nature and they play a key role in many important processes, such as weather pattern, ocean temperature and ecosystem, and sediment transport. The lobe-and-cleft instability is a mechanism that leads to along-front structure on density current fronts. These instabilities are prominent features for identifying the existence of density currents and they are also responsible for kinetic energy dissipation and mixing associated with the density currents. Although lobe-and-cleft instabilities have been observed in river plume fronts, their generation mechanisms remain unclear. In this study, we used a computer model to simulate the phenomena in an idealized domain similar to field observation. The model was able to reproduce the lobe-and-cleft feature that was observed in the field. We found that the instabilities were initiated from the primary Kelvin-Helmholtz instability and were followed by the secondary instability through stretching and tilting. This generates contour-rotating streamwise vortices in the plume and extends to the plume head. The lobe-and-clefts feature is caused by the upwelling associated with these streamwise vortex cells, which bring a slower flow to the plume surface.
  • Thumbnail Image
    Item
    Non-Equilibrium Scour Evolution around an Emerged Structure Exposed to a Transient Wave
    (Journal of Marine Science and Engineering, 2024-06-05) Velioglu Sogut, Deniz; Sogut, Erdinc; Farhadzadeh, Ali; Hsu, Tian-Jian
    The present study evaluates the performance of two numerical approaches in estimating non-equilibrium scour patterns around a non-slender square structure subjected to a transient wave, by comparing numerical findings with experimental data. This study also investigates the impact of the structure’s positioning on bed evolution, analyzing configurations where the structure is either attached to the sidewall or positioned at the centerline of the wave flume. The first numerical method treats sediment particles as a distinct continuum phase, directly solving the continuity and momentum equations for both sediment and fluid phases. The second method estimates sediment transport using the quadratic law of bottom shear stress, yielding robust predictions of bed evolution through meticulous calibration and validation. The findings reveal that both methods underestimate vortex-induced near-bed vertical velocities. Deposits formed along vortex trajectories are overestimated by the first method, while the second method satisfactorily predicts the bed evolution beneath these paths. Scour holes caused by wave impingement tend to backfill as the flow intensity diminishes. The second method cannot sufficiently capture this backfilling, whereas the first method adequately reflects the phenomenon. Overall, this study highlights significant variations in the predictive capabilities of both methods in regard to the evolution of non-equilibrium scour at low Keulegan–Carpenter numbers.
  • Thumbnail Image
    Item
    Numerical investigation of unsteady effects in oscillatory sheet flows
    (Journal of Fluid Mechanics, 2022-06-06) Mathieu, Antoine; Cheng, Zhen; Chauchat, Julien; Bonamy, Cyrille; Hsu, Tian-Jian
    In this paper, two-phase flow simulations of oscillatory sheet flow experimental configurations involving medium and fine sand using a turbulence-resolving two-fluid model are presented. The turbulence-resolving two-phase flow model reproduces the differences of behaviour observed between medium and fine sand whereas turbulence-averaged models require an almost systematic tuning of empirical model coefficients for turbulence–particle interactions. The two-fluid model explicitly resolves these interactions and can be used to study in detail the differences observed experimentally. Detailed analysis of concentration profiles, flow hydrodynamics, turbulent statistics and vertical mass balance allowed the confirmation that unsteady effects, namely phase-lag effect and enhanced boundary layer thickness, for fine sand are not only due to the small settling velocity of the particles relative to the wave period. The occurrence and intensity of unsteady effects are also affected by a complex interplay between flow instabilities, strong solid-phase Reynolds stress and turbulence attenuation caused by the presence of the particles.
  • Thumbnail Image
    Item
    Proceedings of the 2024 DARWIN Computing Symposium
    (Data Science Institute of the University of Delaware, 2024-02-12) Hsu, Tian-Jian; Bagozzi, Benjamin E.; Eigenmann, Rudolf; Jayaraman, Arthi; Totten, William; Wu, Cathy H.; Blaustein, Michael; Blinova, Daria; Carney, Lynette; Huffman, John; Smith, Samantha; Zhang, Jiaye
    The DARWIN Computing Symposium 2024—sponsored by the Data Science Institute (DSI) of the University of Delaware—was held on February 12, 2024. It represented the fifth event in a series of Symposia motivated by a National Science Foundation (NSF) MRI Award, also known as the Delaware Advanced Research Workforce and Innovation Network (DARWIN). As part of an NSF Major Research Instrumentation award (OAC-1919839), DARWIN focuses on catalyzing "research and education at the University of Delaware (UD) and partners by acquiring a big data and high-performance computing system and making this instrument available to the community." In an effort to identify and advance future computing needs for artificial intelligence, to reduce the overhead for domain scientists utilizing HPC, and to develop regional partnerships, this fifth DARWIN Computing Symposium more specifically featured a panel and a keynote talk, as well as a series of research talks on DARWIN-enabled research, on computational and data-intensive (CDI) research/training needs, and on AI-focused CDI research more generally. These talks highlighted the use of AI in HPC to advance sciences and predictive capabilities with societal relevance across a wide range of domains. A panel discussion then facilitated interactions between research software engineers and domain scientists with an eye towards advancing scientific progress in different disciplines. In addition, 30 poster presentations by students and postdocs highlighted a number of relevant CDI research projects. Alongside the NSF and the Data Science Institute, the 2023 DARWIN Computing Symposium was sponsored by Tech Impact, UD’s Delaware Environmental Institute, UD’s Center for Applied Coastal Research, UD Information Technologies, and the University of Delaware Faculty Senate. Dr. Tian-Jian Hsu, University of Delaware Professor of Civil & Environmental Engineering and Director of the Center for Applied Coastal Research served as chair of the 2024 DARWIN Computing Symposium.
  • Thumbnail Image
    Item
    Selection of vortex ripple dimensions in sinusoidal oscillatory flows. Part 1. Ripple dimensions and fluid kinematics
    (Journal of Fluid Mechanics, 2023-04-10) Yue, Liangyi; Hsu, Tian-Jian; Horner-Devine, Alexander R.
    Subaqueous vortex ripples in equilibrium are characterized by their unique geometry and dimensions. Motivated by the recent direct numerical simulation study of oscillatory turbulent flow over a wavy bottom by Önder & Yuan (J. Fluid Mech., vol. 858, 2019, pp. 264–314), the objective of this study is to further investigate the fluid dynamical controls that determine the distinctive equilibrium dimensions of vortex ripples. We use direct numerical simulations to investigate the differences in flow kinetics between sinusoidal oscillatory flow over equilibrium and out-of-equilibrium vortex ripples. In comparison with the equilibrium case, the spanwise coherent vortices, the averaged bottom shear stress on overlying flow and the shear stress distribution on the ripple surface are identified as the key fluid dynamical controls on equilibrium dimensions. Based on these controls, we propose mechanisms in the selection of vortex ripple dimensions. We observe that the flow adjusts in such a way that the interaction between overlying flow and vortex ripples tends to generate the strongest coherent vortices while the ripple surface (or overlying flow) experiences the smallest shear stress averaged over ripple wavelength during the selection process. Through a triple decomposition of the flow, the component of the ripple-induced fluctuation is found to dictate these fluid dynamical controls, which implies that this component plays an important role in the evolution of vortex ripples.